1 99 3
ii
ABSTRACT
This study e xami nes the metacog nitive, cog nitive , social and affective strategy
use of fo reig n lang uage learners who a re studying either in classroom situ atio ns
o r at a d i stance. The i m pact on learn i n g strategy use of a nu m be r of variables is
co nsid e red , re lati ng to 1 ) the lan g u ag e learning co ntext (mode of study, targ et
lang uag e , l evel of study and langu age use o pportu nities) and
2) learne r
charact e ristics (ag e , gender, language learning experience , p ri o r expe rie nce i n
learni n g t h e targ et language, motivation, proficiency) .
2)
a ve rbal report p rocedure , the yoked su bject tech niqu e , ad min iste red to a
su bsample of the questionnai re g ro u p (N=37). Canonical variate an alysis was
applied to the questi on naire data, and instances of strategy use were identified
and classified in the verbal protoco ls usi ng two i ndependent raters.
Resu lts i nd icated that the mai n i nfluences on strategy use we re mode of study
and t h e ag e of l earners; that distance learners were set fu rthe r apart fro m
classroo m learn e rs o n m etacog nitive strategy use measu res when t h e influence
of the targ et lang uag e , p rofici e ncy , p rior target language expe rience and level of
study was conside red; that learn e rs who had had prior expe rie n ce i n learn i n g the
target languag e before e n rolling in a u nive rsity lang uage cou rse we re maximally
disti ng uish ed i n their cog nitive strategy use fro m learners without such p rior
expe ri e nce ; that mode of study exerted some influe nce on cogn itive strategy use,
but this was less than the influe nce of p rior target lang uage experience ; that
diffe re nces i n cog nitive strategy use betwee n learners of Fre nch and learners of
Japanese i n the ve rbal repo rt procedure cou ld not be attributed solely to the
influe nce of the targ et languag e ; and that distance learne rs make less use of
social strateg ies and g reater use o f affective strategies than their class roo m
cou nt e rp a rts.
iii
M ethodological and theoretical i m plicati o n s of the study are p rese nted , and an
appraisal is m ade of the usefu lness of particu lar strategy use m odels for the
investigati o n of lang uage learn ing strategies.
reco m m e nd ations fro m the study are proposed tog ether with suggestions for
fu rther research .
iv
ACKNOWLEDG EMENTS
CONT ENTS
ABSTRACT
ii
ACKN OW L E DG E M ENTS
iv
LIST O F FI G U RES
xii
LIST OF TAB L ES
xiv
Backg rou nd
1 .2
1.3
Distance Educatio n
1 .4
T h e Research P roblem
10
LITERATU RE REVIEW
2.1
I ntroducti on
12
2 .2
E arly Studies
13
2.3
Metacognition
16
in Language Learning
2.4
2 .5
19
22
22
25
26
2 .5 .2 Target Languag e
26
28
31
2 .5 .5 Age
32
2.5 .6 Gender
32
34
35
2.5. 9 Motivation
35
2.5 .1 0 Proficiency
38
Conclusion
41
M ETHODOLOGY
45
2 .6
3. 1
45
3.2
Subjects
46
3.2 . 1 Selection
46
48
3.3
3.4
3 .5
3.6
Questionnaire Study
48
53
53
3.3. 1 Variables
53
55
60
61
62
Instrume ntation
64
64
The Questionnai re
67
67
68
70
71
73
Time
73
Instructions
74
74
74
75
75
76
78
79
79
3.7
Time
79
Instructions
79
Procedures
81
81
82
82
Subject Selection
83
Instrumentation
83
Task Directions
84
84
Subject Selection
85
85
Ecological Validity
86
3 .7 . 3 Reliabi lity
85
86
86
lnterrater Reliability
87
3. 8 Li m itations
87
3. 9 Sum mary
89
90
4. 1
90
4 .2
93
93
94
98
4. 3
103
4 .2 .5 S u m mary
107
109
1 10
114
4. 3. 3 Targ et Languag e
1 17
4.4
1 20
4. 3 .5 Summary
123
124
124
1 25
1 26
C S U Variables
4 .5
4 .6
4.7
129
4. 4.5 Summary
1 33
1 34
1 34
1 36
4.5 . 3 Summary
1 38
1 39
1 39
1 39
S u m m ary
1 40
1 42
5.1
1 42
5.2
1 45
1 46
1 47
1 47
151
1 56
5. 3 . 4 Affective Strategies
1 56
5 .3 .5 Strategy Co m bi nations
1 57
159
5. 3
5.4
Summary
1 62
168
168
on SSU Reports
1 72
on SSU Reports
5.5
1 73
1 74
on M S U Reports
1 76
Summary
1 80
1 82
on S S U Repo rts
1 83
o n AS U Reports
5 .6
Sum mary
1 84
1 85
DISCUSSION
1 87
6. 1
1 87
6 .2
1 91
1 91
1 92
1 93
1 93
1 94
6. 3
Strategy Use
1 95
Strategy Use
1 99
Strategy Use
200
Strategy Use
20 1
6. 4
202
6 .5
205
6 .6
209
209
Strategy Use
6 .7
209
2 10
210
Strategy Use
6 .7 . 3 P roficie ncy and Strategy Use
214
216
6 .8
216
Gender
2 16
2 17
Motivation
2 17
S u m mary
218
CONCLUSION
220
7. 1
220
7.2
222
7.3
224
7. 4
226
S U MMARY
231
APPENDICES
Appendix A
2 39
253
267
Appe ndix D
Sam ple Yoked Subject I nstructi ons
268
Appendix E
Sample Transcri pts of Yoked Subject
Verbal Reports
B I B LI O G RAPHY
270
275
LIST O F FIG U R ES
3.1
3.2
3.3
3.4
3.5
4.1
4.2
4.3
4.4
4.5
4.6
CVA of MSU Variables. Learn e rs u nder thirty and over thirty as G rou ps.
4.7
4.8
4.9
4. 1 0
A Compariso n of Metacog n itive Strategy Use. G rade ' B' Classroom and
Distance Learne rs.
4. 1 1
4. 1 2
4.1 3
4.1 4
4.1 5
4. 1 6
A C o m pariso n of Cog nitive Strategy Use. P rior TL Expe rie nce as G roups.
4. 1 7
4. 1 8
4. 1 9
4.20
4.21
4.22
5. 1
5.2
Categ ori es of Repo rted Strategy Use. Class room and Di stance Learners.
5.3
M e a n Sco res for MSU Categ o ries. Classroom and Distance Learners.
5.4
5.5
Mean Sco res f o r Repo rted Strategy U s e . Learn e rs o f Fre nch and
J apanese.
5.6
Mean Sco res f o r M S U Categ o ries. Learners o f Fre nch and Japanese.
3. 1
3.2
3.3
3.4
3.5
3.6
4. 1
4.2
4.3
4.4
Standardised Coeffici e nts and Corre lation Coefficie nts for MSU
Variables. Age of Lea rn e rs as G roups.
4.5
4.6
4.7
4.8
4.9
4. 1 0
4. 1 1
Standardised Coefficie nts and Co rre lation Coeffi cie nts for MSU
Variables. Grade 'B' Classroom and Distance Le arn e rs as G roups.
4. 1 2
1 . I NTRODUCTI ON
1 .1 B AC KG ROU N D
U ntil the m id-1 970s, a majo r focus i n applied linguistics research ce ntred o n
lan g uage teachi ng methodo logy a n d t h e o ries o f lang uage teach i n g . The possi ble
sig n ificance of learner characteristics such as m otivatio n , learn i n g styles, and the
use of lang uage learning strateg ies was larg e ly ove rlooked. F ro m the mid-1970s
the e m p h asis m oved fro m a conce rn with the m ethods and p rod ucts of lang uag e
teaching to a focus o n the learne r and the actual p rocesses of second o r fo reign
lang uage l earn i ng . The re was a g rowi ng i nq u i ry i nto how language learners
process, sto re , retrieve and use targ et languag e mate rial . Th is n ew e m phasis
invo lved looking at a variety of p rocess factors : the d eve lopment of an
i nterlan g u age (Seli nker 1 972; Selinke r and Lam e ndella 1 9 76) , the ki nds of erro rs
the learn e r makes and the reaso ns fo r them (Richards 19 71, 1 9 74) , the le arn e r's
social and e moti onal adaptation to the new lang uage and c u lture (Sch u mann
19 76, 1 978) , and the co mmu nicati o n strateg ies l earn ers use when faced with a
gap between com m u nicative need and l i ng uistic re pertoi re (Taro ne 1 977; Ca rder
1 983; Faerch and Kasper 1 9 83 ; Tarone 1 983) . As Larse n - Free man and Long
( 1 991 :7) point out, the new d i rection of research i nto lang uage acqui siti on
p rocesses was motivated largely by a desire to u nderstand how some learners
manag e to succeed i n acquiri ng a second languag e and why others fail to do so.
One d i m e nsion of this research i nvo lved atte m pts to fi nd o ut how lang uage
learn e rs manage thei r learning and also to ide ntify the strateg ies they use as a
means of i mprovi ng targ et lang uage compete nce. Research i nto language
learn i n g strateg ies had the u lti m ate ai m of simplifying the rat h e r i mposing task of
lang u ag e learn i n g . This is reflected in an early article by Caro l Hose nfe ld
( 1 976 : 1 1 9) when she u nderlines t h e sig n ificance of he r research i nto languag e
learn i n g strategies b y noting that ' stude nts frequ e ntly fee l t h at foreig n language
learn i n g differs fro m that of othe r subject areas they have e nco u ntered' and that
2
lang u ag e stude nts often meet with new learni n g difficulties. Si nce a languag e is
a hig h ly c o m plex set of system s , st ructure s , and ru les, the p rocesses requ i red to
gai n contro l of this syste m are diffe re nt fro m t hose required in content subjects
such as h i story and socio logy. Learn i n g a lang uage does not merely i nvolve the
u nderstandi ng and m e m o risation of a finite body of public kn owledg e. Rather, a
lang uag e co mprises an i nfinitely variable set of individual performances and the
process of learni ng a second language requ i res that each person learns to
master h i s/her own pe rformative ro le i n a variety of co ntexts. I n o rder to deve lop
co m pete nce in a second lang u age , learners must deve lop a range of language
learning strateg ies wh ich are appropri ate fo r the acquisition of a complex system.
(The term 'second lang uag e' is used throug hout the t hesis to refe r to any
lang uag e othe r than the native lan g u age and for the pu rposes of this study can
be i nte rpreted as synonymous with 'fo reign lang uage'.)
The qu estion of strategy use by lang uag e learners is significant not sim ply
because of the pecu liar demands of language learn i ng but also because , as
We nde n (1 987a:8) n otes , 'one of the lead i ng educati onal g oals of the research
on learner strategies is an auto no mous lang uage learn e r' . S i nce no set of
classroo m experi e nces can provide learne rs with all the language toke ns and
ski l l s t h e y wi ll requ i re in using the languag e , it is i m portant that learn ers deve lop
se lf-di re ction. To do this they need to know how to learn and how to conti nue
learni n g beyond the experiences provided by formal i n structi on. The issue of
strateg y use in re lation to learn er autonomy wi ll be explored further i n secti on 1 . 3 .
I n the f i e ld o f applied ling uistics , research i nto languag e learn i ng strateg ies has
bee n carried out, fo r the most part, with learners who have reg u lar face-to-face
co ntact with teache rs and thei r l earn ing g roup - that is, those who have access
to reg u la r classro o m i nte ractio n s to support the l earning p rocess. Howeve r not
all lang u ag e learn i ng takes p lace in the classroom co ntext. The long list of
handbooks written over the last century attests to the prevale nce of the co ntext
of private study i n lang uage learni ng (Sweet 1 89 9 ; Cu m mi ng s 1 9 1 6; Crawford
and Leitzell 1 930; C o rnelius 1 955 ; N ida 1 957; Po litze r 1 965; Kraft and Kraft 1 966;
3
Mou lto n 1 96 6 ; Pei 1 96 6 ; Pimsle u r 1 980 ; Rubin and Thompson 1 982 ; E llis and
Si nclai r 1 989) . In spite of the long and widespread history of lang uage l earning
i n contexts which do not i nvolve face-to-face classroom i n st ru cti o n , we know
relatively little about how learn e rs in such ci rcu m stances wo rk with target
lang u ag e materi al i n o rder to develop their langu ag e ski lls.
Afte r a n u m ber of prelimi nary studies had bee n co mpleted concern i ng the
strate g i e s of 'good' lang uage learn e rs Stern ( 1 983:41 2) n oted the need to
i nvestig ate strategy use 'in diffe rent social contexts, unde r d iffe re nt lang uag e
learn i n g co nditions'. Ste rn's suggestio n has recently been re ite rated by O'Mal ley
and C hamot ( 1 990 :224) , who, as a final co m m e nt on t h e n ew directions for
research n ote that 'descri ptive work on strategy use . . . in nonclassroom
e nvi ro n m e nts also needs atte ntion'.
Little , i f any, publi shed research appears to e xist re lating to strategy use by
lan g u ag e learners in the no nclassro o m envi ron ment which characte ri ses distance
educatio n . The abse nce of face-to-face contact means that t h e lang uag e learning
context of the distance language learner is quite diffe re nt fro m that of the learner
who has access to reg ular classroom i nteractions to support the learni ng process.
As Sussex (1 991 : 1 89) poi nts out 'languages are m o re difficult than most subjects
to learn in the distance mode because of the co m p lex co m bi n ation of ski l ls and
info rm ation requi red fo r language m aste ry' . The cu rre nt study h as been motivated
by questions about how distance learners manage to deve l o p ski l ls i n the target
lang u ag e , and how their strategy use compares with the strategy use of
classro o m learners. Thi s i s an i n novative direction of study both i n the field of
distance education and in the field of learn i n g strategy research.
B efo re co nti nui ng to deve lop a discussio n of the proble m researched i n this
study, two key co ncepts need to be exami ned namely, strategy use and di stance
educatio n . The fo l lowi ng section ide ntifies, defi n es and di scu sses the terms
associated with these concepts as they re late to the cu rre nt p roble m .
4
1 .2 LEAR N I NG STRATEGI ES
Bialystock ( 1 9 84) poi nts out that the search fo r a set of strateg i e s which u nderlies
the learn i n g of a seco nd languag e h as been m otivated by a n u mber of co ncerns.
From a psychological perspective , the ide ntification of strategies i s see n to be
i m portant in p roviding access to the m ental p rocesses responsible for acquisiti o n .
From a l i n g uistic poi nt o f view, the de l i neatio n o f strateg ies t e l l s us what
resou rce s students possess as lang uage l earn e rs, what they know about their
learni ng and the degree of variability i n approaches to learni n g . Pedagog ical ly,
the i ntention is to i n struct languag e l earn ers in the strategies that have be en
shown t o be effective fo r others . One resu lt of this di sparity in goals, howeve r, is
'a co nco m itant disparity i n defi niti o n . Strategies, that is, are not always co nsidered
in the same terms where each of these alte rnate goals is co ncerned' (Bialystock
1 984 :37).
Strategies used by lang uag e learners have been refe rred to as 'learn ing
strategi es' ( Rubi n 1 975 ; O'Mal ley , C hamot, Stewner-Manzan ares, KUpper and
Russo 1 985a; O' Malley, Chamot, Stewn e r-Manzanares , Russo and KUpper
1 985b) , 'tech niques' (Nai man, F ro h lich , Ste rn and Todesco 1 978) , 'learning
behaviou rs' (Wesche 1 979 ; Politze r and McG roarty 1 985) , 'tactics' (Seliger 1 983) ,
'cognitive p rocesses' (Rubin 1 98 1 ) , and 'learner strateg ies' (Co hen 1 991 ; James
1 99 1 ) These m u ltiple desig nati ons point to the e lusive natu re of the strategy
.
co ncept .
5
I n the fi e ld of applied linguistics a n u m b e r of differe nt ki nds of strategies have
been the su bject of research and i m portant disti nctions h ave been d rawn
betwee n learni ng , co mmu nication and productio n strategi es (Taro n e 1 977, 1 983 ;
Faerch and Kaspe r 1 983, 1 984 ; Bialystock 1 984 ; Chesterfi e ld and Chesterfi eld
1 985) . T h ese disti nctions wi l l now be considered as part of t h e p rocess of
delimit i n g the field of reference for the term learni ng strategies.
The earl i e st defi n ition of lang uage learn i ng strategies was g i ven by Rubin
( 1 975 :43) who i nte rpreted learn i ng strategies as 'the techniqu es o r devices which
a learn e r may use to acquire knowledg e'. This defi nition 'e njoys the widest
curre n cy today' (Larsen-Freeman and Lo ng 1 99 1 : 1 99), and m ost of the variations
on R u b i n's d efi niti o n re late to the issues of i ntentionality and choice. Fo r example,
acco rd i n g to Wei nstei n and Mayer ( 1 9 86 ) , learning strategies are 'i nte nti onal on
the part of the l earner' (O' Malley and Chamot 1 990 :43) and according to
Bialystock learn i ng strategies are 'optional methods for e xploiti ng avai lable
i nfo rm ati o n to i ncrease the profici ency of second lang uag e learni ng' (Bialystock
1 978:76). I n a m o re recent defi n itio n , Rubin (1 987 :1 9) e laborates as fol lows :
Learn e r strateg ies i nclude any set of operations, ste ps, plans,
routines used by the learn e r to faci litate the obtai n i ng , sto rag e ,
retri eval a n d u s e of i nformati o n (after O'Mal ley et a l . 1 983; B rown
et al . 1 983), that is, what learners do to learn and do to regu late
t h e i r learn i n g .
In s h o rt , language l earning strateg i es are the processes which learners deploy
to learn the targ et lang uage (TL) .
Taro n e ( 1 983 :65) defi nes a co m m u nicatio n strategy as a 'mutual atte mpt of two
i nte rlocuto rs to ag re e on a meaning i n situations where requisite meaning
structu res do n ot see m to be shared' . S he argues that co m m u nication strategies
can be diffe re ntiated from learn i n g strateg i es by the i ntent of the strategy use.
The p ri mary motivati on i n using com m u n ication strateg ies i s the desi re to
co m m u nicate and the i m mediate neg otiation of mean i n g . Learn ing strategies, o n
the oth e r hand, are atte m pts t o develop l i ng uistic and soci oli ngu istic com pete nce
6
i n the TL. T h e desi re to learn the TL i s the motivatio n for strategy use. Willing
( 1 988 : 1 47) d i sti nguishes betwee n the terms as follows :
The g oal of a learni ng strategy is the co m pre h e nsio n ,
i nte rnalizati o n , stori ng, a n d setti n g u p o f data accessib l e to the
learner; whereas the focus of communicati o n strategies is the
successful tran smission and receiving of messages.
The n oti o n of com m u nication strategy can also be disti ng u i s h ed from that of a
product i o n strategy which Taro n e (1 983) characterises as an attempt to use the
linguistic syste m effici ently and clearly with a m i n i m u m of effo rt (such as the use
of fo rm u l aic routines), but which does n ot require the negotiatio n of meaning that
defi nes a communication strategy.
The foc u s in this study wi ll not be on strateg ies used for t h e n egotiation of
meani ng nor on p roduction strategies. Rather, the pri nci pal i nte rest lies in the
lang u ag e learning strategies deployed in a di stance learn i ng e nvi ron ment ,
co m pared to those deployed i n a classroom l earning enviro n m e nt. The particular
characte ristics of the distance learn i ng e nvi ro n m e nt are now discussed.
1 .3 D ISTANCE EDUCATION
A n u m ber of defi nitions have bee n formu lated for distance educati on (Moore
1 973, 1 990 ; Holmberg 1 977, 1 985 ; Keegan 1 99 0 ; Wag ne r 1 990) i n which the
co m mo n compone nts are the p hysical separatio n of teac h e r and learner, the
o rg a ni zation of se lf-study by an i n stitution and the use of com m u n ication devices.
T h ese devices may include p ri nt, audio cassettes, video cassettes , fi l m , radio,
te levisio n , te leco nferenci n g , i nte ractive video and co m p ute r n etwo rki ng (Keegan
1 99 0 ) . While the i nterperson al and com m u nicative aspects of languag e
acquisitio n are difficu lt to manage i n the abse nce of face-to-face co ntact, new
7
tech n olog ies h ave e nabled distance education to expand its methodolog ical
options. The result of this is that it is i nc reasingly possible to p rovide a wide
range of learni ng expe rie nces fo r stu d e nts at a distance.
How does the learni ng context of distance educatio n co mpare to that provided
by ope n learning programmes or courses of p rivate study? Sussex (1 99 1 : 1 78)
defi nes o p e n -access learni ng as a p rog ram m e i n wh ich 'stude nts are able to
contro l m u ch of thei r own access, pace and progress t h roug h learn i ng mate rials,
ofte n as an i nteg rated part of a formal course'. H e goes o n to arg u e that too few
educators and learners appreciate the close links betwee n distance education
and ope n -access learni n g and he conti nues to d raw fu rthe r parallels as fol lows :
B ot h distance educatio n and o pe n-access learn i n g i nvo lve high
levels of stude nt co ntrol and di recti o n , p roblems of assessment and
m o nitori ng , and difficulties of i nteraction and di recti o n . And both , if
well desig n ed and delivere d , al low stude nts to learn i n ways that
m ake reduced demands o n scarce human reso u rces . (Sussex
1 99 1 : 1 8 1 )
Rowntree ( 1 992 :30) makes similar co mparisons when he n otes that since 'the
p h i losophy of open learn i ng is to do with i m provi ng access and le arn er-co ntro l ,
t h e n t h e method (thanks t o self-study mate rials) usually i nvolves some element
of di stance learning'. He also lame nts the fact that the lite rature o n di stance
ed ucatio n rare ly refe rs to the lite ratu re o n open learni ng , o r vice ve rsa.
The m a i n diffe rence between distance education and private study has te nded
to be seen i n te rms of the i nflue nce of 'an educational o rgan ization both in the
plan n i ng and p reparatio n of l earn i ng materials and in the p rovision of stude nt
support se rvices' (Keegan 1 99 0 :44) . Thus the disti nctio n is d rawn at the
i nstitutional leve l , rather than i n te rms of specific effects o n the learn i ng context
of t h e i ndividual. Fro m the poi nt of view of the languag e learn e r there are defi n ite
parallels between the learning co ntexts of p rivate study and distance education
particu larly in te rms of the need for se lf-di rection.
lt h as bee n arg ued that self-directio n and i ndepe nde nce for the learn e r i n
d i stance education resu lt from the separation of the learner from the teacher
8
(Tho mpson and Knox 1 987; Calve rt 1 98 9 ; Moo re 1 990). I n t h e absence of a
classroom e nvi ron ment with reg u lar, paced di rectives fro m the i n st ru ctor, distan e
learne rs h ave to give attention to establishi ng their own set of learni n g behaviou rs
and to s hapi ng and managing the co u rse of their learning. The languag e learning
co ntext of d i stance educati o n , as o pposed to that which invo lves reg u lar face-to
face classro o m contact , requires distance learners to be m o re auto nomous i n
Holec's se nse o f h avi ng t h e 'abi lity t o assu me respo nsibility' fo r their learning
(Holec 1 98 1 :3). The importance of learne r auto nomy in re lati o n to lang uag e
learning h as b e e n well docu m e nted in t h e literature o n applied l i nguistics
(Dicki n so n and Carve r 1 980 ; Holec 1 98 1 , 1 987; Hal lgarte n and Rostwo rowska
1 985; Dicki nson 1 987), but as yet we know relatively little about how learners
deve lop f o reign o r second language ski lls i n a learning context which requires of
them a good deal of self-di rectio n .
9
Cu rre nt co m me ntators o n distance educatio n (Bates 1 99 0 ; Keeg an 1 990 ; Marriott
1 99 1 , 1 992; Sussex 1 99 1 ; Rowntre e 1 992) note that an expansion i n the ro le of
distance language educati on is e nvisag e d by educatio nal plan n i ng authorities,
and in some parts of the wo rld has already co m menced, at a time when distance
educatio nal systems are only beg i n ni n g to featu re in educati o nal lite ratu re .
Within the field of distance education data has been g athered o n facto rs such as
stude nt pe rsistence and student characte ristics (e.g . , Kember 1 989 ; Eisenberg
and Dowsett 1 99 0 ; Pau l 1 990 ; Powe ll, Co nway and Ross 1 990) , but re latively
little atte nti o n has been g iven to t h e p rocesses distance learn e rs use as they
wo rk with the learni ng materials. The p rocess-orie nted research that has been
carried out is based large ly on how stude nts use study mate rials in conte nt-based
subjects, n amely education, psychology, and socio logy ( M o rgan , Tay lor and
Gibbs 1 98 2 ; Clyd e , Crowther, Patching , P utt and Store 1 983; Dodds and
Lawren ce 1 983 ; Marland , Patch i n g , P utt and Store 1 984; Marland , Patching , Putt
and Putt 1 990) . More rece ntly i nvestigati o n s have be en made i nto the learni ng
processes of mathe matics stude nts i n distance education (Harper and Kember
1 986; Knight 1 987; Anthony 1 99 1 ).
Research into lang uag e learn i n g at a distance has also te nded to focus on
course conte nt and on a com pari son of the success rates of distance and
classroom learners (cf. Williams and Sharma 1 988). Lambert ( 1 991 ) poi nts to the
need fo r research i nto aspects of the teachi ng- learn i n g p rocess in distance
foreign language education. This cal l has bee n repeated by Sussex ( 1 99 1 : 1 90)
who arg ues that there is u rgent research to be done on distance-mode lang uage
learni n g and that it is critical 'to establish a place for languages in what will
ce rtai n ly be a major mode of learning within a decade'.
10
1 .4 T H E RESEARCH PROBLEM
The study of strategies used by languag e l earners has evolved i n recent years
(Oxford and C rookal l 1 989). I n spite of t h e usefu l and suggestive research that
has been reported, the area is, as Ske h a n (1 989 :98) poi nts out, 'sti l l at an
embryonic stag e' , with co nflicting results and few hard fi ndings. As a
co nsequence of this there is a wide scope for additional research , particularly i n
re lati on t o t h e setti ng fo r learn i ng whe re ' mo re o rde red co mparison' (Skehan
1 989 : 1 49) is required. Strategy u se i n contexts wh e re lang u ag e learners do not
have access to a classroom e nvi ro n m e nt re m ai ns larg e ly u n explored. I n
particu lar, t h e demands that are placed o n lang uage learn e rs i n distance
educati o n , and the strategies such l ea rn e rs e m ploy to succeed in d eve lopi ng TL
skills h as , to date , been neg lected as an ave nue for research.
The educatio nal i nstitution which provides the setting fo r the cu rre nt study is
Massey U niversity , a dual-mode i n stituti o n which provides i n stru ction to students
through reg u lar face-to-face co ntact ('i nternal' students) and also to stude nts
studying at a distance ('extramu ral' students) . The re lationship between mode of
study and the use of language learn i ng strategies wi l l be i nvestigated in this
study.
11
1 .5 S U M MARY
The i nfl ue nce of the distance education setting o n the process of lang u ag e
learni n g h a s not, t o date , been the su bject of research. I n particu lar we know little
about t h e processes w hich distance fo reig n lang u ag e l earners use to deve lop
t h e i r TL competence. The cu rre nt study explore s the i nflue nce of a d istance
learn i n g e nvi ron ment as opposed to a classro o m learn i ng e nvi ro n ment o n the
learni n g behaviours of foreign language learn e rs . Th e i mpact of mode of study
on strategy use re lative to other variab les such as the age of learn e rs , o r thei r
motivatio n , is also evaluated. I n this, the study is recog nisi ng the n eed, as
expre ssed by Oxfo rd and C rookal l (1 989 :4 1 4) 'to expand lang uage learning
strategy studies to i nclude all the re levant p redicto rs and m ediati ng vari ables,
e.g . , age , sex, motivation'.
The fo llowi ng chapter situates the cu rrent study in terms of the existi ng research
i nto strategy use by lang uag e learners . A review is made of the development of
the co ncept of metacog nitive co ntrol in learni n g theory and of its i nflue nce on
lang uag e learning strategy research. Atte m pts to develop classification schemes
for the strategies u sed by language l earners are considered, wit h particu lar
refere nce to the m etacog nitive , cog nitive , socio-affective model which e m e rg ed
fro m research i n cog nitive psychology. The second half of the chapter focu ses
on research into the re latio nship betwee n learne r variables and strategy cho ice
and between lang uage learning co ntext variables and strategy choice.
12
2. LITERATU R E R EVI EW
The early research i nto st rateg ies used by lang uag e learn e rs te nded to focus o n
t h e characteristics o f highly successfu l lang uag e learners ( Rubin 1 975; Ste rn
1 975; N ai man et al. 1 978). These studies were fol lowed by a n u mber of
i nvestigati ons i nto strategies used by learn e rs of varying p roficie ncy , i ncludi ng
t hose who we re n ot ve ry effective in deve lopi ng ski l l s i n the TL (e.g . , Hose nfe ld
1 977; Bialystock 1 979 ; Politzer 1 983). Subsequ e ntly, the sco pe of lang uag e
learni ng strategy research was wide ned to i nclude an exami nation o f t h e
re lationship betwee n strategy u s e and vari ables such as eth nicity (O'Malley et a l .
1 985a ; O'Mal ley e t al. 1 985b ; Politzer a n d McG roarty 1 985; Tyacke and
Mendelso h n 1 986) , level of study (Co h e n and Aphek 1 98 1 ; Tyacke and
Me nde lso h n 1 986 ; Chamot and Kupper 1 989) and, most rece ntly , g e nder
(Ehrman and Oxfo rd 1 989 ; Oxford and Nyikos 1 9 89 ) . While these attempts were
being m ade to exami ne strategy use and to ide ntify influe nces o n strategy use,
researchers we re also developi ng an i nterest i n the feasibi lity of trai n i ng stude nts
in strateg i es fo r languag e learning (Pressley, Levi n and Delaney 1 982 ; H e n n e r
Stanch i n a 1 986 ; O'Malley, Russo, Chamot and Stew n e r- Manzanares 1 988 ;
Chamot and Kuppe r 1 989) .
A n u m be r of i m po rtant studies which h ave been co n ducted i nto lang uage learning
strategy use wi l l be prese nted in this ch apter. Th e studies, which wi l l be
descri bed chro n o log ically, fal l i nto thre e g roups. The fi rst set (section 2 . 2) were
mai n ly carried out i n the 1 970s and are m o re explo rato ry i n n ature (Carto n 1 966;
Ste rn 1 975; Rubin 1 975 , 1 98 1 ; Hose nfeld 1 976 ; Naiman et al. 1 978). The second
g roup of studies in section 2.3 i ncludes a selective review of research conducted
i n the 1 980s. I n this section the i nvestigation of t h e metacog n itive dimensions of
strategy use is re p resented i n the wo rk of We nde n ( 1 986a, 1 986b) and of
13
O'Mal ley et al. (1 985a, 1 985b ) . The various approach es to defi ning and
classifyi ng strategies used in lang uage learn i ng are co nsidered (section 2 . 4) and
key studies relating to the infl u e n ce of a n u mbe r of facto rs on strategy choice are
descri bed and eval uated (section 2 . 5). A synthesis of the findings of th ese key
studies leads to a final statem e nt regarding the ratio nal e for the p resent study .
i n the form of an e xtended rese arch article published by Carton (1 966) e ntitled
The method of inference in foreign language study. In this study atte ntion was
d raw n to the variation that existed among languag e learners i n te rms of thei r
i nclination to make infe re nces. Carto n also noted that the abi lity to make valid,
rational , and reasonable infe re nces varied g reatly between learn ers. H e posited
that the tolerance of risk would vary with the abi lity to make sound i nferences.
Carton fo l lowed up his study in a second article ( 1 971 ) fu rnishi ng a detailed
discussi o n of i nfere nci ng and of the i nfe re ncing cues avai lable to language
learners .
A simi lar direction fo r research was poi nted t o b y Carro l l ( 1 967) w h o suggested
that the learne r might p rovide a rich sou rce of knowledge conce rning successfu l
lang uag e acquisition. H e advocated compiling and analysi ng case histo ries of
adu lts who had learnt o n e o r more foreign languages.
Eight years late r Stern ( 1 975) and Rubin (1 975) explored the noti on that
successfu l languag e learn e rs were effective because of the particu lar learni ng
be haviou rs they e m ployed. This view ran cou nte r to the mai n th read of li nguistic
thought at the time that effective lang uag e learne rs sim ply had an i n here nt ability
fo r language learn i n g , were more m otivated , or had had exte nsive exposure to
n atural lang uage learn i ng situatio n s , m ost p referably in a country where the
lang uag e was spoke n (O'Malley and Chamot 1 990).
14
By means of obse rvation and i nte rviews with learn e rs and teachers Rubin ( 1 975)
ide ntified seve n b road strategi es used by successful language learn ers.
E xamples of these are that the good language learner i s a wi lling and accu rate
g uesser, h as a stro n g d rive to co mmu nicate, p ractises, and attends to m eani ng.
Rubin also sugg ested that learne r strategies vary with task, learn i ng stag e
(beg i n n i n g , i ntermediate and advanced) , age , co ntext (classroom ve rsus n atural
e nviro n m e nt) , i ndividual styles, and cultu ral diffe rences. The i nflu e nce of a
n u mbe r of these factors on strategy choice formed the di recti o n of m uch futu re
research i nto language learn i ng strateg ies (see section 2.5).
At the same time as Rubi n's study, Ste rn (1 975) compi led an i nflu e ntial list of te n
characte ristics of successfu l lang uage learn e rs based o n observatio n and
i ntuiti o n . He obse rved that successful learn e rs exhibited such traits as a
sustai ned search fo r meaning , a wi lli ngn ess to p ractice and experi m e nt , and an
active approach to the learning task.
Ste rn's specu lati o n s i nspi red a g roup of researchers to u nde rtake a larg e scale
study of the characte ristics of good language learners. This work was carri ed out
by N ai man et al. ( 1 9 78) and has beco me known as the 'To ro nto' study of g ood
lang uag e l earners. The successful language learners who were the subjects for
the study had been identifi ed through co lleag u e s fro m the u n ive rsity e nvi ro n ment
and we re mostly hig h ly educated people. N ai m an et al. e licited i nformati o n from
the su bjects (N=34) usi ng multiple data col lection procedu re s , which i ncluded
o bservatio n s , i nte rviews and question nai res . Five pri mary strateg ies we re found
to be co m m o n to all good language learn e rs : an active task appro ach, reali zation
of l ang uag e as a system, realization of lan g uage as a means of co m m u n ication
and i nteractio n , m a nage me nt of affective d e m ands, and m o n itori n g of seco nd
lang uag e pe rfo rmance. Among their m o re i nteresting co nclusions was that
lang uag e success appears to be attributable n ot so much to an 'i n n ate gift' o r an
'ear' fo r languag e , as to constant effort and i n volve ment with t h e languag e , eve n
to the e xtent of c re ating useful opportu nities t o practise the languag e . This study
marks the fi rst attempt to use a m u lti method d esig n to e licit data o n strategy use.
15
F u rthe r prog ress was made i n the development of research techniques for
investig ati ng learn er strategies in the work of Hose nfeld (1 976 ) . Hosenfeld
i ntroduced the 'think-aloud' i ntrospective p rocess to ascertai n the strategies used
by 25 ju nior high schoo l learners of Fre nch whi le perform i n g lang uag e learn i ng
tasks. Th e results i ndicated that stude nts cou ld ide ntify and ve rbalize their
strateg ies, and that teachers' assu mptions about thei r stude nts' strategies were
often wro n g . H ose nfeld (1 976 : 1 27) p rese nts seve ral cautionary stateme nts i n the
co nclusions to her study i ncludi ng the fo llowi ng :
lt m ust be rem e m be red that the procedu re d oes not have a long
h isto ry of widespread use (in fo re i g n language education or
e lsewh e re). As a research tool it does not meet the t radition al
requi re m e nt for scie ntific rigour.
Si nce Hosenfeld's g rou nd-breaki ng study, a variety of ve rbal report m easu res
have been deve loped and used in studi es which have added to our
understandi ng of strategy use (Cohen and Aphek 1 981 ; H ose nfeld 1 977, 1 984;
Block 1 986 ; Cavalcanti 1 987 ; Cohen and Cavalcanti 1 9 87; Mangub hai 1 99 1 ). The
'yoked subject technique' em ployed in the present study is o n e such verbal re port
measu re which has recently been developed to investigate strategies used by
languag e learne rs (Nayak, Hanse n , Krueger and Mclaug hli n 1 990) .
A fu rther means of gatheri ng data on strategy use was explored by Rubi n (1 981 )
as she co l lected repo rts o n strategy use t h rou g h directed d iary studies. The
subjects fo r the study, you ng adu lt learners, were give n explicit instructions o n
how to keep t h e di ary . Rubi n repo rted that s o m e stude nts we re bette r able to
descri be strategies than others. In analysi ng t h e results a disti nctio n was d rawn
between strategies that contribute di rectly to language l earn i ng (e.g . , g u essi n g ,
deductive reasoning) and those which help i ndi rectly (e .g . , m o nitoring , creating
opportu nities to practise).
At aro u n d the same time that these early studies we re being conducted i nto the
strategies used by language learn e rs, a n u m b e r of insig hts were e merg i n g from
the wo rk of researchers i n the fie ld of cog n itive psychology which were to have
a m arked i mpact o n ways of co nceptualisi n g strategy use. The co ncept of
16
?r
17
deciding in advance to attend to specific input; monitoring or correcting one's
language production.
The term metacognition gained wide currency and proved to be the conceptual
starting point for literally hundreds of studies into memory strategies, study skills,
attention and reading. In the field of applied linguistics a number of studies drew
upon research into metacognition conducted in the field of cognitive psychology
and used this theoretical basis for research into language learning strategies
(see, for example, O'Malley et al. 1 985a; Wenden 1 986a, 1 986b; Garner 1 988a,
1 988b; Carrell 1 989 ; Carrell, Pharis and Liberto 1 989). These studies provided
support for the powerful role that metacognition plays in different learning
situations.
Unfortunately, in spite of all the empirical research that has been done,
metacognition has remained a concept that is difficult to define. Brown (1 980)
contended that the primary problem with the metacognition construct is that it is
very difficult to separate from cognition. Many researchers and educators have
wondered if metacognition is merely a set of cognitive skills 'elevated and
dignified with a new title' (Brown 1 978:7). The prediction made by Wertsch
(1 977:5) that 'as we pursue problems in this area further and further the
distinctions between cognitive and metacognitive abilities will become less and
less clear' has proved to be very apt. A number of authors have suggested that
it is difficult to separate metacognition from general aptitude or cognition (e.g. ,
Wertsch 1 977; Cavanaugh and Perlmutter 1 982; Slife, Weiss and Bell 1 985;
Jacobs and Paris 1 987; Garner and Alexander 1 989; Cohen 1 991 ).
As Sparks ( 1 988:209) points out 'the literature ... does not always make a clear
distinction between the higher-level metacognitive strategies that manage
comprehension and the lower-level cognitive strategies that carry out the
executive decisions'. (For a definition and discussion of cognitive strategy use see
section 2.4. 1 and section 2.5. 1 .) An activity such as summarising has been called
metacognitive in one study (Brozo, Stahl and Gordon 1 985) and cognitive in
18
another (Mosenthal 1987). The difficulties which are encountered when
attempting to demarcate the boundary between metacognitive strategies and
cognitive strategies will be taken up again later in the thesis.
While the parameters of the metacognitive concept are still under dispute, a
number of researchers have identified the importance of metacognitive control in
language learning (O'Malley et al. 1985a, 1985b ; Wenden 1986a, 1986b, 1987a,
1987b; Duran 1987; Carrell 1989; Carrell et al. 1989). For example, in an early
study by O'Malley et al. (1985a:24), it was found that 'students without
metacognitive approaches are essentially without direction and ability to review
their progress, accomplishments, and future learning directions'. Holec (1987)
argues that it is the use of metacognitive strategies that distinguishes learners
who are consumers of language courses from those who are actively involved in
their own learning programs and who are self-directed. Galloway and Labarca
(1990:144) note that learners who do not use metacognitive strategies, or who
use them infrequently, 'often equate . .. the process of learning with 'being taught'
- that is, doing what the teacher and teaching materials say to do'.
Skehan (1989:95) describes the investigation of metacognition in language
learning as 'the most exciting development in recent strategy research'. The first
studies to explore the role played by metacognitive processes in language
learning were carried out by Wenden (1986a, 1986b) and O'Malley et al. (1985a,
1985b). Wenden's work explored the dimension of metacognitive knowledge in
language learning, while O'Malley and eo-workers focused on the other dimension
of metacognition, namely the 'executive processes' or self-management
processes used by second language learners. These studies are now reviewed.
19
2.3.1 The I nvesti gation of Metacog nition in Lan g uage Learn i ng
Ink
20
about language learning and about ways to learn a language (e.g. , Ellis and
Sinclair 1 989; Willing 1 989; Oxford 1 990).
The other main research into metacognitive aspects of language learning was
carried out by O'Malley and eo-workers (1 985a, 1 985b). They investigated the
self-regulatory aspects of metacognition as part of an examination of the
strategies used by language learners. The subjects for their studies were 70 high
school-age students enrolled in ESL classes. Data was collected through
inteNiews with, and observations of, students carrying out typical classroom
activities, such as pronunciation exercises, oral drills, vocabulary learning, and
language use.
From the findings it was very clear that inteNiews were much more productive
as a means of data collection than obseNations. The fact that obseNations were
not very productive in revealing the strategies used by language learners was
consistent with the findings of earlier researchers (Naiman et al. 1 975 ; Rubin
1 981 ; Cohen and Aphek 1 981 ). From the inteNiew data 23 strategy types were
identified. The researchers found that it was possible to define and organize
these strategies within the basic classification scheme proposed by Brown and
Palincsar (1 982) consisting of metacognitive and cognitive strategies. The
following definitions are given for these terms by O'Malley et al. (1 985b:560-1 ) :
Metacognitive strategies involve thinking about the lerning process/
planning for learning onitoring of comprehension or production while
it is taking place, and self-evaluation of learning after the language
activity is completed. Cognitive strategies are more directly related to
individual learning tasks and entail direct manipulation or
transformation of the learning materials.
A third category of strategy use onsisting of social mediation strategie as
added and this was 'most clearly evidenced in cooperative learning' (O'Malley et
-- ---
al. 1 985b:56 1 ).
_t!l'
--
21
strategies were identified, examples of which are inferencing, or working out the
meaning from the text, and elaboration, or relating new information to other
concepts in the memory. The two social strategies were co-operation, involving
working with peers, and questioning, involving asking the teacher or native
speaker for repetition or clarification. In addition a distinction was drawn between
metacognitive strategies used for planning, monitoring and evaluating learning.
Results of the study showed that beginning and intermediate students used more
cognitive strategies (especially repetition and note-taking) than metacognitive
ones. Of the metacognitive strategies almost all were related to planning. Few
social strategies were used. The classification scheme was used and developed
in later studies (Chamot 1 987; Chamot and O'Malley 1 987; O'Malley et al. 1 988;
O'Malley, Chamot and KUpper 1 989; Chamot and KUpper 1 989).
O'Malley and colleagues continued with a number of descriptive and longitudinal
studies of strategy use as well as studies of learner training in strategy u s/'
Aspects of their researc '} wllich are directly relevant to the issue s.-Hfv8 stigated in
this study are explored in section 2.5.
22
2.4 D EFI N ITION AND CLASSI FICATION
2.4.1 The Devel o pment of Classification Schemes
23
Dansereau (1 978, 1 985). In Oxford's classification, the class of primary strategies
includes three groups of strategies (memory strategies, cognitive strategies,
compensation strategies), while the class of secondary strategies contains three
other groups of strategies (metacognitive, affective, social). The six strategy
groups are subdivided into a total of 1 9 strategy sets. The entire learning strategy
system includes 62 strategy types.
This classification scheme was used to generate items to tap strategy use in the
form of a 1 21 -item questionnaire, The Strategy Inventory for Language Learning
(SI LL). This questionnaire has been used in a number of studies including those
relating gender and motivation to strategy use. In addition, the questionnaire has
been used to enable learners to develop a profile of the kinds of strategies they
use in learning a new language (Oxford 1 990).
The two major criticisms of Oxford's classification scheme relate to how the
scheme was developed and to the absence of any links between that particular
scheme and important insights from the field of cognitive psychology. These two
criticisms are interconnected. In developing a classification scheme which
subsumed every strategy that had been cited in previous learning strategy
research, Oxford's extended listing was 'far removed from any underlying
cognitive theory' (O'Malley and Chamot 1 990 :1 03). This is a serious limitation of
Oxford's scheme since over the last fifteen years a number of information
processing models have emerged alongside studies and theories in cognitive
psychology as to the role of cognitive processes in learning (e.g., Anderson 1 983,
1 985; Gagne 1 985; Wagner and Sternberg 1 985; Weinstein and Mayer 1 986;
Mayer 1 988).
Apart from Oxford's classification scheme, the other main theoretical framework
to have emerged over the last fifteen years has been developed from the work
of Brown and Palincsar (1 982). This taxonomy produced by O'Malley and eo
workers (mentioned in section 2.3. 1 ) comprises three main categories of strategy
use: metacognitive, cognitive and socio-affective, depending on the kind and level
24
o/
Cognitive strategies are more directly related to individual learning task han
metacognitive strategies and involve the manipulation or transformation of the
material to be learned. The following definition of cognitive strategies is given by
Rubin (1 987 :23) :
Cognitive strategies refer to the steps or operations used in learning
or problem-solving that require direct analysi0ransformation or
synthesis of learning materials.
Chamot and O'Malley (1 987 :242) state that while engaging in cognitive strategy
use the learner:
interacts with the material t
earned by manipulating it mentally (as
in making mental images or relating new information to previously
acquired concepts or skills) or physically (as in grouping items to be
learned in meaningful categories or taking notes on or making
summaries of important information to be remembered).
p/
tj
25
control over affect' (O'Malley and Chamot 1 990:45). An example of social strategy
use is cooperation, or working with peers to solve a problem, to share
information, or to get feedback on a task. An example of affective strategy use
is self-talk, or 'reducing anxiety for learning by using mental techniques that make
one feel competent to do the learning task' (O'Malley and Chamot 1990:45).
o/,
,/o
r)S'd
Research has revealed that variables such as the TL (Politzer 1983; Chamot et
al. 1 987), the gender of learners (Ehrman and Oxford 1989; Oxford and Nyikos
1 989) and the ethnicity of learners (Politzer and McGroarty 1 985) appear to
influence the strategies language learners employ. The present study furthers this
research by exploring the role played by four language learning context variables
(mode of study, TL, level of study, language use opportunities) in the strategy
choice of undergraduate foreign language learners. In addition, the influence of
particular learner characteristics (age, gender, language learning experience, prior
experience in learning the TL, motivation, proficiency) on strategy choice is
investigated. The relative impact of these variables, and particular combinations
of these variables, on strategy use is explored in the context of undergraduate
foreign language study.
26
This section presents research into factors which are related to strategy choice.
The discussion is organised in ten sections, each one dealing with prior studies
of the relationship between strategy choice and the variables relating to the
learning context or learner characteristics which will be explored in the current
study.
2.5.1 Mode of Study
27
co nsisted of three sections, (i) an 1 8-ite m scale o n g e n e ral b e haviours ; (ii) a 1 4item scal e o n classroom behavi o u rs ; and (iii) a 1 9-item scale o n i nte ractions with
others outside the classroo m . Stude nts we re asked to respond in te rms of the
freque n cy with which they e ngaged in each of the behaviou rs on a scale of 4-0.
Fi ndi ngs revealed that the behavio u r scales followed a patte rn that set off French
cou rses from the others. Politze r (1 983:62) poi nts out that o ne possible reason
for the singu lar learni ng behavi o u rs of stude nts of Fre nch could be ascribed to
the fact that ' u n like the cou rses i n t h e Spanish and Ge rman de part m e nts, those
in the French departm ent . . . fo llow ve ry strictly a ' rational, di rect method'
methodology' . The fi ndi ngs also revealed that stude nts of Spanish e ngaged i n
fewe r positive strategies than d i d students o f othe r lang uages. I n summarising the
fi ndi n g s of the study, Po litze r ( 1 983 :62) notes that 'even within similar ove rall
co ntexts, some lang uage learn i n g be h aviours vary sig nificantly acco rding to
lang u ag e and level'.
Howeve r, it is like ly that lang uag e of study i nte racts with a host of other variables
i ncludi n g that of lang uage teachi n g methodology hig h lig hted by Po litzer. Oxfo rd
( 1 989a) suggests that it is the brig hter o r m o re 'strategy-wi se' students who tend
to take Russian rather than S panish , si nce S panish , she clai m s , is pe rceived to
be easi e r fo r E n g lish speakers. F u rtherm o re , stude nts m i g ht be learni ng the Tls
for different pu rposes, and this could i nfluence thei r choice of strategies.
28
reported usi ng more than twice the nu mber of strategies t han students of
Span i s h , and that they used strategies the other g roup did n ot m e ntion (e.g . ,
re h earsal , s u m m arisi ng) , cou ld not automatically b e attri buted sole ly to the
i nflue n ce of the TL acco rding t o the researchers. At least two other possi ble
effect s could n ot be ru led out. Firstly , it is possible that di rect i n structio n by the
teach e r could e nco u rage stude nts to learn i n particular ways , and h e nce i nflue nce
t h e i r strategy choice. Secondly, the o bjectives of a particular cou rse cou ld
i nflue nce strategy choice. In a subseq u e nt lo ngitudinal study of t h e same fo reign
lan g u age stu d e nts (Chamot, KOpper and l mpink- Hernandez 1 988a, 1 988b) the
o bjectives of t h e course we re see n to play a ro le i n influe nci ng how stude nts
learnt the TL. In reviewi ng the study O' Malley and Chamot (1 9 9 0 : 1 40) note :
A classroom emphasising the g rammatical structure of the foreign
langu ag e and an analytical co m parison of the TL to the native
languag e fosters strategies such as deduction and translatio n . On the
other hand a classroom focusing on proficie ncy fosters strategies such
as i nfere nci ng and su bstituti o n .
Does learni n g a languag e at the e le m e ntary stage requi re the use of strategies
which are different fro m those used at i ntermediate o r advanced levels of
lang uag e learni ng? Is it possible fo r learne rs to use a st rategy such as
i nfe rencing o n ly when they have reached a particular stag e of co m pete nce in the
TL? Are diffe re nt strategies differe nti ally appropriate at diffe rent stages of
lang uag e learni ng? A nu mber of studies have add ressed the question of the
effect of the leve l of study o n strategy use by language learners.
Tyacke and Mendelsohn ( 1 986) carried out a research project with seve n
learne rs of E ng lish each drawn fro m a different leve l of proficiency i n an eight
level pro g ra m m e . The learners kept diarie s of their 'learni ng development' on a
daily basis ove r a matte r of weeks thoug h the exact du rati on of the project is not
i nd icated. Tyacke and Mendelso h n ( 1 986 : 1 76) report the resu lts as fo llows :
E xcerpts from the diaries of lowe r leve l students i ndicate a much
g reater dependence on the teacher, and o n the ling uistic code, as we
wou ld expect than those of the m o re advanced student.
29
The study was not sufficie ntly focused i n te rms of the i n st ructio n s g ive n to
learners to p rovide i nsig hts i nto specific i n stances of st rategy use at diffe re nt
levels of lang uage study. I n addition the fi ndi ngs were rat h e r g e n e ralised and
precise defi niti o n s or examples of these g e n e ralisations were n ot give n . Fo r
example, it i s not clear what was meant by 'depende nce o n the li ng u istic code' .
H oweve r the resu lts of the study are suggestive t h at strategy use at the lower
level diffe rs fro m strategy use at more advanced levels.
McDo nou g h and McN erney ( repo rted by Tyacke and M e ndelso h n 1 986)
co mpared notes on vocabulary learni ng strategies in two classes of diffe rent
p rofici ency , low-i ntermediate and advanced , usi ng Rubin's ( 1 981 ) Observation
Oxford and Nyikos in their 1 989 study m easu red the level of languag e study in
te rms of the n umber of years spent study i ng the target fore i g n lang uage. With a
30
sample of 1 ,200 u nderg raduate foreig n l ang uag e learners they e licited reports on
strategy use t h rough the Strategy Inventory for Language Learning. This 1 2 1 -item
i nstru m e nt asked learn e rs to i ndicate i n a m u lti-choice fashion the freque ncy with
which they u sed certai n lang uag e learn i ng strategies. The researchers fou nd that
foreig n lang uag e students who h ad stu died the new lang uag e fo r a mini m u m of
fou r o r five years used strateg ies far more often than did less experienced
lang uag e learners . More precisely, stude nts who had studied for at least five
years used fu nctional practice strategies such as attending fore i g n language fi lms
and readi ng authentic material i n the TL sig nificantly more ofte n than did less
expe rie nced stude nts. These m o re frequ e ntly used strategies all required
lang uag e practice i n natural settings o utside the lang uage classroom. Stude nts
who had studied fo r at least five years also used more conversatio nal/i nput
e licitation strateg ies such as requ esti ng slowe r speech and aski ng for
pronu nciation co rrectio n .
Cohen and Aphek's study (1 9 8 1 ) i ndicated that advance m e nt i n cou rse leve l o r
in years of study d i d not necessarily mean that stude nts used better strategies
in eve ry instance . The subjects for the study we re 1 9 native Eng lish-speaki ng
stude nts learning Hebrew i n Israe l. The stude nts were d ivided into three leve ls
of profici ency : begi nners (N=9), i ntermediate stude nts ( N =6 ) , and advanced
students (N=4). The researchers atte m pted to fi nd out what insights cou ld be
g ai ned about g ood and bad co m municative strateg ies i n t h e classroom from
e m pi rical obse rvati on coupled with verification by the stude nts the mselves. Cohe n
and Aphek ( 1 981 :233) reported that 'both g ood and bad co m m u nicative strateg ies
appeared across c lass levels, and were used by better and poorer stude nts'.
They also e m p hasised that the research 'raised some real q uestions as to what
'good' strateg ies consist of' (Co hen and Aphek 1 98 1 :233) noting that such
questions warranted further i nvestigati o n .
31
Oxford goes o n to suggest t h ree possible reasons fo r this. T h e fi rst is that
stude nts ' might spontaneously develop new and bette r strategies as t hey become
more advanced' (Oxfo rd 1 989 :237). Secondly, there i s the possi ble effect of task
requ i re m e nts changing as students move to higher leve l cou rses, thus prompti ng
stude nts to respo nd with strategies tai lored to the task req u i re m e nts. Thirdly,
there i s the q uestion of attrition rates si nce 'stude nts with poo re r strategies might
perform worse than stude nts with better strategies and t h e refore d ro p out of
lang uag e study b efo re reachi ng higher level co urses' (Oxfo rd 1 989 :237) .
Oxfo rd's discussio n of changes i n strategy use ove r time is significant, but what
co n stitutes a 'bette r' strategy is by no means o bvious. lt is not appro p riate at this
point to atte m pt to investigate fu rther the value judg eme nts which have been
attached to strategy use i n various studies (e.g . , Cohen and Aphek 1 98 1 ; Politze r
and McGroarty 1 985) si nce the present study is concerned wit h ide ntifying the
strategies students use rather than evaluati ng whether they are g ood o r bad .
Therefo re Oxfo rd's discu ssion is pro bably bette r co nsidered i n terms of the use
of diffe re nt st rateg ies at diffe re nt levels rather than in terms of the evaluative
labels that are used.
The wo rk of Bi alystock ( 1 979) and Huang and van Naerssen ( 1 987) revealed the
i m po rtance of fu nctional practice outside fo rmal classro o m requ i re m e nts fo r
languag e learni n g . What has not been i nvestigated is whet h e r opportu nities to
use lang uag e , beyond those provided by the cou rse, have an effect o n the ki nds
of strateg i e s stude nts choose to use in the context of p rivate study. That is, do
students who can reg u larly practise the TL with native spe akers , or friends,
e m ploy diffe rent strateg ies to those u sed by stude nts who have n o such fu rther
opportunities? The question of the
relationship between
extra p ractice
opportu niti es and strategy use wi l l be exami ned i n the p resent study.
32
2.5.5 Age
There h as been little di rect researc h on the i nte ractio n between the choice of
lang u ag e learning strategi es and the age of learne rs . Skehan ( 1 989 :97) poi nts
out that 'we cu rre ntly have to d eal with studies which h ave not syste matically
cove red the age range of learne rs' . Instead , strategy use studies have focused
o n particular ag e g roups. For exam ple, y o u ng chi ldre n at school e ntry ag e were
the su bjects for the studies of Wo ng-Fi l l mo re (1 979) and Chesterfi eld and
Chesterfield ( 1 985) . High school stude nts were the subjects i n other studies (e.g . ,
O'Malley et al. 1 985a, 1 985b ; Ramirez 1 986) as well as adults and u niversity
stud e nts (e.g . , Nai man et al. 1 978 ; Bialystock 1 979 ; Rubin 1 98 1 ; H uang and van
Naerssen 1 985 ; Politze r and McG roarty 1 985 ; E h rman and Oxfo rd 1 989) .
2.5.6 Gender
Wo men are ' better' language learne rs than men acco rdi ng to lang uage learni ng
folklo re. H oweve r, possible gender diffe re nces i n strategy use by lang uage
learners h ave received relatively scant attentio n . Four studies to date have
add ressed this question .
Politzer (1 983) , i n h i s q uesti on naire-b ased study of the lang uage learn i ng
behavio u rs of u nde rg radu ate language stude nts (detai led i n sectio n 2 . 5 . 2) , found
statistically sig nificant gender diffe re nces on his i nteracti o n behavi o u r scale (p <
33
.05). Po litze r does not attempt to explai n t h e sig nificant gender diffe re nces, and
attaches re latively little i m po rtance to t h e findings re lated to g e nder differences:
Even wit h i n simi lar ove ral l contexts, some lang u ag e learni ng
behavi o u rs vary sig n ificantly acco rdi n g to language and l evel (vari ance
due to t h e sex of the learner s e e m s relatively mi nor, but does exist
with regard to such vari ables as social inte raction). (Po litzer 1 983:62)
The Oxfo rd and Nyikos study ( 1 989) descri bed i n secti o n 2 . 5 . 3 revealed that
wo men u sed languag e learning st rategies significantly m o re ofte n than men i n
three of five possi ble strategy facto rs : formal ru le-related practice strategies (such
as analysi ng wo rds, finding si mi lariti es betwee n languag es, g e nerati ng and
revising ru les) , gene ral study strateg ies (such all-purpose techniques as studyi ng
hard , i g n o ri ng distracti ons, being pre pared) and co nversational/i n put elicitational
strategies.
In reviewing the fi ndings of Oxfo rd and Nyikos (1 989) and of E h rman and Oxford
(1 989 ) , Oxfo rd (1 989a :238) sugg ests that the sex differe nces ' might have been
associated with women's g reate r social o rie ntation , stronger ve rbal ski l ls, and
g reater co nform ity to n o rms, both linguistic and academic, demonstrated by
earlier research'.
34
m etacognitive , cog nitive , socio-affective model. As Oxford , Nyiko s and E h rman
(1 988 :327) n ot e :
I nstru m e nts s hould clearly e xamine . . . other ki nds of strateg i e s , such
as cog nitive , metacog nitive and co m pe nsatory .
A necdotal evide nce suggests that o nce a person has studi ed seve ral lang uages,
the process of language learning beco m es e asier. This expectation i s based upon
the hypothesis that whe n adu lts learn a new language they 'wi l l approach tasks
with the strateg ies and be haviour that t hey co nsider productive , and these
strategies wi l l be drawn from past experie nce' ( Ramsay 1 980 :90). One re latively
u n explored approach to strategy use by language learne rs is to i nvestigate
wheth er lang uage learn ing experience has an effect on the strategies learners
use.
A n u mber of studies i n the fi eld of cog nitive psychology ( Ra m say 1 980 ; N ation
and Mclau g h l i n 1 986 ; Nayak et al. 1 990) have explored the strateg ies used by
'expert' as opposed to 'novice' lang uag e learners. The fi ndi ngs of t hese studies,
however, are n ot applicable to the cu rre nt study in that the novice l earn ers were
m o nolingual s ubjects and the exp e rt learn ers were bili n g u al or multili ngual
subjects. The expert learners h ad been raised in bilingual or multi li ngual
e nviro n m e nts and possessed an advanced , native-like leve l of proficie ncy .
Other research i n t h e fi eld o f cog n itive psycho logy has e stablished that t h e
amou nt of knowledg e possessed by the learne r has a substantial i mpact o n the
learning p rocess (e.g . , Chi, G lase r and Rees 1 982). In additi o n , i ndividuals who
know a g reat deal about so m ething encode new mate rial related to that
knowledge i n a diffe re nt way to that of i ndividuals who know little about the topic
(e . g . , Chi et al. 1 982 ; Ande rso n 1 983, 1 985). A perti nent question fo r the cu rre nt
study is w h ether students who are al ready experie nced in foreig n languag e
learning e m p loy diffe re nt strategies to those used by stude nts who are e mbarki ng
o n fo reig n lang uag e learni ng for the fi rst time.
35
The sing le piece of publi shed research which s heds some light on this question
comes from the longitudi nal study of C hamot and KOpper ( 1 989) . In a brief
discussion of the study O'Malley and Chamot ( 1 990) reported t h at novice foreign
lang u ag e learn e rs someti mes panicked w h e n they realized that t hey lacked
procedu ral ski lls for so lvi ng language problems. Expert learn ers , who had already
studied another foreign languag e , approached new l ang uage tasks cal m ly and
we re able to d eploy procedu ral ski lls developed in other lang uag e learning
situ atio ns.
When learne rs e n rol in a fo reig n languag e cou rse t hey may d iffe r in terms of
wheth er they h ave already had some experie nce in learni n g the TL. Some
learners may h ave already spe nt time in the host count ry , others may have
studied the lang u age at schoo l , throug h n i g ht classes, or t h rough co ntact with
native speake rs. Research effo rts h ave not be e n explicitly di rected towards
exploring whether the presence or abse nce of prior experie nce in the TL
i nflue nces strategy choice. Re lated to this q uestion is the effect of leve l of study
on strategy choice as di scussed in section 2.5.3.
2.5.9 Motivati o n
36
they h ave not been successfu l i n the past, m ay h ave d eveloped a
negative attitude toward thei r ability to learn .
Most of the research and debate su rrounding m otivati o n and lang uage learn i ng
strategies relates to the i mportance of m otivatio n i n strategy i n structio n . For
e xample , an ESL trai ni ng study (O'Mal ley et al. 1 985b) clearly showed the
i mportance of motivation i n learn i ng strategy instructi o n . Since the wi l l to learn
appears to be esse nti al for deve loping the ski l l to learn (Pari s 1 988) , the
researchers concluded that the success of learni ng strategy i n structio n i s
dependent o n , among other factors, t h e abi lity t o provide a motivati onal
framework that can co nvi nce students of the valu e o f learni ng strategies. While
research i nto the effectiveness of strategy inst ruction has very o bvious
importance , it is beyo nd the scope of the prese nt study. The ro le of m otivati o n
wi l l be explored i n re lati o n t o choice o f lang uag e learning strategies, rather than
in relation to the success of strategy trai ning prog rams.
Motivation is a ve ry broad concept. A usefu l disti nction can be drawn between the
leve l or i ntensity of m otivatio n and m otivatio nal o rie ntation , as these co nce pts
re late to strategy choice.
m otivatio nal level o n strategy choice. Bialystock and Frohlich (1 978) conducted
a study of the strategy use of 1 57 hig h school stude nts learni ng Fre nch as a
seco nd lang u ag e . Four factors we re exami ned fo r their effects o n the various
types of language achievement. The fou r factors were
aptitude , fi e ld
i ndepe ndence, attitude and strategy use. Four tests measuri ng ach ieve m e nt were
used
and
described
as
fo rmal/o ral ,
formal/written,
fu ncti onal/o ra l ,
and
37
strategies. Bialystock (1 979 :392) prese nts the fi ndings in a l ater discussion as
follows :
the use of these strategies is re lated pri mari ly to the attitude of the
l a n g u age learner and u n re lated to his language l earn ing aptitude. Thus
it i s t hose learners who are particularly m otivated to m aster the
languag e who engage i n these strateg ies.
I n Oxfo rd and Nyikos' (1 989) study (detai led i n secti on 2 . 5 . 3 ) , it was found that
of all t h e variables they measured (such as proficie ncy rati ng , gender, years of
study) the deg re e of expressed motivatio n to learn the lan g u age was the most
powe rfu l i nfluence on strategy choice. Motivation had sig n ificant effects o n the
te nden cy of students to use (or not use) strategies in fo u r o f the five factors :
formal ru le-related practice strategies, functio nal practice strategies, g e n e ral study
strate g i e s , and conversatio nal i n put elicitation strateg ies. The more moti vated
stu d e nts used these fou r ki nds of learning strategies more often than did less
m otivated stude nts.
O n e d ifficu lty with the Oxford and Nyikos study is that an o pe rational defi nition
of m otivation is not given, neither is there any i ndicati on of how motivation was
m easu red. Only a very general description of the backg ro u nd questio nnaire is
give n :
I n addition to the S I LL, we also administered a backg ro u nd
question naire cove ri ng sex, years of fo reig n lang u ag e study, e lective
vs. requ i red cou rse status, self-perception of p rofici e ncy and
motivatio n , and other to pics. (Oxford and Nyikos 1 989 :293)
W h i l e the operationalizations of some of the variables, such as g e nder, are fai rly
o bvious, in the case of self-perception of proficiency and motivation, this is n ot
t h e case . Were students si mply asked 'How motivated are you'? The lack of
u nambiguous defi nitions and clear operationalizatio n s of a variable such as
m otivation is a major weakness of this study.
The effects of m otivational o rientation (instru m e ntal vs. i nteg rative) o n reported
strategy have not been an explicit focus fo r research , though the study of E h rman
38
and Oxford ( 1 989), described in sectio n 2 . 5 . 2 , p rovides some i nsig hts. E h rman
and Oxfo rd fou nd much m o re frequ ent use of fu nctional practice strategies
among adu lt languag e learn e rs who were learni ng foreign lang uages for career
reaso ns. Th ese learners , who had an instru me ntal motivation to learn a fo reig n
languag e , e m ployed language use strateg i es such as seeki ng out native
s peakers to talk to, reading authe nti c texts and i nitiating conve rsati ons in the new
languag e . One difficulty with the E h rman and Oxford study is that motivati onal
o rie ntati o n is i nferred fro m the career of the su bjects and so the fo llowi ng
co n necti o n s are made :
the professio nal lang uag e trai ners may have both i ntegrative and
i nstru me ntal drives fo r learni n g (and teach i n g ) lang uages . . . Fo r
stud e nts who are g overnment e m ployees, the motivatio n for learn i ng
the other lang uag e is likely to be i n stru me ntal in nature ; fo r their
spou ses, it may be either i nstru me ntal o r i nteg rative . (Ehrman and
Oxfo rd 1 989 :9)
Motivati o n al o rientation , then, is not m easured but is infe rred from the career of
the su bjects. This is a crude measure of a co mplex co nstruct and is a major
li mitation of Ehrman and Oxfo rd's exploratio n of the effect of motivatio n al
orie ntation on strategy use.
2.5 . 1 0 Proficiency
Th e i m petus for research into languag e learn i n g strateg ies came from a desire
to find out about the particu lar strateg ies used by 'good' , or 'successful' lang uag e
learners . Su bsequently there were fu rthe r atte m pts to co rrelate t h e use of
particu lar strategies with learning effective ness, or proficie ncy. The resu lts of
these studies suggest that m o re proficie nt l earners use a wider range of
strategies, but that the relationship betwee n strategy use and proficiency is
co mplex (Oxford and Crookall 1 989). A re prese ntative sample of studies which
have investigated the relationship betwee n proficie ncy and strategy use wi ll be
prese nted in this section.
Bialystock ( 1 979) developed a fou r-part model of learni ng behavio u rs co m pri sing
i nfere n ci ng , monitori n g , formal practice and fu ncti onal practice. The effects of
39
these fou r catego ries of strategy use on proficie ncy were exami n ed in a study of
high scho o l stude nts learning French. She fou nd that fu nctio n al practice , which
occu rs w h e n learne rs increase t h e i r opportunities to use the languag e fo r
com m u n i cation (by g oi ng to the m ovies, readi ng books o r talki ng to native
speake rs , fo r e xample) had a st ro ng e r co rre lati on with ach ieve ment than any
other strategy and pro moted ach ieve ment in both oral and writte n tasks.
Huang and van Naerssen ( 1 987) co m pared the strategies used by successful
Chinese E F L learners ('success' was defi n ed i n terms of o ral co m m u nicative
abi lities) and less successfu l Chi nese EFL learners . The subjects for the study
we re the top and bottom thirds in a g roup of 60 g raduating stude nts from the
Guangzhou Foreig n Lang uages I nstitute. Strategy data we re co llected by means
of a question nai re on learner strategies, and an in-depth intervi ew. The strategies
they g athe red data on were re lated to the Rubin/Stern inve ntories (Rubin 1 975;
Ste rn 1 975) : fo rmal practice, fu ncti onal practice and monitori n g . These strateg ies
had been fou nd to be i m portant i n Bialystock's ( 1 979) study. After Bialystock
( 1 979 ) , the questi o n n ai re emphasised strategy use outside any classroom tuiti o n ,
that i s , opportu niti es fo r lang uag e use which t h e learners themselves chose. A
similar co nditio n i s included i n the present study i n order to e nsure that the
tech niques were the ones the learn ers chose to use.
In a c o mpari son of high and low p roficie ncy g roups, no sig nificant diffe re nces
we re fou nd for the use of fo rmal practice and monito ri n g . Howeve r high-
profici en cy stude nts reported g reater use of fu nctional practice strategies, and the
diffe re n ce betwe e n the two g roups was sig nificant (at the 0.05 level). The fi ndings
of H uang and van Naerssen ( 1 987) co rrobo rated the research reported by
Bialystock ( 1 979) that more successfu l students employed fu ncti o n al practice
strate g i es significantly more often than less successfu l ones. Ske h an ( 1 989 ) ,
howeve r, sugg ests that some caution is requi red when i nterpreting the resu lts of
the H uang and van Naerssen study. Firstly, si nce the subjects were g raduates
of a Foreig n Lang uages I nstitute , one wou ld expect that they we re already
talented languag e learners. Generali zation to other g roups is therefore
40
hazardous, 'since the li nguistic accompli s h m e nts of the low g ro u p m i g ht well be
the e nvy of most people' (Ske han 1 989:92). This n ote of caution cou ld also be
applied to the pre sent study in that u nive rsity students must possess a good deal
of educatio nal e x pertise and abi lity in o rder to reach the te rti ary level of study.
A more i mportant issu e , relating to all t h e studies to date which have exami ned
the effect of p roficie ncy on strategy use, concerns causality. G reater strategy use
might lead s o m e stude nts to higher levels of perfo rmance. Equally, higher
performance m i g ht faci litate the use of more strategies (Ske han 1 989). One
aspect of the Huang and Van Naersse n research lends s o m e su pport to the
profici ency-causes-strategies explanati o n , when t hey re port that :
althoug h certai n u nsuccessfu l stude nts . . . attempted to adopt the . . .
tech niques used by their more successfu l pee rs , they fo und that those
techniques were not very helpfu l i n their cases. ( H u ang and van
Naerssen 1 987 :293)
A recent stu dy i nto the effect of proficie ncy on strategy choice was carried out by
Chamot and Kuppe r (1 989). The subjects for this study were h i g h school students
of Russian and Spanish (N=67) . The descri ptive and longitudinal phases of the
study e xami ned the strategy use of i n effective versus effective learners. In the
descri ptive study teachers identified 'effective', 'average', and 'in effective' learners.
Stude nts retrospectively repo rted their strategy use throug h g roup i nte rviews.
Analysis of the ve rbal repo rts revealed that stude nts at all ability leve ls used
strategies and were able to talk about them. Howeve r, m o re effective stude nts
used learn i n g strategies more ofte n , and had a wider rep e rtoi re of learni ng
strategies, than did less effective stude nts.
The longitudi nal phase of the study was conducted with a sample of the same
students who had partici pated in the descri ptive study, with ave rag e students
excluded. Learners thought aloud while wo rki ng thro u g h typical lang uage
activities. In general, more effective students used a g reate r variety of strateg ies,
and used them i n ways that helped with the successfu l co m p letion of the task.
Conversely, i neffective students u sed fewer strategies, and also frequently used
41
strategies that we re i nappropriate to the task. Qualitative analysis of strategy use
revealed that effective stude nts were more p u rposeful i n thei r app roach to a task
than i n effective students. I n additio n , they m o n itored their co m prehe nsi o n and
p roductio n for ove rall meani ngfu l n e ss rat h e r than fo r individual co mponents, and
effectively used p rior general knowledg e as well as their linguistic knowledge
whi le worki ng on a task.
2.6 C ONCLUSION
I nvestigations i nto the strategie s u sed by lang uag e learn ers we re origi nally
m otivated by a desi re to i solate the characte ri stics of successfu l learn ers. With
the g rowing recognition that strategy use was a complex phenomenon varyi ng
acco rding to facto rs such as task, level of study and the ethnicity of learners,
m o re recent research has soug ht to ide ntify the influe nces on language learne rs'
choice of strateg ies. A n u mber of fi ndings fro m such studies have bee n presented
i n this chapte r.
In some cases, for example with reg ard to proficie ncy , there appears to be
evide nce fo r quite a strong re lationship betwee n effective ness in lang uag e
learning and strategy choice. S i m i larly, the research sugg ests that the leve l of
languag e study affects the strategies learne rs use, with g reater strategy use
being associated with m o re advanced leve ls of lang uag e study. Th e cu rre nt study
i nvestig ates whether the obse rved diffe re nces in strategy use with reg ard to
profici ency and leve l of study are co nsistent in the context of fo reig n lang uag e
underg raduate students studyi ng i n a dual-mode setti ng.
42
With reg a rd to the i nfluence of the TL and the age of learne rs on strategy use,
a wide ni n g of the research focus is requ i red. Research i nto the ro l e played by the
TL in strategy choice by lang u ag e l earne rs has, to date, bee n confi ned to
situ atio n s i nvolvi n g l ndo-European Tls. The current study e xtends the scope of
previ ous studi es to i nclude a co mparison of the strateg ies used by learne rs of
Asian lang uages (Japanese, Chi nese) and of l ndo-E u ropean languages (Fre nch ,
German). I n response to a call by Skehan ( 1 989 : 1 48) the cu rre nt study makes
an o rd e red co m parison of the strategies used by diffe rent ag e-g roups. Data on
strateg y use i s co l lected fro m learners who vary in ag e fro m late tee ns to
advanced adu lthood , thus syste matical ly coveri ng an age range of al most five
decades.
The mai n research question of this study i s co ncerned with the i nfl ue nce of m ode
o f study, that is, the p rese nce or absence of a classroo m learn i ng e nviro n m e nt ,
43
on strategy choice by u nde rg radu ate foreign language learners. The role of mode
of study in dete rmi n i ng strategy use , re lative to the ro le of oth e r aspects of the
langu age learni ng context (th e TL, level of study, and languag e use opportunities)
is examined. I n additio n , the study e xplores the contribution made by l earn e r
characte ristics (ag e , gender, languag e learn i ng experience , p ri o r co ntact with t h e
T L , motivation a n d proficie ncy) i n dete rm i n i n g the st rategies learn e rs em ploy t o
develop TL co m pete nce .
I n the research published to d ate g e n erally only o n e o r two facto rs have bee n
e xami ned for t h e i r i nflu e nce o n strategy use. I n such studies it is o n ly possi ble
to g auge the relative importance of, say , level of learning as o pposed to g e nder
diffe rences, with oth e r possi bly critical vari ables such as proficiency be i ng absent
fro m the equation. Th u s it has not been possi ble to measu re the re lative
sig n ificance of a number of key vari ables. Larsen-Freeman and Lo ng (1 99 1 )
argue for the n eed fo r m ore co mplex research designs rather than 'si mple
correlations b etween a si ngle i ndividual variable and learner perfo rmance', which ,
i n the case of strategy research , h as be e n re ported strategy use. The cu rrent
study acknowledges that lang uage learning is a co mplex p rocess and seeks to
e xami ne the relationship betwee n strategy use and both learn e r characte ristics
and aspects of the lang uag e learn i n g co ntext.
I n short, this study seeks to ide ntify the factors which exert a strong i nflu e nce on
strategy choice , th ose which exert a lesser i nflu e nce , and those which do not
appear to b e i mportant. lt also attempts to add ress the question of the ro le of
m ode of study re lative to oth e r variables, i n determining the strategies used by
fo reign lang u ag e learn e rs . Once a set of key variables which are associated with
strategy choice has bee n identified, a fu rther question arises : are the two g roups
of learne rs (classroom and distance) homogeneous in the extent to which the key
variables h ave an effect on their strategy choice? I n oth e r words, does mode of
study i nte ract with other key variables to i nflue nce the strategy choice of
languag e learners? This aspect of the p roblem is a further way of examining the
44
i nfluence of the learn i ng e nvi ronment on the operations that learne rs use i n
accomplishing languag e tasks.
45
3. M ETHODOLOGY
This ch apte r describes the m ethods employed to i nvestig ate i nflue nces on the
choice of l anguage learni ng st rategies by u nderg raduate foreign languag e
learners. T h e early part of the chapter describes t h e populati o n a n d subjects fo r
t h e study as well as the vari ables and research questi ons (sections 3 . 1 -3. 3) . The
m i dd l e part is devoted to a discussio n of i nstru me ntati o n , i nstructi ons and
procedures, and m ethods used for processi ng the data (sectio n s 3 .4-3.6). A
critical evaluation of the validity and reliability of the study tog ether wit h its
l i m itations co mprise the final sections of the chapter.
T h e setti ng for the study is Massey U nive rsity , a 'dual-mode' i nstitution offe ri ng
c o u rses of study through an i nte rnal prog ram m e where tuiti o n takes place
t h rough reg u lar classroom co ntact and also , si nce the 1 960s, throug h an
e xtramural programme fo r people who wish to pu rsue u n ive rsity study 'at a
distance' . For the latter g roup tuition takes place larg e ly t h roug h the study
m aterials despatched to learners (wo rkbooks , study guides, taped m aterial) and
t h rough reg ular assign ments. Four fo reign languages are tau g ht through the dual
m ode syste m : Fre nch , German, J apanese and Chi nese.
Distance lang uage l earners have some opportu n ities for classroom learn i n g but
t hese are ve ry li mited. Once a year they m ay attend on-campus vacation cou rses
46
usually lasti ng fro m three to five days . I n t h e case of 1 00-leve l learne rs t hese are
voluntary but fo r the higher leve l s , co m pu lsory.
Classroom learn ers tend to be school leave rs in their late teens who usually
pursue ful l-ti m e study. Distance l earn e rs t e n d to be more matu re i n ag e , mostly
in the 20-40 year age g roup, and usually p u rsue part-ti me prog ram mes due to
family or wo rk co m mitm e nts.
The isolation of distance l earners has a lways been a conce rn and a range of
su pport netwo rks is avai lable i ncluding reg ional co-ordinators, the E xtramural
Stude nts Society , newslette rs and the provision of reg ular opportunities to
telephone cou rse controllers.
Much of the research i nto strategy u s e by lang uage learn e rs has drawn on
re latively small samples of learners (Politze r 1 983, N=37; Po litze r and McG roarty
1 985, N=90; O'M alley et al. 1 985a, N=70 ; H uang and van Naersse n 1 9 87, N=37) ,
with the notabl e exception of the work of Oxfo rd and Nyikos ( 1 989) whose work
on variables affecting strategy use was based on a 1 ,200-person u n iversity
sample. Give n the hig h ly i ndividualised n ature of strategy use (Carre ll 1 989) , it
is desirabl e to g ai n access to a re latively larg e number of subjects particu larly
when i nvestigating the effect of a n u m be r of variables on re ported st rategy use.
3.2.1 Selecti o n
The subjects fo r this study were fo reig n languag e learners at Massey U niversity ,
studyi ng F re nc h , German , Japanese o r C hi nese.
Each year approxi m ately 600 stude nts e nrol as either classro o m or di stance
learners i n fo reig n languag e cou rses. There is a sig nificant attrition rate for
distance lang uage learners, particu larly at the 1 00-leve l , t hat i s , among fi rst year
students. Th i s can be largely attri buted to the de mands of lang u ag e learning in
terms of c o n stant practice and the need to avoid falling b e h i nd with assig n ments.
47
The attritio n rate is hig hest i n the fi rst half of the year. I n selecting subjects fo r
this study it was i m portant to wait u ntil at least three months of the academic year
had e lapsed when one cou ld m o re safe ly ass u m e that the majo rity of learners
partici pati ng i n t h e study were like ly to be conti nuing students.
The study took place with an i ntact g roup, and the mode of se lection was by
vo l u ntary partici patio n . Two p rocedures were used to co llect data and the size of
sample varied according to the procedure. In the case of the questio n n ai re
procedure all c lassroom and distance learners were i nvited to partici pate i n the
study. Fo r the ve rbal report procedure 200- level (second year) learners of French
and J apanese served as subjects . The ve rbal report g roup (N=37) was a
subgroup of the q uestio n n ai re g roup (N=41 7) and co ntai ned both classroom and
distance learn e rs. Partici patio n i n the ve rbal re port procedure was e nti re ly
volu ntary. Learn ers who expressed re luctance to take part, often due to a
p rofessed lack of confide nce i n their 'study habits' , were not i ncluded i n the study .
P e rso nal characteristics of the verbal report subjects, such as ag e and g e nder,
were not co l lected , but the mode of study and TL of each subj ect was recorded.
More detai led biog raphical data was required of the questio n nai re subjects in
o rd e r to i nvestigate the influence of a wider range of variables o n strategy use.
H e nce i nfo rm ation was e licited about a number of learne r characte ristics,
i ncluding age , ge nder, m otivation and prior experience in learning the TL.
I nformation re lati ng to the co ntext of learning was also requested. Th is m eant
aski ng learn e rs about any o pportunities they had to practise the TL beyo nd those
provided by the course. Detai ls of the mod e , leve l of study and TL we re i n se rted
for each s u bject o nce the qu estionnaire was com pleted .
T h e fo l lowing section outlines the characte ristics of the question naire subjects
and of the ve rbal repo rt su bjects.
48
3.2.2 Characteristics of Su bjects
Questionnaire Study
Fou r h u n d red and seve ntee n u nd e rg raduate learners (284 wome n , 1 33 men)
e n rolle d in fore i g n lang uage cou rse s at Massey U nive rsity w e re the su bjects of
the q u estionnaire study. Learne rs were e n rolled to pursu e fo re i g n lang uag e
cou rses t h rough the classroom mode of study (N=1 43) o r t h roug h the distance
mode ( N=274) .
Ag e Distribution of S u bjects
C la ssroo m and Distan ce Lea rners as G ro u ps
"""
80
-40
.,
<S<S
?s:X!
c J1
-- ClaMroorn t.lode
__,
_
_
ll1 - !0
1----
31 - -40
t - 110
Wode
51 -110
10
-----1
Fig u re 3. 1
49
The distributio n of learners acco rdi n g to level of study, mode of study and TL is
prese nted in Fig u re 3.2.
1 20
100
20
FNnoh
Cennn
,._
C'*-
FIWIIIh
study
1 00-level
Ch!MM
o.nnc.t
200-level
I122S
300-level
Fi g u re 3.2
The fig u re shows learners of Japanese formed the larg est TL g ro u p . Chi nese was
n ot offered as a 300-leve l course throug h t h e distance m ode fo r the academic
year i n which the study took place . This is reflected in Fig u re 3.2. lt is also
evid e nt that t h e hig hest leve l of enro l m e nts in fo reig n lang uage pape rs occu rs at
t h e 1 00-level of study.
50
8i]
:s:
88(
18: !SS
w.:
l1.E No ill
Fnnch
liE No LLE
Cerrn<m
LLE No LLE
J<poneae
ill No l1.E
ChlneM
Fig u re 3.3
51
As far a s leve l o f study a n d experi ence i n learn i n g oth e r languages is concerned ,
Fig u re 3 . 4 c learly i ndicates that among l earners w h o had p rog ressed i n their
languag e studies to the 200-leve l and the 300-leve l , an incre asing p ro po rtion of
them h ad h ad so m e experience in learning othe r lang uages.
80
eo
!!
l .ro -
m"
c..
,.,,
0
LLI
lie
1 00-levl
LLI
lie
LLI
2QO-IIMII
lie
LLI
.300-level
Figure 3.4
52
Learne rs were also asked whether they h ad had p rior experi e nce in learni ng the
T L befo re e n rolling at U nive rsity. Taki ng the sam ple as a whole 69. 1 % of l earn ers
i ndicated they had had such experie nce. For t h ose who h ad had such prior
experie nce the majo rity i ndicated this was at school whi le fo r the next largest
g roup this was i n 'ot h e r' ci rcu mstances (e.g . , t h roug h eve ning classes, at
Polytech nic courses) . Overall, distance learn e rs tended to have h ad less
experie nce of learni ng the lang uage at school and m o re of learn i n g it in the host
cou ntry o r i n other ci rcu m stances (see Fig u re 3.5).
Context of P rior T L Experience
1 00
....
0
L.
..8 ..o
E
::I
:s<5<
Figure 3.5
The most freque ntly i ndicated le ngths of ti m e of prior l earn i ng of the language
were less than one year (most notably for l earn ers of Japanese) and ove r a
pe ri od of five years ( most subjects i n this category were learn i n g Fre nch or
German).
53
Verbal Report Study
The ve rbal report study was desig ned to provide co nverg e nt data rel evant to
questions about the i nflu e nce of mode of study and the TL on strategy choice .
Verbal rep o rt su bjects we re both classroom learn e rs and distance l earn e rs d rawn
from 200-level Fre nch and Japanese cou rses. Table 3 . 1 prese nts the distri bution
of the ve rbal re port su bjects i n terms of TL and mode of study.
Table 3 . 1
Dist ribution of Ve rbal Repo rt Subjects
Classroom Mode
Distance Mode
French
15
Japanese
13
Further backg ro u nd i nfo rmation o n the ve rbal repo rt su bjects such as age , g e nder
o r languag e learn i n g e xperie nce was not qu antified. Possible effects of the
u n even dist ribution of t h e verbal re port su bjects accordi ng to mode of study wi ll
be discussed late r in this chapter and in Chapte r 5 . Th e fol lowi ng section outlines
the research desig n of the study.
The p rese nt study co mpares the i nflue nce of a number of variables on the
metaco g n itive , cog nitive, social and affective st rategy use of classroom and
distance language learners. Potenti al i nflue nces o n strategy use we re seen to be
derived fro m the l earn ing context (such as mode of study and TL) and from
particu lar characteristics of learners (such as age , gender and lang uage learni ng
experience).
54
The variab l e s , i ncluded i n the cu rre nt study, are set out below:
Operatio nal definitio n s fo r the variables in the study and the scales used to
measu re th ese variables are now discussed.
55
Strategy Use Variables
Strategy use was operatio nalised t h roug h the model of metacog nitive, cogn itive
and socio-affective strategy use de rived from B rown and Pali ncsar ( 1 982) and
d eve loped by, among othe rs , O'Mal ley et al . (1 985a) , O'Malley et al. (1 985b) ,
O' Malley et al. ( 1 989 ) , Chamot and KUppe r ( 1 989 ) . Th e re were two mai n reasons
why this model was chosen fo r the cu rre nt study.
Particu lar types of strategy use within this g e n e ri c classification were defi n ed and
identifi ed followi ng the strategy descri ptio ns that appear in the lite rature (Chamot
1 987; Chamot et al. 1 988a, 1 988b ; Ellis and S i n clai r 1 989 ; O'Mal ley and Chamot
1 990; Oxfo rd 1 990). H oweve r there were a n u m b e r of difficulties i n re lati on to the
classification of strategies which needed to be resolved , and t hese are detai led
be low.
One source of confusio n e ncou ntered in the early stages of the p resent study
was the categorisation of advance p reparation as a m etacognitive strategy
(O'Malley et al. 1 985a) and of rehearsal as a cog nitive strategy (Chamot et al.
1 987). The confusi o n arose because the operati o nal d efi n iti o n s for these two
strategies appear to be almost identical , with both referri ng to the re hearsi n g of
lingui stic compone nts fo r an
upco ming
task.
lt
reclassification took place afte r 1 985 and that the gen e ral co nse nsus was that
re hearsi n g language for an u pco m i n g task i s a cog nitive strategy . I n this study
56
re hearsal i s co nsidered to be a cog n itive strategy i n that it i nvo lves e ngaging with
the TL to p repare for a particu lar task.
Once classification problems had been reso lved , Metacog nitive Strategy Use
(MSU) was operatio nalised according to the ten ite ms defi ned i n Table 3.2.
57
Table 3 . 2
Metacognitive Strategy Descriptions
M etacognitive Strategy
Description
Advance Organisation
Selective Attentio n
D i rected Attentio n
Delayed P roduction
Self-manageme nt
P roblem Identification
Self-monitoring
Self-evaluation
P rioritising
Revision
Cog nitive strategy use (CSU) was ope rationalised acco rdi n g to t h e seve nteen
items defi ned i n Table 3.3.
58
Table 3 . 3
Cognitive Strategy D escriptions
Cognitive Strategy
Description
Repetition
Resou rcing
Grouping
Note taking
Deduction
Substitution
I magery E laboration
Visual isation
World Elaboration
Relating parts of the task to each other. O'M alley and Chamot
( 1 990:1 38)
Contextualisation
Summarisation
Translation- To English
59
l nferencing
Transfer
Rehearsal
Social strategy u se (SS U ) was measu red by items based o n the defi nitions i n
Table 3 . 4 .
Table 3 .4
Social Strategy Descriptions
Social Strategy
Definition
Questioning
Co-operatio n
60
Affective strategy use (AS U) was m easu red by t h re e ite ms d rawn fro m the
categories defi ned i n Table 3 . 5.
Table 3 .5
Affective Strategy Descriptions
Definition
Affective Strategy
Self-talk
Self-reinforcement
Self-encourag ement
The ope ratio nal defi niti ons fo r strategy use p rovided above were used to d evelop
a questio n n ai re to p rovide o rdered categorical data on the frequency of strategy
use by u nderg raduate foreign language learners. Detai ls co ncerning the
developme nt of the strategy use questio n nai re are prese nted i n sectio n 3 . 5 . 2 .
T h e re mai nde r of sectio n 3 . 3 . 1 co nce rns t h e ways o f defi ning a n d m easu ri ng the
learning context and learn e r characte ristic variables in the study.
The variables re lating to the co ntext of learn i n g were m easu red as fo llows. Firstly ,
61
there were two cou rses at the 1 00-level : o n e fo r ze ro begi n ners o r learners with
so m e elementary knowledge of the TL (1 OOE-Ievel ) , the other for learn e rs who
have h ad more e xtensi ve experi ence of learn i n g t h e TL usually at school or i n the
host country. Language use opportunities were m easu red in t e rm s of whether
subjects had oppo rtunities to use the languag e above and beyo nd those provided
by the cou rse.
The variables relati ng to the co ntext of learn i n g were all categorical variables.
Six variables relating to learner characte ristics we re used i n the study. There
were six categ o ri es of respo nse as a measu re fo r age (< 21 , 2 1 -30 , 3 1 -40, 41 -50 ,
5 1 -60 , >60), and subjects we re asked to mark the age g roup to which they
belonged. Su bjects indicated their gender by marki n g eith e r t h e cate g o ry of male
or fe male . The i ndex fo r language learning experience was whether su bjects had
learnt lang uages other than their mother tongu e(s) and the languag e they we re
cu rre ntly studyi ng. Prior experience in learning the TL was m easu red in terms of
whether subjects had learnt the TL before e n ro l l i n g at u n ive rsity. Su bjects we re
also asked to indicate where they h ad h ad this p rior experi e n ce (schoo l , host
cou ntry , other).
62
i mportance the i ndividual attaches to beco ming proficie nt i n the TL and to
co nti n ui ng the study o f the TL. N i ne q u e stions we re generated to tap these
diffe rent facets of motivation and su bjects were requi red to respond to them o n
a Li kert scale o f i mpo rtance (1 -5) .
The measu re for proficiency was the g rade for the course received by the l earner
com prising both internal assessm e nt (based on course wo rk) and e xternal
assessment (based on final exam i n ations) of language skills. For the pu rposes
of this study the g rades 'A' , 'B', and 'C' were used for learn e rs who passed the
p rog ramme in descendi ng deg rees of p roficie ncy and 'D' was u sed for learners
who did not pass.
For the variables re lating to learner characte ristics all the variables were
categorical. Fo r the motivation variable the responses p rovided ordered
categorical d ata about motivatio nal i ntensity.
P reli m i nary i n vestigation of strategy use i n this study co ncerns the freque ncy with
which strateg ies are deployed by u nd e rg raduate fore i g n language learners.
Answe rs to this qu estion p rovide base line data through which i n s ights may be
gai ned i nto the strategy use of the sample as a who le.
I n section 3 . 3. 1 it was hypot hesised that di stance learn e rs wou ld make g reater
use of metacog nitive strategies and l ess use of soci al strateg i e s than classroom
learners. The pri nci pal research question for this study re lates t o the infl ue nce of
mode of study on metacog nitive , cog nitive , soci al and affective strategy use. I n
additio n , the role that the T L may play i n the strategy choice o f languag e learners
i s exami ned. Research i nto these two aspects of the problem i s carri ed out using
conve rg e nt data obtai n e d t h ro u g h question naire and ve rbal repo rt procedures.
The rol e of oth e r variables in strategy choice is also exami n ed . These variables
include those which h ave a l ready bee n i nvestigated in the research literatu re
63
(namely, level of study, g e nder, profici ency and motivation) as wel l as a further
set of variables, the effect of which on strategy use has not yet been explored
(e.g . , prior TL e xperie nce and lang u ag e u se opportu nities) . The re lati onship
between these vari ables and learners' Istrategy choice i s i nvestigated and is also
u sed to assess the relative i m pact of m ode of study and the TL o n strategy
choice.
1.
2.
What is the i nflue nce o f mode of study o n metacog nitive, cog nitive , social
and affective strategy use by u nderg raduate foreig n lang u ag e learners?
3.
4.
What role does the TL play i n the strategies learn e rs choose to use
( m etacognitive, cog nitive , socia l , affective) to improve their TL com pete nce?
5.
Do aspects of the lang uag e learn i ng co ntext (namely level of study and
lang u ag e u se opportu nities) affect the metacog nitive and cog nitive
strategies e mployed by u nd e rgraduate foreign lan g u ag e learners?
6.
64
In this study the questio n n ai re procedu re is e mployed to investigate the i nflue nce
of a rang e of variables re lati n g to learner characteri stics and the l earning context
o n strategy choice. The questionnai re data i s particu larly suitable for the
m u ltivariate analysis of the variables i n the cu rre nt study (see section 3.5. 1 and
section 4 . 1 ). The verbal report study has a m o re limited focus and i s used to
e xami ne in more detai l the re lationship betwee n learn i ng strategy use and both
m ode of study and the TL.
In previous studies of strategy use by lang uage learners various data col lecti on
tech niques have been used : questionnai res ( Ram i rez 1 986 ; Oxfo rd et al . 1 988) ,
i nte rvi ews (We nden 1 986a; Pearson 1 988) , observation of classroom behaviou r
( Nai man et al. 1 978 ; Cohen and Aphek 1 98 1 ; Rubi n 1 98 1 ) , learner reports i n
d iary form (Tyacke and Mendelsohn 1 986) , and ve rbal p rotoco ls (Co h e n and
Aphek 1 980, 1 981 ; Cohen and Cavalcanti 1 987, 1 990 ; Mangubhai 1 99 1 ). I n
selecting appropriate data co llection tech niques fo r the curre nt study i t was
i m po rtant to conside r the experiences of earlier researchers regard i ng the vi abi lity
and usefu ln ess of the p rocedures they had used to el icit strategy use data.
One of the earliest studies of strategy use by Naiman et al. ( 1 978) revealed that
it was vi rtually i m possible to obtai n accu rate i n sig hts about learn e rs' co nsci ous
thoug ht p rocesses through co nve nti o n al observations of teacher-ce ntred
classroom sessions. After a n u mber of h o u rs of classroom o bse rvation Naiman
et al. co ncluded that very few learning techniques were overtly displayed i n the
classroom . They fe lt t hat only through intervi ews cou ld one h ave access to
tech n iques that were i nvisible to any obse rvers - such as 'atte m pti ng to answer
65
to themse lves eve ry questio n asked by the teacher' Naiman et al. (1 978 :68). The
fi ndi n g s of subseq u e nt studies by Rubin (1 981 ) and Cohen and Aphek ( 1 981 )
were consiste nt with those of N ai man et al . , i n that observations of lang uage
classrooms we re not productive i n reveali ng strategies u sed by lang uage
learn e rs.
Various i nvestigators (e.g . Rubin 1 98 1 ) h ave attem pted to use di ary studies
i nvo lvi ng retrospective accounts of l earn i ng experiences. There has bee n some
success with these, particu larly after people have received trai n i ng in the
tech niques of se lf-observation and recordi ng , b ut such studies requi re a deg ree
of sophistication and application on the part of the learner th at can not be
g e n e rally assumed.
The major difficu lty fo r the i nvestigato r when attempti ng to collect data on
strategy use lies i n the fact that subjects fi nd it very difficu lt to arti culate the
strategies they e mploy, particu larly if they are asked what they do out of context.
Research in oth e r fi elds has shown that it is hard for expert learners , as opposed
to novices, to m ake i ntu itive knowledge exp licit. Brown and Burton ( 1 978) , in their
study of eleme ntary school mathematics teachers acquiri n g arith m etic ski lls,
fou nd that the teachers becam e expert in performi ng arit h m etic operations, but
had difficu lty in expressi ng what these operations were.
Anothe r co mplication for the researcher is that different types of data col lection
m ay l ead to d iffe rent co nclusions about the character and use of languag e
learning strategies. O'Mal l ey and Ch amot (1 990 :95-96) make the fo llowi ng
obse rvations on data collection proced u res:
I n o u r earlier studies we were atte mpti n g to obtai n a broad su rvey of
the types of strateg ies learners used and we collected d ata with both
small g ro u p i nterviews and questi o n n aires. One of t h e co nclusions we
reached from these studies is that the strategies reported depend on
the data collection m ethodology. This seems disco ncerting o n ly if one
ignores t h e e xtreme diffe rences in the way that questions are asked
of respo nde nts usi ng these differe nt methodolog ies. W h e n t hese
differe nces are acknowledged, finding varyi ng resu lts fro m diffe re nt
data collecti o n procedu res can be expected.
66
Such a co nclusion i s not surprising whe n one co nside rs the differe nt demands
placed on su bjects who are provided with instances of strateg ies (as i n a
recog niti o n task) com pared to learn e rs who m ust come fo rt h with their own
strategies (as i n a productio n task) . The general tendency for production tasks
to u ndere sti m ate com pete nce and for recog nition or co m p re h e n sion tasks to
ove resti m ate compete nce is well known in the literatu re (Frase r, Bellugi and
Brown 1 96 3 ) .
Two i n stru m e nts were used to gather data for this study: ( 1 ) a questionnaire
relati ng to t h e strategy u se and biographical characte ristics of subjects ; (2) a
ve rbal repo rt procedure known as the 'yoked subject tech n ique' to e licit
retrospective ve rbal accou nts of lang u age learning strategy use.
Descri pti o n s of each of these i nstru m e nts, the proced u res u sed to trial and
ad mi niste r t he m , and the methods of p rocessi ng the data are d etailed in section
3.5 (qu e sti o n nai re p rocedure) and section 3.6 (ve rbal report procedu re ) .
67
3.5 THE QU E STIONNAIRE
Fi rstly, one advantage of qu estio n n aires is that no prior trai n i ng of the subjects
is requi red i n the use of the procedure as co m pared to diary studies and
i ntrospective accounts. Seco ndly, t h e respo nses of learners who complete the
questionnai re can be considered to be re lative ly fre e of i nvestigator i nfluence ,
co m pared to i nterviews. I n te rms of feasi bi lity a questi o n n ai re is easy to
ad m i niste r. This i s an i m portant co nside rati on given the larg e sample and the
particu lar context of this study: distance learn e rs are spread t h roug hout N ew
Zealand (and a small n u mber are overseas) and at t h e fi rst year level o n ly some
attend the volu ntary o n-cam pus co urses. Fi nally, a structu red questio n nai re such
as the one deve loped i n this study yields quantifi able i nfo rmation on the rang e ,
type and frequency of strategy use of subjects. Respo nses are lim ited to
informatio n that is re levant , and this i nformatio n can be re adily coded and
analysed usi ng com pute r prog ram mes.
There are a n u m ber of pote ntial problems with question nai res and th ese will be
considered i n the context of this study. Fi rstly t h e re is often a low retu rn rate with
\
m ai led qu estio n nai res which raises questions about the reaso n s why certai n
s ubjects respo nd and others do not. However i n t h i s study t h e response rate was
ve ry hig h (see sections 3.5.3 and 3.5.5).
A noth e r potential p roblem is that the questions posed m ay not have been
p roperly u nderstood. lt i s also possible that not all q uestio ns are i nterpreted i n the
same way by all subjects. In an attempt to ove rco me these pote ntial difficulties
the questio n nai re ite ms we re tri al l ed i n a pi lot study which is d etai led later i n this
sectio n . Pi lot subjects were asked to m ake co m me nts n ext to any questions that
appeared to be ambig uous or u nclear.
68
The pro babi lity of m isunde rstanding i s of cou rse g reatest when learners are
asked to complete a questionnai re i n a fore i g n languag e . In this study 1 8 learners
whose m other tong u e was not E n g lish we re re moved from the sample. I n the
re mai n i ng sample all subjects h ad E ng lish as their mother tongue and so
co m pre h e nsion difficulties were n ot co nsidere d to be a major problem.
The mai n pote ntial weakness of question n ai re data is that o n e can n ot be sure if
partici pants are respondi ng i n te rm s of what t h ey thi nk they should do rather than
i n terms of what t h ey actually do. Section 3 . 6 . 1 deals with attempts to mini mise
the 'social desi rabi lity' influence o n subjects.
A se lf-com pleti o n p i lot questio n n ai re (see Appendix A) was d eve loped to gather
i nformation on learner strategies (Parts 1 -3) and re levant backg rou nd i nformation
re lated to the variables i n this study (Part 4).
Parts 1 -3 of the q uestion naire co ntai ned scaled questions about the frequency
with which learn e rs employed particular strategies : cog nitive strategies (Part 1 )
m etacog nitive strategi e s (Part 2), and socio-affective strateg i es (Part 3). This
tri partite division e nabled the research e r to provide an appropriate i ntroduction
to questi ons d i rected at the particular strategy g rouping u nder focus. For
example, in Part 1 learners we re asked to t h i nk about the t h i n g s t hey do (or do
not do) when t h e y are actually e ngaged i n wo rki ng with the language (cog nitive
strategies). I n Part 2 the co ntext fo r strategy use was the period when learners
are plan n i n g , m o nito ri ng and evaluati n g their learning (metacog nitive strategy
use). Instructio n s relati ng to Part 3 alerted subjects to the fact t hat a slig htly
differe nt response scale was to be used for some of the q uesti ons (the social
strategy use q uestions) allowing t h e m to i ndicate if they did not have the
opportu nity to u se a particul ar strategy. The tri partite division u sed fo r the
strategy ite m s was also i nte nded as a means to break u p what wou ld otherwise
appear as a l o n g list of ite ms i n successio n (32 ite ms).
69
Q u e stions re lati n g to strategy use were g e n e rated fro m defi niti o n s of i ndividual
strategies identified i n the lite rature (see section 3 . 3 . 1 ). G reat care was take n n ot
to i nt roduce i nto the questio n n ai re any tech nical terms such as metacognition o r
inferencing.
Part 1 of the questionnaire co ntai ned 1 7 scaled qu estions about the frequ ency
with which learners employed particu lar cog nitive strategies. A n example of an
ite m fro m P art 1 e nqui ri n g about i nferenci n g is given below :
1 5.
P art
which certai n metacog nitive strategies we re used. The followi n g ite m concerned
with monitoring appeared in Part
8.
Ite m
2:
While you are doing a language task do you monitor your use o f
language, correcting, i f necessary, your pronunciation, grammar,
style, etc. ?
2 b e low appeared i n Part 3 as one of two qu esti ons re lati ng to the use of
How often do you work together with your fellow learners to solve
a problem, practise conversations, check over a task. . . ?
I n Part 4 of the pi lot qu estion naire biog raphical info rmatio n was sought o n each
subject. This i nformatio n i ncluded name , age , g e nder, mother to ngue, previ o u s
e xperi e nce i n learni ng t h e T L (i ncludi ng the ci rcumstances a n d length of such
experi e nce), languag e use oppo rtunities beyond those provided by the co u rse
(i nclud i ng where and with who m ) , and expe ri e nce i n learn i n g othe r lang uages
70
(apart f ro m moth e r tongue and the c u rrent TL). This i nfo rmati o n was soug ht i n
such a way that t h e respo nses cou ld be easi ly coded. A n i n e ite m scale was also
deve loped to measure motivation (see section 3.3. 1 ) an exa m p le fro m which is
g iven b e low:
9.
3.5.3 P i l o t Study
Once t h e question naire was deve loped it was pi loted i n ord e r to g ai n information
o n practical aspects of ad mi nisteri n g the i nstru me nt, such as testi ng the clarity
of the i n structions and qu estions and dete rmi ni n g the ti m e requi red to com plete
t h e fou r parts of the questionnaire . Two further ai ms of t h e pilot study we re to
gauge the readiness of l earners to respo nd to the questi o n n ai re and to test the
i nternal consistency reliabi lity of the instru ment.
71
Table 3 . 6
Distribution o f Subjects i n Questi o n nai re Pilot Study
Target Language and Mode of Study as G roups
Classro o m Learn e rs
Distance Learne rs
F re nch
G e rman
J apanese
22
While you are doing a language task do you monitor your use of
language
...
was replaced by
While you are doing a language task do you check on your use of
language
...
Anot h e r change was required to make the q uestionnai re item co ncerning the
learn i n g of word forms more app ropriate fo r learn e rs of J apanese and Chi nese.
lt does not m ake se nse to ask about the spelli ng of wo rds when refe rri ng to the
form of J apanese and Chi nese wo rds (cou rse materials avoid Pin Yi n or
72
ro manised spelling). So when learne rs of F re nch and German were asked the
followi ng questi o n :
How o ften do you make a mental picture o f the spelling of a word you
want to remember, or of the object itself?
l earners of J apanese were asked :
How o ften do you make a mental picture of the form of a word you
want to remember (e. g. of the kanji), or of the object itself ?
and learners of Chi nese we re asked :
How o ften do you make a mental picture of the form of a word you
want to remember (i. e. of the character), or of the object itself ?
The frequ e n cy rati ng scale i ncluded i n the pilot study is give n below:
usually
5
o ften
4
sometimes 3
rarely
2
1
never
This scale was fou nd to be co nfusi n g by th ree learners who co m m e nted
i ndepende ntly that for them often was more frequ e nt than usually. To ove rcome
this co m p lication the followi ng rati ng scale from Huan g and van Naerssen ( 1 987)
was adopted and proved to present no difficu lties for respo nde nts :
very often 5
4
often
sometimes 3
2
rarely
never
1
I nformation collected t h rough the questi on naire was coded and analyzed and the
internal consiste ncy re liability of the scales was e xamined usi ng S P SSX. This
step was i m po rtant si nce a major c riticism of the questio n nai re used in the
Politzer and McG roarty (1 985) study was that the scales used had low
homog e neity and as such were not reliable (Ske han 1 989). Using Cronbach's
Alpha test the i nternal co nsiste ncy re liabi lity of the strategy scales was as follows :
metacog nitive strategy u se scale r=.73, cog nitive strategy use scale r=.78, socio
affective strategy u s e scale r=.48.
73
The low i nte rnal co nsiste n cy of the socio-affective scale was to be expected si nce
social strategies relate to quite a diffe re nt constru ct fro m the co n struct u nderlyi ng
affective strategy use. The socio-affective scale was divided i nto two scales in the
main study : one dealing with affective matte rs such as self-re i nforce ment and the
oth e r deal i n g with social aspects such as co-operative learni n g . The i nte rnal
co nsiste ncy re liabi lity of the revised strategy use scales is prese nted in section
3 . 7.3.
A sample of the revised question naire u sed i n the main stu dy fo r learners of
Chinese appears i n Appe ndi x B.
Beari n g these poi nts i n m i nd, it was decided to collect the data ove r a fo u r month
peri od fro m May to Aug ust 1 99 1 , thus e ncompassi ng the two o n-campus cou rse
peri ods for distance learners.
74
Instructions
The fi rst pag e of the questi o n nai re g ave su bjects i nformation about the survey.
They were i nfo rmed that the questio n nai re was part of a comparative study of
how classroom and distance languag e learne rs m an ag e the p rocess of languag e
learning. S u bjects we re asked t o respond i n terms of what they actually d o i n the
context of private study , that is how t hey m anag e their l ang uag e learn i ng beyond
any classroom contact they may h ave . Care was take n to stress that individuals
learn in diffe re nt ways and that there i s no set of learning b e h aviou rs which can
be consi dered rig ht for everyon e. This latter poi nt was an atte mpt to reduce the
'social d esi rabi lity' i nflu e nce on respo nde nts. lt was also e m p hasised that
participation was enti rely volu ntary , that all resu lts wou ld be kept co nfidential and
that the qu estions in no way formed part of the assessment of the cou rse. The
q uestionnai re was exactly the same fo r classro o m and distance learn e rs .
A ti metable of class visits was d rawn u p i n consu ltation with the lectu re rs and
tutors i nvolved in each cou rse. Sixtee n classes we re visited ove r a ten week
period fro m May to J u ly 1 99 1 . Subjects we re asked to co mplete the questi on nai re
at the start of each class to avoid any possible i nflue nce fro m earlier class roo m
activitie s .
75
aske d , d u ri n g t h e i r o n-campus co urses i n August , to give accou nts of their
lang u ag e learning strategies usi ng the yoked subject tech nique. Attached to the
mailed q uesti o n n ai res was a letter providing backg rou nd i nformation about the
question n ai re (see Appendix C). This i nfo rmati on was g iven to the i n-class
subjects (both i n term time and at o n-campus courses) by t h e researche r as an
i nfo rmal introductio n . In all 249 mai led questio n n ai res we re despatched to
distance learn ers and 1 58 were returned. This re prese nts a h i g h respo nse once
the attrition rate i s co nsidered. O n e month after the mai ling of the questionnai res
the n u mber of withdrawn stude nts was checked and it was found that 37 of the
orig i nal 249 learn e rs had officially withdrawn from their lan g u ag e cou rses by the
e nd of May. This set the respo nse rate for the postal questi o n naires at 74.5%.
Once received the questi o n n ai res were coded acco rdi ng to the course nu mber
of the stude nt co ncerned and eac h pag e was ch ecked. If parts of a questionnai re
we re i nco mp lete it was not i ncluded i n the sample. Any co mments were
transcribed and records were kept as to which particular items the co mments
refe rred to (if this was the case) . I nfo rmation o n each subject (mode of study, TL,
leve l o f study) was i n serted i nto the qu estion naire. The qu estio n nai re data was
t h e n e nte red i nto an ASC I I fi l e for future analysis usi ng SAS.
The methods for analysi ng the questionnaire data are detai led i n Chapte r 4
(sectio n 4 . 1 ).
Verbal report p rocedures, also known a s verbal protoco ls, have played a ro le in
a sig n ificant n u mber of the studie s which have atte m pted to identify learner
strategies. Cohen ( 1 984) classifi es ve rbal report m easu res i nto three types. Fi rstly
there are se lf-report measu res which co llect learners' g e n e ralised stateme nts
about their learn i ng behaviou r o r characte ristics. Secondly se lf-observation
measures i nvolve the inspectio n of specific, not g e n e ralise d , language be haviour,
e ithe r i ntrospectively o r retro spectively. Thirdly think-aloud techniques, also
76
known as self- revelation, consist of the stream-of-co nsciousness disclosu re of
t h o u g ht p rocesses du ring the e xecuti o n of a task.
The se lf-obse rvatio n technique used fo r the cu rre nt study is d escribed later in this
section o nce some of the i ssues of controve rsy reg arding ve rbal report data have
been considered.
3.6. "1 Li mitat i o n s of Verbal Reports
Conce rns about the use of data based on ve rbal reports have bee n expressed
in t h e psych olog ical lit.eratu re (e . g . , E ricsson and Simon 1 980 , 1 984, 1 987;
Seli g e r 1 983 ; Dob ri n 1 986; E ricsson 1 988 ; Cohen 1 983, 1 99 1 ). A s u m m ary and
discussion of t hese concerns are prese nted below.
Another dime nsion of the potential p roblems of verbal re port d ata conce rns
i nflu e nces on the content of what is re ported. For exam p l e , i n st ructions, p robes
or p ro m pts m ay act as cues wh ich shape the things subjects report o n , or how
t h ey repo rt t h e m . There is also the possible effect of the task of ve rbali satio n
itself: t h e need fo r additio nal ve rbal p rocessi ng may i nte rfe re with the p rocessi ng
t h at i s being com mented o n .
A fu rther concern is t h e possibility that resu lts obtained t h roug h ve rbal report data
wi l l vary according to the characte ristics of participati ng subjects . Considerable
77
individual diffe re nces i n tendency to ve rbalise exist (Miyacke and N o rman 1 979)
and respo nde nts may differ with respect to their verb al ski l ls , such as
articulate n ess and specificity. This prese nts us with an eve n g reater co mplicati o n ,
as Skehan ( 1 989 :80) sugg ests, namely t h e possibility that :
what accou nts for t h e re porting o f st rategies and the lang uag e learning
success are the same thing - g reate r powe rs of articulate ness. lt is
possib l e , i n other wo rds, that so m e people are capabl e of more
preci s e , d etai led and organi sed thought perhaps because of
decontextualization abi lity, analytic capacities with ve rbal m aterial, or
memory , o r othe r factors . This is what e nables them to reflect o n their
own lang uage learni ng experi e nces effective ly, and re port t h e m so
we l l .
Skehan contends that ve rbal repo rts i n s u c h cases may not e nable u s t o ide ntify
whether the strategies the mse lves, o r the p owe rs of articulate n ess, contribute to
language learn ing success. A further perspective is offe red by Garner (1 988a)
when she poi nts to situations whe re i n sufficient data are obtai ned and co ncludes
that in such circu mstances it is not possib l e to know whether this is the result of
limited cog n itio n , limited languag e ski l l , or some co mbinatio n of these factors.
Cohen (1 9 9 1 : 1 37) writi ng about the co ntrove rsy regard i n g ve rbal report data
notes :
The critics wou ld suggest that these n u m e rous problems with ve rbal
repo rt m e asures seriously li mit the g e n e rali sability of the findings and
might eve n preclude their use. H owever, proponents of verbal report
wou ld arg ue that cog nizance of t hese problems i n plan n i ng the
research design may help to avoid some of them and that others wi l l
sim ply p revai l , just a s proble m s are i n he re nt i n the use of other
research measures as we l l .
b.
c.
78
The ve rbal report procedu re used i n this study has bee n termed the yoked
subject technique. lt fol lows Garner's g uideli nes i n aski ng learn e rs to reveal what
they do i n t h e context of a section of t h e i r study mate rials and i s also used as
another m ethod of data collecti on to co m p lement the findi ngs. Th e i m portance of
such guide l i n e s is e m phasised by C o h e n ( 1 991 : 1 37-8) :
Whereas the re liabi lity of m e ntalistic measures has been questio ned
in co m parison with behaviouri stic m easu res, research has
demonstrated that ve rbal repo rts, e licited with care and i nte rpreted
with f u l l u nderstanding of the ci rc u m stances u nder which they we re
o btai ned, are , i n fact, a valuable and thoroug h ly re liable source of
i nformation about cog nitive processes (Ericsson and Si m o n 1 980).
Verbal report tech niques are used in the prese nt study si nce they can provide
fi ne-g rai ned i n formation about learn e r p rocesses , i nformation that is otherwise
lost to the i nvestigator (Ericsso n and Si m o n 1 9 84; Ericsso n 1 988). The particular
ve rbal report technique used i n the cu rre nt study is i ntroduced below.
The yoked subject tech nique was fi rst used i n an i nve stigation carried out by
Nayak et al. ( 1 990) i nto whether m u lti l i ngual su bjects wou ld perfo rm better than
monol i n g u al subjects in learning a m i n i ature linguistic syste m . T h ey i ntroduce this
proced u re fo r o btai ning verbal reports as fol lows :
S u bjects were asked to make t h e i r strategies as explicit as possi ble fo r
anot h e r (yoked) subject who is to perform the task. We h o ped that the
yoked subject procedure wou ld yield cleare r informati on about the
strateg ies differe nt g roups of subj ects used under diffe re nt conditions.
(Nayak et al. 1 990 :226)
79
The i n stru m e nt is located towards the low e nd of structuredness in that the object
of the ve rbalization is li mited to a section of the curre nt learn i n g m aterials and the
specific fo rm and co ntent of the re port is at the i nfo rmant's d iscreti o n . As such
it also allows learners to reflect on the aspects of their strategy u se which they
co nsid e r to be significant.
A descri ptio n of the pi lot study used to trial this tech nique is given below.
A pi lot study usi ng the yoked su bject tech nique was undert ake n with four 300leve l lang u ag e learners who had al so partici pated i n the qu esti o n nai re pi lot study.
The ai m of the trial was to test the viabi lity of both the warm - u p procedure and
the actual procedu re , the adequ acy of the instructio ns give n and the time take n .
The proced u re took approximately t h i rty m i n utes a n d su bj ects reported that the
instruct i o n s were clear and that o nce they began talking the p rocedu re fe lt quite
natural si nce they were reporting o n their i nteractions with t h e study materials i n
fro nt of t h e m .
The m ai n yoked subject procedu re s took place over a period of three weeks i n
Aug u st. There were fou r sessio ns, two for distance learners and two for
classro o m learners.
Instructions
As an i ntroduction the researcher explai ned that the purpose of the session was
to find out how stude nts go about their lang uage learn i ng s o that th ese reports
cou ld b e i ncluded i n a strategy u s e g uide . lt was mentioned that this g uide wou ld
80
be dist ri buted to al l classroom and distance lang u ag e learne rs that year. The
i mportance of i ncluding exce rpts fro m learners' own accou nts was e m phasised
as well as the fact that al l reports would be kept confide ntial and wou ld not be
shown to staff i nvolved with the i r lang u ag e cou rses. The fact that partici pation
was volu ntary was also e m phasi sed.
Fo llowi ng the sugg estion of Ericsso n and S i m o n ( 1 980) and Rubin ( 1 981 ), a
warm-up phase was included befo re l earne rs were asked to report o n their
strategy use. The distance partici pants we re p rovided with a sam ple cove r sheet
and retu rn add ress card which they wo u ld n o rmally su bmit with each assig nment.
They were then asked to talk about what they did with this material. This provided
an authentic task re levant to the procedu res fo llowed by distance learners with
which to p ractise produci ng self-observatio n data. The task howeve r did not
i nte rfe re with the mai n task re lati ng to the actual learning materials.
The warm-up task for classroom learners requ i red them to talk about thei r weekly
schedule of l ang uage classes and assessment p rocedu res. lt was e mphasised
that this was a p ractice task to fam i l iarise them with the p rocess of reporting to
another stu d e nt. Agai n , describing class schedules and assessment is an
authentic task which is related to, but does not i mpi nge on the main task of
reporti ng strategy use.
Subjects were then provided with a copy of the study guide (in the case of
learners of Fre nch) o r workbook (fo r learners of J apanese) fo r the lang uag e u nit
they had al most co mpleted. They were asked to talk about how t hey go about
studyi ng eithe r Fre nch or J apanese as if they were actu ally talki n g to a fellow
classroom o r distance learn e r (the yoked subject) who was plan n i ng to e n rol in
the same cou rse the fo llowi ng year. They were asked to refe r to a particular
section of the mate rial in front of t h e m , in o rder to make the i r reports more
specific. The w ritten instructions g iven to subjects (see Appe ndix D) provide both
an approp riate co ntext fo r talki ng about strategy use and also a n u mber of
81
p rompts which ai m ed t o serve the same fu nctio n a s t h e kinds o f questions that
would be asked by the hypothetical yoked subject.
Procedures
Subjects reco rded their reports i n language labo ratory booths and were able to
replay , de lete or add co m m e nts
as
they wished.
The advantag es of reco rding reports i n the booths are many. Fi rstly there is less
likelihood of subj ects being influe nced by the researcher because the researcher
can not be seen o nce the p rocedure i s u nderway. Secondly, si nce subjects are
asked to i m agi ne they are talki ng to another language learne r, the more p rivate
co ntext of the booth i s less distracting and therefo re more co nducive to focusi ng
on the task. Thi rdly, it would not have been possible to gai n access to a larg e
number o f distance l earners i n a particular cou rse on an individual basis si nce
they are scattered t h roug hout N ew Zealand and the o n-campus cou rse sch edules
are e xtrem e ly tight. Lang uage learne rs are very kee n to have as m uch exposure
as possible to the TL and it would have bee n u n reaso nable to ask them if they
wou ld mind withd rawi ng from sessions o n an i ndivid ual basis. Final l y , the fact that
respo nses were recorded meant that there was a permanent reco rd of the data.
Thus the pote ntial for i n accu racy rai sed by O'Malley and Chamot ( 1 990), when
an i nte rviewe r atte mpts to code strategy occu rre nces i m mediately after the
i nfo rmant m e ntions t h e m , was avoided.
Th rough the yoked subject tech niqu e , 1 1 reports were collected fro m classroom
learn e rs . lt was o n ly possible to use n i n e of these , si nce two were g iven by non
E ng lish n ative speake rs who had so me difficu lty in repo rti ng on their strategy use.
Twenty- n i n e ve rbal re ports were o btai n ed fro m distance learne rs , and of these
o n ly one could not be used due to a fau lty reco rding.
82
Once the tapes were col lected , each tape was n u m b e red , with the TL and mode
of study rep resented by the appro p ri ate i n itials ( F ,J ) .Thus t h e t h i rd ve rbal report
for a classro o m l earner of French was ide ntified as 3FC.
Chapter 5 co ntai ns a description of the methods used to analyse the ve rbal report
data (section 5. 1 ) .
Any d ata collection procedu re , by t h e ve ry fact of its use, creates some effect on
the d ata. In o rder to assu re the qual ity of data co l lection p roce d u res, the criteria
of re liab i lity and validity were app lied to this study.
83
Subject Selection
Four learne rs e n ro l led in J apanese language classes were in fact native speake rs
of Japanese. They did not form part of the sample for the questi o n nai re study or
the verbal re port study.
Subjects co mpleting a questionnai re in a lang uag e other than thei r mothe r tongue
may h ave experie nced difficu lti es in u nderstanding some ite m s or may not have
u nderstood them accurately, thus i nfluenci ng the validity of the data. Subjects
were asked to give thei r mothe r tong u e thus al lowi ng the researcher to identify
learne rs who we re not native spe ake rs of E ng lish. Twe nty-fou r su bjects noted
that they had two mother tongues, o n e of which was English , and these people
were retai n ed in the qu estionnai re sample . Eig hteen subjects g ave a language
other than E ng lish as their mother to ng ue and these stude nts we re removed from
the question nai re sample since it was not possi ble to check o n their leve l of
p rofici e ncy in English. Two classro o m learn e rs were re moved from the ve rbal
report sample si nce they we re not native speake rs of E n g li s h .
Instrumen tation
lt was also i m portant that the observatio ns used i n this study were valid and
consist e nt and that the definitions of t h e key terms reflected t h e constructs u nde r
study. To d efi ne the co nstructs of m et acog nitive strategy use , cognitive strategy
use , social strategy use and affective strategy use a rep re s e ntative set of items
84
were d rawn f ro m the literatu re (see section 3.3. 1 ) . Ope ratio nal defi nitions have
bee n g ive n fo r the learn i ng context variables and the learne r characte ristic
variables used in the research, also i n section 3 . 3. 1 .
Task Directions
I n structio n s were plan ned and pi loted to e nsure that they were clear and that
su bjects knew what was e xpected of the m i n co mpleting the qu estionnai re and
in followi ng the yoked su bject procedure. Identical procedu res were followed i n
col lecti ng d ata from diffe re nt classes. Th e o ral i ntroduction g ive n t o classroo m
learners p rese nted the i nfo rmatio n co ntai n ed i n the acco mpanyi ng letter mai led
to distance learners. In the mai n study a simi lar consistency in the col lecti o n
procedu re was fol lowed .
H atch and Lazaraton (1 9 9 1 :39) deal with a fu rthe r aspect of i nternal validity i n t h e
fo l lowi ng way :
To b e valid the data-g atheri n g procedu re should allow u s to tap the
t ru e abi lities of the learners . . . . I n qu estio nnaire research we suggest
that d e m og raphic informati o n . . . be placed at the e nd of the
quest i o n n ai re . . . . You are more like ly to g et full participation if other
data a re col lected fi rst and personal i nformation seco nd.
I n the q u e sti onnai re study, su bjects we re fi rstly asked to respo nd to qu estio n s
about thei r strategy use and t h e n t o provide backg rou nd info rmation o n
th emselves. T h e fi nal section o f the backg rou nd info rmati o n consisted o f a n i n e
item scale relating to lang uage learn i ng motivation. This requ i red su bjects to t h i n k
about whether certai n i ssues were important t o them in learning t h e TL. Since
responses to t hese questions are less automatic than those about age , g e n d e r,
p rior languag e learn ing experience and so o n , the validity of the motivation data
could p robably have been i mp roved by putting the ni ne-ite m scale at the start of
the backg round info rm atio n section rather than at the end. Subjects who did n ot
complete t h e questio n s re lating to motivation were not included i n the sample.
85
3.7.2 E xternal Val i dity
Subject Selection
The qu estionnaire study was carried out with i ntact g roups of languag e learn ers.
S u bjects were se lf-se lected i n that partici pation was e nti re ly vo lu ntary. Not all
classroom learners we re p resent whe n the qu estion n ai re was ad m i n iste red and
not all distance learners respo nded to the questio n n ai re . Thus, it was not possible
to be e ntirely su re that the subjects fo r the study fo rm a representative sample.
T h e yoked su bject procedure was also carri ed out with i ntact g roups, namely
2 00-level classroom and distance learn ers of J apanese and French. Self
s e l ection also took place . lt was noted that a n u mber of learn e rs felt re luctant to
partici pate in the yoked subject procedu re becau se t hey did not feel suffici e ntly
co nfident about thei r own 'study habits' or about their abi lity to re port on the
p rocesses they use. Th us subject selectio n may li mit the g e ne rali sabi lity of the
fi ndings of the study.
E ach method of data co llection has u nde rlyi ng theoretical assu m pti ons about the
n atu re of the data. The qu estio n nai re procedu re requires subjects to respo nd to
s pe cific ite ms while the yoked subject procedure is a less structu red i nstru ment
and requires subjects to p roduce verbal re ports o n t h e i r strategy use. These
d iffe re nces in the degree of structuredness of t he i n stru m ent and in the differi ng
relative e mphases o n productive ve rsus receptive com petence i n re po rti ng on
strategy use can be expected t o i nfluence the fi ndings o btai ned . G ive n the h i g h ly
i ndividualised nature of strategy use care was taken to obtai n co nve rgent data
o n the strategy u se of languag e learn e rs.
86
Ecological Validity
The yoked subject p rocedu re has a particu lar ecolog ical vali dity i n the co ntext of
the cu rre nt study i n that learne rs a re asked to repo rt on t h e i r strategy use i n a
realistic context - that is to report o n how t hey g o about l earni ng the TL by
looki ng at a particular section of t h e workbook materials and talki ng as if they
we re with o n e of t h e i r peers who was plan n i ng to e n rol in t h e p rog ram m e . The
procedure is not artificial in that it co ntai ns many of the eleme nts of peer tutoring
and as such is a fitting procedu re t h roug h which to exte rnalise the strategic
repe rtoi re s of classro o m and distance learn e rs.
The criteri o n of re liabi lity was applied t o the study i n o rder to d ete rmi n e the exte nt
to which the data collecti on p roced u res can be conside red accu rate and the
resu lts can be co nside red to be stab le.
87
lnterrater Reliability
3.8 LI MITATIONS
Li mitati o n s of ve rbal report data were considered in sectio n 3 . 6 . 1 , and ag ain the
possibility that learn e rs are reporting what they think they should do, rather than
what they actually do can not be altog eth e r ru led out. F u rtherm o re , su bjects
probably repo rted o n a particu lar subset of the strategies they used , that is, those
strategies which they were conscious of at the ti me of the p roced u re . Thus, the
ve rbal report data can not be seen as p roviding a compre h e nsive accou nt of the
strateg i c rep e rtoi re of each languag e l earner. I n addition , the subjects may have
varied c o nsiderably in the i r abi lity to talk about the strategies they use, and thus
the data fo r some subjects may be much less rich and relatively i ncom plete
com pared to their actual strategy use.
88
As Po litze r and McG roarty ( 1 985 : 1 1 8) note, g iven the number of suspici ons t hat
have b e e n raised about self- report data, this data 'should, w h e n ever possi ble,
i nclude a check of the se lf-reports ag ai nst actual o bse rvatio n s' . I n thi s study it
was n ot actually possible to o bse rve learn e rs wo rki ng on particu lar language
tasks , d u e to limitati ons of time and access. H oweve r, it re mains a hig hly
desi rable sou rce of co nfi rmation about the re liabi lity of se lf-re p o rt data.
I n this study , the questionnai re data and the ve rbal repo rt data we re not d rawi ng
on exactly the same sou rce si nce they were col lected at diffe rent times, and did
not require learn e rs to think abo ut strategy use in relation to exactly the same
tasks. Th us, while m u ltiple measu res were u sed, the fact that they were not
di rected at exactly t he same sources is a lim itation of this study.
One fu rthe r weakness relates to the re latively small number of classroom learners
co m pared to distance learn e rs who partici pated i n the yoked su bject p rocedure .
T h i s was u navoidable i n that the classroom learners o f French were a small
g ro u p and the classroom learn e rs of J apanese appeared diffident about
participation in the procedure . Howeve r, a larg e r re prese ntation of classroom
learners wou ld h ave been desirable .
89
3.9 SU MMARY
This chapte r has p re sented the research desi g n used to i nvestig ate the freque ncy
of strategy use and i nflue nces o n the strategy use of u nderg radu ate foreign
language learn e rs.
Resu lts fro m t h e analysis of the questio n n ai re data are presented in Chapter 4
and provide evidence relati ng to the research qu estions p rese nted i n section
3 . 3 . 2 . Resu lts fro m the analysi s of the ve rbal report data are d etai led i n Chapter
5 , and the fi ndings shed fu rthe r light o n questions co ncerning the influence of the
TL and mode of study o n the re po rted strategy use of fo reign lang uage learners.
90
This chapte r p resents resu lts from the analysis of the q uestion nai re data.
Fi ndi ngs are o rg an ised according to i nflue nces o n m etacog nitive strategy use
(sections 4.2 and 4.3) , o n cog nitive strategy use (sectio n 4 . 4) , o n social strategy
use (secti on 4.5) and on affective strategy use (section 4.6) . Fi nd i ng s co nce rni ng
the frequency of m etacognitive , cog n itive , social and affective strategy use also
form t he i nt rod ucti on to each of these sectio ns. Reg u lar su m m ari es are g iven at
t h e e nd of e ach stag e of the p rese ntation of results. The p ri nci pal findings are
s u m marised i n section 4. 7. The fo l lowi ng section i nt rodu ces the statistical
m et hods applied to the questio n n ai re data, i n particu lar, cano nical vari ate
a nalysi s.
The central issues in this study co ncern the re lationship between a range of
strategy use variables and a variety of vari ables re lated to the lang u age learning
co ntext and learner characteristics. To i nvestig ate such a re lationship an
i nt ri nsical ly m u ltivariate technique is required. The choice of a mu ltivari ate
techniqu e to apply to the questionnaire data was dete rm i ned by the larg e nu mber
of vari ables. As Brown ( 1 992 :649) states in an article e ntitled Statistics as a
Canonical co rrelatio n analysis, also known as can o n ical variate an alysi s (CVA) ,
p ovides a means 'fo r studying the re lationships a m o ng two sets of variables and
for studying the n u m be r and natu re of di mensions of co rrespondence' (Tabachnik
and Fide l l 1 989 :22 1 ). In the co ntext of the p rese nt study CVA is used to highlight
which particular set of strategies, if any, co ntribute to the diffe rentiation of
learners according to, for example, mode of study. Thus fro m among the te n
91
m etacog n itive strategies w e are able t o ide ntify two o r t h ree st rateg ies which
exert the stro n g est i nflue nce in setti ng classroo m learn e rs apart from di stance
l earners in term s of their m etacognitive strategy use (MSU). I n this way CVA
reduces the d i mensions i n a d ata set where there were o rigi nally a large nu mber
of vari ables. Tabachnick and Fide l! ( 1 989 : 1 93) sugg est that canon ical variate
analysis 'is best considered a descri ptive tech nique or a scre e ning procedure
rather than a hypothesis-testi ng p roced u re'.
The process of CVA i nvolves constructing a li near combinati o n of the vari ables
which separates the g roups (e .g . classroo m learn e rs and distance learn ers) as
well as possib l e . Coefficie nt s i n this co mbination are chosen which maxi mise the
between-g ro u p variatio n and m i n i mise the withi n-g roup vari ation . So meti m es it
i s possible to dete rmine several li n ear combinations which se parate the g roups.
In such i n stances, the fi rst cano n ical variate captures as much i nte r-g rou p
diffe re nce as possi ble and this i s the most i m po rtant variate. The second
canonical vari ate then reflects as m uch as possi ble of the g roup diffe re nces not
captured by the fi rst o n e , and is the seco nd most important. Thus cano n ical
variates are computed i n descending o rde r of mag nitude , and the fi rst few (one
o r two) are generally sufficient to accou nt for almost all of the i m portant g roup
diffe re nces. I n this study all the g roup diffe re nces could be acco u nted for by the
fi rst can o ni cal variate , e xcept in the case of the i nvestig ation of the re lationship
betwee n age and strategy use , i n which case two canon ical variates were
needed. As Manly ( 1 986 :89) poi nts out, o n e of the major attractions of canon ical
variate a nalysis is that if o n ly o ne or two of the canonical variates are needed ,
then 'a si mple g raphical re p rese ntation of the re latio nship betwee n the various
g roups is possible'. This is o btai ned by plotti ng the values of the variates for the
sample o bse rvations. The advantage of the scatte rplots is as an aid to the visual
i nte rp retation of g rou p differe nces .
I n the curre nt study o nce plots of the canonical variate sco res were obtai ned,
they were examined by the researche r and a statistician to dete rmi n e which
showed the most m arked separation of learne rs according to vari ables such as
92
age , g e nder and TL. T h e significance of t h e cano nical variate sco res f o r these
key plots was evaluated usi ng F values from a nalysis of variance (ANOVA) . The
standardised coefficie nts and correlation coefficie nts we re t h e n obtai ned to
ide ntify which strategies were responsible fo r the se paration of learne rs as
d i s played by the plots. A g raphical p resentation was made of t h e frequency of
use of the key strategies which had been identifi ed.
When exa m i n i ng the re lati onship betwee n age and MSU a post hoc com parison
of means was carried out usi ng Du ncan's m ulti ple range test to locate p recisely
where the most m a rked diffe re nces i n M S U occu rred amo n g the various ag e
g rou ps i n the populatio n . Brown (1 992 :648) poi nts to the use of such p rocedu res
when he suggests that 'mean comparison p rocedures may be followed by m o re
detai led co m parisons like Scheffe , Tu key , D u n n . . . to dete rmi n e e xactly where
any sign ificant diffe re nces may be located'.
lt was possi ble to calculate mean frequ e ncies for the metacog nitive , cog nitive and
affective strategy use measure s which are in the form of o rdered categ o rical data
and for which the ran k scales are i nterval-like . The mean freque ncies cou ld then
be re lated back to the o riginal respo n se scale (5=ve ry ofte n , 4=ofte n ,
3=so metimes, 2=rarely, 1 =n ever). A simi lar i nterpretatio n of frequency m easures
was made by H uang and van Nae rsse n ( 1 987) , Politzer (1 983) and Politze r and
McG roarty ( 1 9 85) .
Howeve r, s i n ce the social strategy use scale i s a categorical scale (5=ve ry ofte n ,
4=ofte n , 3=som eti m e s , 2=rare ly , 1 =never, O=no o pportunity) analysis o f the
i nfluence of mode of study o n SSU was confi ned to a com pariso n of the
respo nses of classroom and d i stance learn e rs on questions about their use of
questi o n i ng and co-ope ration.
93
To s u m marise , the p ri ncipal techniques u sed to analyse the questio n nai re data
were (i) canonical variate analysis (CVA) , (ii) u nivari ate analysis of variance
(ANOVA) and (iii) comparison of m eans (Du ncan's m u lti ple ran g e test) , all of
which we re perfo rmed using SAS (1 989) software .
Mean
S.D.
4.06
3.28
3 . 47
3 . 74
3.65
3 . 86
4.02
4.1 6
3.27
3. 1 5
0.97
1 .09
1 . 09
1 .1 2
1 . 26
0 . 94
0.94
0 . 99
1 .1 4
0.99
The mean responses were related back to the fo l l owi n g frequ e ncy values : 5=ve ry
ofte n , 4=ofte n , 3=sometimes, 2=rarely, 1 =never. They i ndicate that u nde rg raduate
fo reig n lang uage learners m ake frequent use of m etacog n itive strategies. The
most frequ e nt strategies relate to the thre e d i m e nsio n s of metacog nitio n : planning
(advance o rganisation) , monitoring (se lf- mon ito ri n g ) and evaluatio n (self
evaluati on). The mean respo nse fo r these t h re e strategies relates to the category
'ofte n'. Revisio n is the least used metacognitive strategy fo r lang uage learners
with the m ean sco re rep resenti ng a response close to 'so m eti mes' .
94
CVA was used in o rder to assess the degree of re lationship betwe e n vari ables
re lati ng to the learn i n g co ntext ( mode of study, TL, level of stu dy , languag e use
opportu nities) and the m easures of m etacog nitive strategy use ( M S U ) .
T h e fi rst relationship to b e exa m i n ed was that betwee n mode of study and the
MSU m easu res. CVA was pe rfo rmed on the data and a plot of the canonical
variate sco res was obtai ned ( Fig u re 4 . 1 ). The plot indicated that classroom and
distance l earners are separate d acco rdi ng to their use of metacog nitive strategies
along the X-axis , that is, the axis rep rese nti ng the first can onical vari ate.
Similar p rocedu res were used to o btain p lots to exami ne the rel ationship between
the TL and MSU m easure s , between level of study and MSU measu res and
betwee n language use opp o rtu n ities and MSU measu res. An exami nation of the
plots revealed that the TL appeare d to have some i nfluence on MS U , thoug h this
was n ot so marked as the i nfluence of m ode of study.
A s i m i lar set of six can o n i cal variate analyses was pe rformed between the
variables relati ng to lea rn e r characteristics (ag e , gender, p roficie ncy, p rior TL
experience , experie nce i n the l e arn i ng of other lang uag es and motivation) and the
MSU variables. Ag ai n , p lots of the cano nical variate scores we re obtai ned . The
six p lots were exam ined to see whethe r learners were diffe re nti ated o n the MSU
m e asu res according to t h e learner characteristic vari ables . The clearest i nflue nce
o n MSU fro m this set of analyses was the ag e of subjects (Figu re 4 . 2 ) .
CVA o f M S U Va ri a b l e s
M o d e o f Stu dy a s G ro u p s
Q)
-+'
-
()0
c
0
c
1!1
1!1
1!1
-3
-4
1!1
1!1
""0
c
0
o -1
Q)
(f)
-2
..
1!1 .
.,
1!1
1!11!1 _. 1!1 . .. 1!1
1!1
.
llb
1!1 ...1!1
!la
.1!1
1!1
1!1
l!ljl .
1!1
1!1
1!1
1!1 1!1 '=
Jl '
1!11!1 . . 1!1
. 1!1
1!1
1!1
d
1!1
...
1!1
1!1
:;
.. I!J 1!1""
_..
1!1
., . 1!1
1!1
1!1
1!1
1!1 1!1 -.m a 1!1
1!1
I!II!J
1!1
1!1 1!1 1!1
I
1!1 1!1
1!1
1!1
.
.
1!1
1!1
1!1 1!1
.
1!1
.. 1!1
1!1 I. ..
1!1
....
1!1
1!11!1
1!1
1!1
1!1
1!1
1!1
-1
-2
-3
Fi rst c a n o n i c a l va riate
Mo d e
of
Learn ing
l!l
l!l
l!l
Classroom
Fig u re 4 . 1
Distance
Q)
+'
r:
m 1!1
c:
'\.
-.p
1!1 1!1
0
c
0
0
-g
0
0
Q)
(/)
-1
-2
1!1
JP
-3
-4
-1
-2
-3
-4
1!1
1!1
1!1
c 21
.. .. .. 21-30
1-40
* *
* 4HIO
51-1
+ +
10
97
Findings fro m the set of canonical vari ate analyses i ntroduced i n this section are
s u m marised i n Table 4 . 2 . Ove ral l , an e xami natio n of the plots revealed that
learne rs were maxi mally separated i n thei r M S U accord i ng to mode of study
(Fi g u re 4. 1 ) and age (Fi g u re 4.2). P roficiency and the TL also exe rted some
i nfluence on the se paratio n of language learners o n MSU measu res (ranked thi rd
and fou rth i n terms of i nfluence ) . Th e re were lesse r effects fo r p rior TL
experi e nce. Learne rs we re not diffe re ntiated i n thei r MSU accord i ng to thei r level
of study, lang uag e use opportu n ities, g e nder, e xpe ri ence in the learn i ng of oth er
languag e s o r motivation.
Table 4.2
I nflue nces o n Metacognitive Strategy Use
(In Rank Order of I m portance)
Learning
Co ntext
Variables
Mode of Study
Targ et Lang uag e
Leve l of Study
Language Use Opportu nities
1
4
Learn e r
C haracte ristic
Variables
Age
Gender
P roficie ncy
Pri o r Target Language Expe ri ence
Ot her Lang uage Learn i ng
Motivation
2
3
5
To evaluate the significance of the fi rst canonical variate scores used to p roduce
the plot in Fig u re 4 . 1 , F values we re o btai ned t h roug h analysis of variance
(ANOVA) . The results, F=78.92 (1 ,41 1 ) , p<.0001 i ndicate that the separation of
class room and distance learn e rs o n M S U measures as displayed i n Fig u re 4. 1
is significant.
98
F values were also o btai ned through ANOVA to test the sign ificance of the fi rst
can o nical variate sco res displayed in Fig u re 4 . 2 re lating to the i nfluence of ag e
o n M S U . Agai n , sig nificant values were o btained : F=1 7 . 6 1 (5,406) , p<.000 1 .
Fu rther an alyses of the relationship betwee n mode of study and M S U measu res
(section 4.2.3) and between age and MSU measu res (secti on 4 . 2 . 4 ) we re then
carried out.
The coeffici e nts listed in Table 4.3 re late to the fi rst cano nical vari ate which
accou nts fo r all the variation rep resented i n the plot in Fig u re 4 . 1 . This variation
occurs along the hori zo ntal , fi rst cano nical variate axis.
99
Table 4.3
Standardised Coefficie nts and Co rre lati on Coeffi cients for MSU Variables
Mode of Study as G ro u ps
Strategy
Advance Org ani sation
Selective Attention
Di rected Attention
De layed P roduction
Self-man ag e ment
P roblem Ide ntification
Self- m onito ri ng
Self-evaluation
P rio rities
Revision
Standardised
Coefficie nts
Co rre lation
Coefficie nts
.50
-. 1 6
.12
.07
. 70
-.1 0
-.1 1
.21
-.31
.38
.60
.19
. 37
.22
. 72
.08
.21
. 37
-.01
.53
Inspecti on of the standardi sed coefficie nts i ndicates that a co m bi nation of t h ree
strateg ies i nfluences the separation of class roo m and distance learn e rs : self
manag e ment (.70 ) , advance o rganisatio n (.50) and revi sion
(. 38 ) .
An exami nation o f t h e corre lation coeffici e nts reveals that t h e same t h ree
strategies also , i ndepende ntly of each oth e r, s e rve to diffe rentiate between
classroom and distance learn e rs : se lf- manag e m e nt (.72 ) , advance o rgani sation
( . 6 0) and revision (.53).
1 00
Respo nse frequencies for the t h re e key m etacognitive variables which maxi m ally
separate classroom and d istance learne rs were obtai ned. These are displayed
in Fig u res 4.3, 4 . 4 and 4.5.
Freq u e n cy of Use of Self M a n agement
Cl assroom a n d Distance Lea rners
1 20
00
80
60
40
20
l%
<:5<5
0
Never
100000<:
:5<5-i
Never
Distance ----1
t---- Classroo m
Fig u re 4.3
The respo nses of classroom (N=1 43) and distance (N=274) learners acco rding
to t hei r use of se lf-man agement are p resented i n Fig . 4.3. This i llust rates the fact
that distance learn e rs 've ry often' e m p loy self- m anage ment i n their lang uag e
learning while the m ost freque nt response for c lassroom learn e rs is that they
'someti mes' use t h i s strategy.
1 01
Fig. 4 . 4 s hows the response of classroom and distance learners i n terms of their
use of advance o rg anisation. Classroom learne rs respond in a l most equal
n u m b e rs to the catego ries of 'so meti mes' 'ofte n' and 've ry often' wh i l e the m ost
frequent response fo r distance learners is 'very ofte n'.
1 40
1 20
f!
1 00
80
'0
...
.8 60
E
:::1
z
40
20
B888888&:
Never
1------
Classroom
-----1
N.ver
Fig u re 4.4
1-----
Dl9tance
----1
1 02
Fig . 4 . 5 i llustrates the response to the third strategy, revisio n , which contri butes
to a sig nificant diffe rence betwee n classroom and d i stance l earners i n te rms of
MSU . Classroom learn e rs tend to respo nd at the l owe r e n d of the frequency
scale while the distance learners respond towards t h e hig h e r e n d of the scale .
1 20
1 00
f 80
'0 eo
:I
z
:>0<
40
i8
20
0
Never
Never
r----- cm ----
t-----
Diatal'le4J
-----1
Fig u re 4.5
Thus three metacognitive strategies we re ide ntified which accou nt for the
separation of classroom and distance learners in the fi rst plot ( Fig u re 4. 1 ).
1 03
4.2.4 The I nfluence of Age o n MSU Variables
Fig u re 4 . 2 i ncluded i n section 4.2.2 presented a visual i nte rpretation of the
separation of learn e rs of diffe re nt ag e g roups accordi ng to their MS U .
CVA was performed t o ide ntify the influence o f particular metacog nitive variables
on the separation of the age g roups. Co rrelation coeffici e nts and standardi sed
coefficie nts fo r the fi rst two can o nical variates a re prese nted in Table 4.4. The
fi rst can o n i cal variate e xtracts 60% of the variatio n i n the data and the second
can o nical variate e xt racts 26% of the vari atio n . Together, the two cano n ical
vari ates accou nt fo r 86% of the vari atio n in the d ata rep rese nted i n the second
plot (Fi g u re 4.2). I n i nterpreting Fig u re 4 . 2 it is evident that the dispersal of
learne rs acco rd i n g to ag e g roups occu rs mostly along the horizo ntal axi s. While
the re i s appare ntly some variation (26%) along the vertical axis, no clear ove rall
pattern e m e rges for the dispe rsal of age g roups along the ve rtical axis (seco nd
can oni cal variate ) . This is another i ndicato r that the fi rst cano nical vari ate is the
most i m portant re p rese ntation of the variation in the data.
Table 4 . 4
Standardised Coefficie nts and Co rre lati o n Co effi cients f o r M S U Variables
Age of Learn e rs as G roups
First Canonical Variate
Strategy
Advance O rganisation
Selective Attentio n
D i rected Attentio n
D elayed P roduction
Self-management
P roble m Identification
Self-monitoring
Self-evaluation
P riorities
R evision
Standardised
Coefficients
.24
-.1 0
. 09
.27
.76
.15
. 03
.20
-. 1 8
.22
Co rrelation
Coeff icients
. 44
.20
.34
.41
.81
.31
.33
.33
-.07
.38
Correlation
Coefficients
.08
.12
.34
-. 1 7
-.25
-.04
.42
-.04
-.19
.54
1 04
The standardi sed coefficie nts fo r the fi rst can o nical variate reveal that se lf
manag e m e nt (.76) makes the pri nci pal co ntri buti o n to the separation of l earne rs
acco rd i ng to their ag e . The other strategies which co ntri bute joi ntly to the
separatio n of learne rs by age are de layed production (.27) , advance o rg a n isation
(. 24) and revi sion (.22).
When we consider the i ndividual contri bution of metacog n itive variab les by
referri ng to the co rre lation coefficie nts the most significant one is ag ai n se lf
manag e m ent (.81 ) . Advance o rganisati on (.44) , d e layed p roductio n ( . 4 1 ) and
revision (.38) co ntri bute i ndividually to the separatio n of learners by age g roups
re prese nted in Fig u re 4 . 2 .
Fo r the second cano nical vari ate , se lf-monitori n g and revi si o n a re the key
strategies and they are i mpo rtant both joi ntly and i ndividually. The strateg i es
ide ntified t h roug h the second canon ical variate can explai n a much smaller
perce nt of the variati o n t han those i n the fi rst cano n ical vari ate.
Table 4 . 5 presents the m ean sco res f o r t h e five m etacog n itive strategies which
m axi mally di sti nguish learners i n diffe re nt ag e g ro u p s : self-manag e me nt , advance
o rganisati o n , de layed p roductio n , revi sio n and se lf- monitori ng.
Table 4.5
M ean Sco res fo r Use of Metacognitive Strategies by Age G roups
Age G roups
Strategy
<21
2 1 -30
3 1 -40
41 -50
5 1 -60
>60
Self-m anag e m e nt
Adv. O rganisation
De layed P roduction
Revision
Self- m o n itori ng
3.02
3.88
3.50
2.89
3.80
3.48
3.93
3.60
3. 1 5
4.07
4.07
4.21
4.07
3.08
4.00
4. 1 6
4.32
3.83
3.42
4.31
4.20
4.08
3 . 83
3 . 33
3.91
3 . 57
4 . 42
4.21
3 . 85
4.28
1 05
Perusal of Table 4.5 reveals that the t rend is fo r the freque ncy of MSU to
i ncrease steadily with ag e , thoug h the re are so m e fluctuations in this pattern ,
particu larly withi n the h ig h e r ag e g roups.
Table 4 . 6
Mean Sco res fo r U se o f Metacognitive Strategies b y Age G roups
Classroom Learn e rs
Age G roups
Strategy
<2 1
N=84
2 1 -30
N=43
3 1 -40
N=9
4 1 -50
N=5
Self- manag e me nt
Adv. Organi sati on
Delayed P roduction
Revision
Self-monitori n g
2.91
3.77
3.47
2.76
3.83
3. 1 8
3.65
3.69
2.83
4.07
4.22
3.55
4.1 1
3.22
3.77
4.40
4.20
4 . 40
3 . 40
4 . 20
Perusal of the fig u re s i n Table 4.6 reveals that classroom learn e rs disp lay steady
increases in MSU with age, particularly fo r se lf- manageme nt, delayed producti on,
revision and self- m o n itoring.
1 06
Table 4.7
Mean Sco res for U se of Metacog n itive Strategies by Age Groups
Distance Learners
Age G roups
Strategy
<21
N=22
2 1 -30
N=73
3 1 -40
N =74
41 -50
N=68
5 1 -60
N=24
>60
N=1 4
S elf-man agement
Adv. O rganisation
Delayed P roductio n
Revision
Self- mon itori ng
3.54
4.27
3.72
3.45
3.63
3.64
4.08
3 . 53
3 . 30
4.06
4.06
4.31
4.08
3.08
4.04
4. 1 4
4.33
3.79
3.42
4.32
4 . 20
4 . 08
3 . 83
3 . 33
3.91
3 . 57
4 . 42
4.21
3.85
4.28
Perusal of the means prese nted earl i e r i n Tab le 4.6 sugge sts that the mai n
separatio n fo r learne rs i n te rms of M S U occ u rs between the u nder t h i rty and the
ove r t h i rty age g ro u ps . To dete rm i n e e xactly where the sig n ificant diffe rences
between the age g ro u ps are located , Du ncan's m u ltiple-range test (alpha=0 .05)
was applied to the fi rst canonical variate. The resu lts are p rese nted in Table 4.8.
1 07
Table 4 . 8
D u n can's Test fo r Compariso n of Means
Age G roups for the Fi rst Canonical Variate
Du ncan
G rouping
Mean
Age G roup
A
A
A
A
B
B
.59
.52
.39
.35
-.26
-.57
72
14
82
24
115
1 06
41 -50
>60
31 -40
51 -60
21 -30
<21
I n Table 4 . 8 , age g roups with the same lette r have m eans which are not
sig nificantly diffe re nt. lt is evident th e n that t h e re is a statistically sig nificant
diffe re n ce o n MSU m easu res between the learners who are u nder thi rty and the
rest of the popu latio n . Figure 4.6 shows the separation of the u nd e r thi rty and
ove r t h i rty age g roups in te rms of MS U . CVA i ndicated that this se paration can
be l arg e ly attri buted to the use of the self- manag e ment strategy.
4.2.5 Summary
Mode of study and the age of learne rs were found to be the p ri nci pal i nfluences
on M S U m easu res. C lassroom and distance learn e rs we re diffe re nti ated i n terms
of thei r use of se lf-management, advance org ani sati o n , revision and self
monitori n g . As far as the i nflue nce of ag e o n MSU variables was concerned, the
most marked separation between ag e g roups occu rred for learn e rs u nder thirty
and ove r thirty, i rrespective of mode of study. Differences i n MSU acco rd i n g to
age were related to t h e use of self-manag e ment, de layed p roducti o n , advance
org a n i satio n and revision. The tende ncy was fo r the freque n cy of use of t hese
m etacog n itive strateg ies i ncreased with age.
CVA of M S U Va ri a b l es
Lea rn e rs U n d e r Th i rty a n d Over Th i rty a s Groups
1!1
1!1 .
-2
1!1
-3
-4
-4
-3
-1
<
30 y e a r s
Figure 4.6
>
30 years
1 09
Havi n g established the i n fluence of mode of study on MSU , the next stage was
to determ i n e which vari ables i nte ract with mode of study to contribute to the
diffe re nces i n MSU
who show a marked diffe re nce i n MSU acco rdi ng to mode of study.
Te n sets of CVA we re perfo rmed in which mode of study was co mbined with
each of the learn i n g co ntext and learn e r characte ristic variables. P lots of the
values of the cano nical variate sco res were o btai n ed to aid i n the i nterp retation
of the fi ndi ngs. Analysis of the plots revealed that fo ur variables i nfluenced the
separation betwee n classroom learne rs and distance learners o n MSU meas u res.
These were the leve l of study , proficie ncy , the TL and prior expe rience in learning
the TL.
Listed be low are the values for each of the key variables which appear to further
i nflue nce the separatio n of cl assroo m and distance learn ers o n MSU measu res.
Level of study, for example, is re prese nted by t h re e values: 1 00-level, 200-level,
300-leve l.
110
Table 4.9
Value s fo r Variables I nteracting with Mode of Study and MSU Variables
Variable
Value
Leve I of Study
1 00-level
200-leve l
300-leve l
Proficie ncy
'A'
' B'
'C'
' D'
French
German
J apanese
Chi nese
Yes
No
Further canonical vari ate analyses we re perfo rmed to ide ntify which of the above
values m axi mi ses t h e difference between classro o m and distance learn e rs for the
MSU m e as u res.
As indicated i n Tab le 4 . 9 , there are t h ree val ues for the leve l of study vari able.
CVA was pe rfo rmed to determ i n e at which leve l of study , class room and distance
learners were maxi mally separated in terms of thei r MS U . An analysis of the p lots
of the cano n ical variate scores revealed that the g reatest diffe re nce occu rred at
the 200-level (Figu re 4. 7) . All the variatio n between the two g roups is captu red
by the fi rst canon ical variate , that is, along the h o rizontal axis. This is true fo r all
the analyses presented in the re mai nder of the chapte r.
CVA of
Va ri a b l es
MSU
3
1!1
2 -l
Q)
0
c
0
of-
1!1
>
1!1
1!1
0
c
1!1
iJ
-
1!1
1!1
'
1!1 1!1
1!1
1!1
1!1
1!1
..
1!1
1!1
1!1 .
1!1
1!1
1!1
-2 -l
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
"
1!1
-3
-3
-2
-1
Le a r n i n g
!!l
!!l
!!l
Classroom
Figure 4.7
Distance
112
The standardised coefficie nts and co rre lation coefficie nts presented i n Table 4. 1 0
show the contributi on of each of the ten m etacog nitive strategies to the
separation of classroom and distance learners at the 200-leve l . The sig nifi cance
of the can o n ical variate sco res was evaluated usi ng the F distributi o n o btai ned
t h rough AN OVA : F=43 .90 (1 , 1 1 2) , p<.000 1 .
Table 4 . 1 0
Standardised Coefficie nts and Co rre lati o n Coeffi cients for MSU Measu res
200-level Classroom and Distance Learners as G roups
Strategy
Standardised
Coefficie nts
Advance O rg a nisatio n
Selective Atte nti on
Di rected Atte ntion
De layed P roduction
Self-man ag e ment
P roble m Ide ntificatio n
Self-monitori ng
Self-evaluatio n
P riorities
Revisi o n
.91
-.1 3
.35
.12
.1 5
.05
-.05
.10
.01
.29
Corre latio n
Coefficie nts
. 84
.13
.47
. 24
. 44
.24
.28
.48
.23
.49
A n analysis of the coeffici ents i n Table 4 . 1 0 reveals that the key strategies
i nflu encing the separation of classroom and distance learners both se parately and
jointly at the 200- leve l are advance o rgani sati o n , revisio n , se lf-evaluatio n , di rected
atte nti o n and self- management. The
o rganisation strategy ( . 9 1 , .84) show the hig h contri buti on it makes to the
separatio n of learn e rs in the plot (Fig u re 4.7) .
113
Fig u re 4 . 8 prese nts a co mparison of mean strategy use fo r the five key M S U
measures for class ro o m a n d distance learners at t h e 200-leve l.
:50<5
<:5<5
:so<
38
<:5<5
<:5<5
:so<:
:so<
>Q<;
0
Mv.
9'1
Orgonieotn.
Rwiaion
Self-8'J'OIU<rlion
l:sO<:so<:
lm:s<!5/l
Dir. Attention
Self-management
Fig u re 4.8
114
4.3.2 Proficiency
Stand ardi sed and co rrelation coefficie nts were o btai ned and are listed in Table
4. 1 1 .
Table 4 . 1 1
Standardised Coefficients and Co rre lati o n Coeffi cients fo r MSU Variables
G rade 'B' Classroom and Distance Learners as G ro u ps
Strategy
Advance Organisation
Selective Attention
Di rected Attention
Delayed P roductio n
Self- manag e me nt
P roblem Identification
Self-monitoring
Self-evaluation
P riorities
Revisio n
Standardised
Coefficie nts
Co rrelati o n
Coefficie nts
.58
.04
.1 8
-.1 8
.57
-.1 2
.1 4
-.01
-.1 0
.39
.71
. 43
.55
.06
.67
.18
.29
.37
.21
.62
AN OVA perfo rmed on the cano ni cal variate sco res revealed that the separation
of g rade ' B' classroo m learne rs fro m g rade 'B' distance learn e rs on MSU
measu res i s hig h ly sig nificant : F=46.63 ( 1 , 1 30) , p<.000 1 .
CVA of M S U Va ria b l es
3 .
2
Q)
0
c
0
1!1
1!1.
1!11!1
1!1 1!1
1!1
0
c
1!1
1!1
1!1
-2
-3
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
1!1
""0
c
'
;::
(/)
1!1
1!1
1!1
0
Q)
1!1
>
()
1!1
.....
1!1
1!1
1!1
1!1
!J
..
1!1
-4
-4
-1
-2
-3
Le a r n i n g
l!l
l!l
l!l
Classroom
Figure 4.9
Distance
1 16
::
!\X(
1%(
:><X
hN.
Organlsotn.
Self-management
Fig u re 4. 1 0
Revtslon
117
4.3.3 Target Lan guage
I n terms of the i nflue nce of the TL on MSU, the most marked effects take place
when the TL is J apanese. The resu lti ng separati o n of classro o m and distance
learne rs is d isplayed i n Figure 4 . 1 1 .
The standardised and co rre lati on coefficie nts p rese nted i n Table 4 . 1 2 reveal that
th ree strategies are p ri ncipally responsible fo r t h e separation displayed i n Fig u re
4. 1 1 : advance o rgani satio n , se lf-manag e ment, and revision. Th e contri buti on
m ade by t h e advance organisation strategy e xceeds that made by other
m etacog n itive strategies.
Table 4 . 1 2
Standardised Coefficie nts and Corre lati on Coefficients for M S U variables
Class roo m and Distance Learne rs of Japanese as G roups
Strategy
Advance Organisation
Selective Atte ntion
Di rected Attention
Delayed P roduction
Self- m an agement
P rob le m Ide ntification
Self- m o n ito ring
Se lf-evaluation
P rioritie s
Revision
Standardised
Coeffici e nts
.80
-.25
.29
.05
.37
-.1 4
-.1 3
.06
-.31
. 38
Co rrelation
Coefficie nts
.81
.1 4
.44
.1 6
.59
.12
. 23
.29
.00
.50
The sig nificance of the canon ical variate sco res was evaluated using the F
distributio n obtai ned throug h AN OVA : F=52.03 ( 1 , 1 56), p<.000 1 .
CVA
of M S U
Va ria bles
Q)
-+J
0
c
0
>
0
ij
c
0
c
0
l!l
()
'"0
c
0
l!l
-2
j
1
l!l
l!l l!l
l!l
l!l
l!l l!l
. l!l
l!l
l!l
l!l .
l!l
l!l
l!l .
l!l
l!l
l!l
_._____Il
l_
1!1
l!l .
l!l
l!l
. l!l ltl
l!l
l!l
l!l
l!l
. .,
"
. . .
l!l l!l
'
l!l
..
l!l
-3
-4
-3
-2
-1
Le a r n i n g
l!l
l!l
l!l
C l assroom
Figure 4. 1 1
Distance
1 19
A com parison of the use of t h e t h ree strategies which diffe re ntiate classroo m
learne rs o f J apanese fro m distance learn e rs o f J apanese is presented i n Fig u re
4. 1 2.
<95
99
<95
0
CkiMrOOm Dleta
Mv. OrganJ.atn.
ClaMroom Dlatance
Self-
Fi gu re 4. 1 2
38
ClcHroom Dim!
lalon
1 20
4.3.4 Prior Target Language Experience
Classroom and d i stance learners who had no p rior experie nce in the TL before
e n ro l li ng i n a lang u ag e cou rse at Massey U nive rsity were maxi m ally disti nguished
in terms of their M S U . This separation is p rese nted i n Fig u re 4 . 1 3 .
Table 4 . 1 3
Standardised Coefficie nts and Co rre lati on Coefficients fo r M S U Vari ables
Classroom and Distance Learners with No P rior TL Expe rie nce as G ro u ps
Strategy
Advance Organisation
Selective Atte nti on
Dire cted Atte ntion
De layed Production
Self- m anagement
P ro b l e m Ide ntification
Self- m o nitori n g
Self-evaluati on
P ri o riti es
Revision
Standardi sed
Coefficie nts
.41
-.04
.41
.23
.70
-.28
-.01
-.08
- . 33
.27
Co rrelation
Coeffici e nts
.57
.23
.49
.35
.76
-.1 5
.14
.1 9
-.07
. 38
The standardised and co rre lati on coefficie nts p rese nted in Table 4 . 1 3 reveal that
t h re e strateg i es co ntri bute to the separation displayed i n Figu re 4 . 1 3 : self
m anagem ent, advance o rganisation and di rected atte nti o n . I n t h i s case the
g reatest joi nt ( . 70 ) and i ndividual (.76) co ntri bution is made by the self
m anag e ment strategy.
CVA of M S U Va ri a b l e s
C l a s s ro o m a n d D i sta n c e Le a rn e rs w i t h N o P ri o r T L E x p e ri e n c e a s G ro u p s
2 -,
1!1
Q)
......
0
c
0
>
0
()
c
0
c
0
u
1!1 '
1!1
1!1
1!1
1!1
1!1
1!1
1!1
'U
c
0
()
Q) - 2
(J)
1!1
fi
!!JI!I
1!1 1!1
1!1
1!1
1!1
1!1
1!1
1!1 .
'
1!1
1!1
1!1
-3
., .
..
1!1
1!1
1!1
1!1
11
1!1 1!1
1!1
1!1
1!1
-4
-4
-3
-2
-1
First C a n o n i c a l Va ri ate
Mode
of
Learn ing
l!l
l!l
l!l
Classroom
Fig u re 4. 1 3
e e
Distance
1 22
:5<5<
I)
::1
!
c
::E
:888
0
Cloearoom
Dietonce
Self-management
Claeeroom
D1st<1no.
Adv. Or9anfsatn.
Claearoom
Dletonce
Dlr. Attention
Fi g u re 4. 1 4
F values were o btai ned t h rough ANOVA performed o n the canon ical variate
sco res. They were found to be sig nificant at the .0001 level : F=59.51 ( 1 , 1 24) ,
p<.0001 .
1 23
4.3.5 S u m mary
CVA revealed that mode of study had the g reatest effect on the MSU of
u nderg raduate fo reign lang uag e students. The key strategi e s contri buti ng to the
separation of classroom and distance learn e rs were the use of se lf-manage ment
and advance o rganisatio n , and to a lesse r extent revisi on.
The d iffe re nces betwee n class roo m and distance learners i n te rms of their MSU
were particularly notable in fou r ci rcu mstances: fi rstly wh e n languag e learning
took place at the 200- level, secondly when the TL was J apan ese , thi rd ly fo r
learn e rs who achieved a proficie ncy level of 'B' and fi nally wh e n learn e rs had no
p rior experi ence of the TL befo re e n rolling i n the university languag e co u rses. I n
the fi rst t h ree sets of ci rcu mstances the strategy wh ich had the g reatest i m pact
o n t h e diffe re ntiation of classroom and distance learners was the use of advance
o rganisatio n . When learne rs had no prior experience of the TL, classroom and
distance learn ers were maximally separated in terms of thei r use of se lf
manag e ment.
1 24
4.4 COG N ITIVE STRATEGY USE
4.4.1 Freq uency of Cognitive Strategy Use
Table 4 . 1 4 l i sts descriptive statistics fo r the cog nitive strategy use (CSU)
measures fo r the underg raduate lang u ag e l earn e rs i n this study.
Table 4 . 1 4
Freque ncy of Cog n itive Strategy Use
( Means and Standard D eviations N=41 7)
Strategi es
Repetition
Resourcing
G roupi ng
N ote-taki ng
Deduction
S u b stitution
E laborati on-I magery
Visualisation
E laborati on-World
E laborati on- Parts
Contextuali sation
S u m marisation
T ranslation TE
T ra nslation FE
l nfe renci ng
T ransfer
R e hearsal
Mean
S.D.
3.94
4.21
2.93
3.42
3.55
4.00
3.48
3.60
3 . 54
3.93
3.05
3 . 28
3 . 84
3 . 68
4 . 42
4 . 02
2 . 99
.91
.94
1 . 09
1 .1 0
.98
.94
1 . 26
1 .1 2
1 . 09
.93
1 .02
1 .1 0
.99
1 .1 0
.72
.94
1 .1 6
Refe rri ng to Table 4. 1 4 , the fou r m ost frequently used cog n itive st rategies are
infe re nci ng , resou rci ng , transfe r and sub stitutio n . The mean respo nse for these
strategies re lates closest to the categ o ry of 'ofte n'. The least u sed st rategi es are
re hearsal and g roupi ng which are used l ess than 'someti m es'.
1 25
4.4.2 I nfl uences on Cogn itive Strategy Use
A si mi lar procedu re was followed to assess the re lationsh i p betwee n the learner
characte ristic variables and CSU. The most m arked separation of l earners
occu rred i n terms of their p rior TL experience. N o effects we re fo u nd for the
variables of age , gender, proficie ncy , oth er lang uage learn i ng o r motivatio n .
Mode of Study
Target Language
Leve l of Study
Language Use Opportu nities
Learn e r
Characte ristic
Variables
Age
Gender
P roficie ncy
P rior Target Lang uage Experience
Other Lang uag e Lea rn i n g
Motivation
2
4
3
The two most important variables, i n te rms of t h e i r influ e nce on CSU m easu res,
name ly prior TL experience and mode of study, were then se lected for further
analysis.
1 26
4.4.3 The I nfluence of Prior TL Experience o n CSU Variables
I n o rde r t o assess t h e sig nificance o f t h e cano nical variate sco res , F values were
o btained t h ro u g h ANOVA: F=64.86 ( 1 ,41 1 ) , p<.000 1 .
T h e next stag e was to ide ntify wh ich cog nitive variables co ntri bute to the
separation in Fig u re 4. 1 5 . Standardised coefficie nts and corre lation coeffici e nts
w e re obtai n ed and are listed i n Table 4. 1 6 .
Table 4 . 1 6
Standardised Coefficie nts and Corre lation Coefficients for CSU Measu res :
P rior TL Experie nce as G roups
Strategy
Re petitio n
Resourcing
G rouping
Note-taki ng
Deductio n
Substituti o n
E labo ration-I mag ery
Visualisatio n
E labo rati on-World
E labo ration- Parts
Co ntextualisatio n
S u m m arisation
Translatio n TE
Translatio n FE
l nfe re nci ng
Tran sfe r
Rehearsal
Standardised
Coefficie nts
-.06
.61
-.24
.1 2
.31
.29
- . 20
.09
.21
-.23
.1 1
-.1 6
- . 25
- . 25
. 08
.04
- . 25
Co rrelation
Coefficie nts
.04
. 54
-.1 7
.07
.23
.53
-.1 3
.08
.25
-.20
.1 5
-.22
- .39
-.42
.09
.1 1
-.24
CVA of C S U Va ri a b l e s
P ri o r T L E x p e ri e n ce a s G ro u ps
4
1!1
2
Q)
1!1
>
+J
1!1
1!1 1!1
1!1
1!11!]
0 -1
()
1!1
1!1
1!1
'
() -2
Q)
(/)
1!1
I!I L!P .
'
u -':
" 1!1
1!1
1!1 l!l
l!l i
1!1i
!!l.ri . 1!1
I!Plii!J
\
!!!Jr,
1!1
..
1!] 1!1
1!1 1!1
1!1 1!11!1
1!1
1!1 1!1 1!1
1!1 1!1
1!1 1!1
1!1 1!11!] 1!1 1!1
1!1 1!1 1!1 1!1
') 1!11!] 1!1 1!1
1!1
1!1 1!1
1!1
1!1
1!1 1!1
1!1
1!1
1!1 1!1 !!l l!fJ
1!1
1!1 1!1 1!1
1!1
1!1 1!1 1!1
1!1
1!1 1!1 1!1
1!1 1!1
1!1
1!1
1!1 . 1!1
1!1
1!1
1!1
1!1
"'0
c
0
()
c
0
c
1!1
r!'
1!1
-4
-5
-4
-2
-1
in
TL
learn ing
Fig ure 4. 1 5
!!l
!!l
!!l
Yes
No
1 28
Perusal of t h e coefficie nts i n Table 4 . 1 6 i ndicates that the two main cog nitive
variables which contribute individually and jointly to the separation of the g roups
i n Fig ure 4. 1 5 are resou rcing and su bstituti on. The re is also a joint contribution
made by deduction (.31 ), though i ndividual ly this strategy is not important.
Relatively hig h coeffici e nts appear fo r tran s lation both to and fro m E ng lish. Th e
translatio n strateg ies have negative values for both the standardised coeffici ents
(-.25) (-. 25) and for the co rre lation coefficie nts (-. 42) (-.39) . This i ndicates that the
use of translatio n both to and fro m E ng lish contributes to the separation of
learners acco rd i ng to pri o r TL experi e nce , b ut that the co ntri bution operates in the
opposite di recti o n to that of strateg ies such as resourci ng and su bstitutio n . This
can be seen by refe rri ng to Fig u re 4 . 1 6 .
:so
%88:
%<
....
Reaourctr.g
Subatttutn.
Fig u re 4. 1 6
Translatn.
FE
Transkrtn. TE
1 29
Figu re 4 . 1 6 reveals that l earners with pri o r experi e nce of the TL make more use
of resou rcing and su bstitution than do learners without such expe ri e nce. On the
othe r hand they m ake less use of translation, both t o and from E ng li s h , than d o
learners with no p ri o r experie nce of learni ng t h e T L before e n ro l l i ng i n t h e
u nive rsity language cou rse.
I n section 4 . 4.2 the i nfluence of mode of study o n CSU was investig ated and
some se paration of classroom and distance learn e rs was fo u nd . This i s displayed
in Fig u re 4. 1 7.
From a com parison of Fig u re 4. 1 (presenting the influe nce of mode on MSU
measures) and of Fig u re 4. 1 7 (showi ng the i nflue nce of mode o n CSU
measures) , it can be seen that the i nflue nce of mode of study is more marked fo r
MSU than fo r CSU . This is also i ndicated by the fact that the m eans on the fi rst
canonical variate are fu rther apart fo r the i nflue nce of mode on MSU measures
than for the means re lati ng to the i nflu e nce of mode on CSU m easu res. A
comparison of the means is p rese nted i n Tab le 4 . 1 7 .
Table 4 . 1 7
Class Means on Canonical Variate One
MSU Measures and CSU Measu res
Mode of Study
C lassro o m
Distance
-.61
.31
- . 56
.28
To evaluate the significance of the fi rst can o n ical variate sco res F values were
obtai ned thro u g h AN OVA : F=64.86 (1 ,41 1 ) , p<.000 1 .
[!)
[!)
Q)
0
c
0
+'
>
lli!J
c:
c
0
"'0
Q)
(/)
[!)
l!l
l!ll!l
[!)
[!)
-2
[!)
[!) [!)
l!l
l!l
l!l
l!l
l!l
l!l
l!l l!J
l!l
l!l
'i!i
..
[!) [!)
[!)
[!)
[!)
l!l l!J
[!)
[!)
[!)
l!l
[!)
[!)
l!l
l!ll!l
l!l
l!l
. .
,_
. ...
[!)
.
l!l A .
,..
[!) . l!l
l!l
l!l
l!li!l.
l!l [!)
-4
l
-2
-3
. .
[!)
t;i
-1
l!l
Le a r n i n g
[!)
[!)
[!)
Classroom
Fi g u re 4. 1 7
[!)
[!)
l!l
,
'
11\
l!l
l!l
. , l!l . ..
...
.l!l
[!) [!)
.. -
.,..
...
""
'.l!l ... ... . 1!1 . . .. :
.
.
[!)
[!)
1!1
efil
'
-3
Distance
1 31
I n o rder to discove r which set of cog nitive variables is influe nci n g the tendency
for classroom and distance learne rs to separate i n Fig u re 4 . 1 7 , standardised
coefficie nts and co rrel ati on coeffici e nts were o btained. These coeffici e nts appear
in Table 4. 1 8.
Table 4 . 1 8
Standardised and Corre lation Coeffi ci e nts fo r CSU Variables
Mode of Study as G roups
Strategy
Standardi sed
Coefficie nts
Repetition
Resou rci ng
G rouping
N ote-taki ng
Deduction
Substitution
E labo rati on-I magery
Visualisation
E labo rati on-Wo rld
E labo rati on- Parts
Contextualisation
S u m m arisation
Translation TE
Tran slation FE
l nfe re nci ng
Transfer
Rehearsal
. 36
. 49
.1 6
. 05
.09
-. 1 3
-.04
-.01
.28
.02
-.1 6
.01
-. 1 1
-.07
.02
.37
.32
Corre latio n
Coefficie nts
.58
.62
.39
.32
.37
.04
.21
.21
. 35
.22
.1 1
. 33
-.05
-.02
. 20
.51
.46
Refe rri ng to Table 4. 1 8 , three cog n itive strateg ies co ntri bute i ndividual ly and
jointly to the separation of classroom and distance learners i n Fig u re 4 . 1 7. These
are resou rci n g , repetition and t ransfe r.
1 32
A co mparison of mean frequencies of classroom and distance learn e rs for their
use of these t h re e strategies is displayed in Fig u re 4 . 1 8.
:s:
B:
XXX:
0
Claun>om
llstance
R980urcing
ClaAroom
Dt.tanc.
Repetition
Claaroom
Distance
Tronr
Fig u re 4.1 8
The investig ation of the influe nce of m ode of study on C S U m easu res was
fu rth ered u s i n g procedu res ide ntical to t h ose described i n secti o n 4.3. Th is was
to discove r if there were marked effects for mode on CSU o nce t h e influence of
other variables, such as prior experi e nce , was take n i nto accou nt. However, no
fu rther marked separation of classroo m and distance learners was fou nd.
1 33
4.4.5 Summary
Mode of study had less i m pact o n CSU measures than o n MSU meas u res.
Diffe rences o n CSU measu res according to mode of study we re dominated by
t h e use of resourci n g , re petition and transfe r.
1 34
4.5 SOCIAL STRATEGY USE
4.5.1 F req uency of Social Strategy Use
As disc ussed in sectio n 4. 1 , the S S U scale i s a catego rical scale, and thus it is
not appro p riate to o btai n statistics such as means and standard deviations from
the scale. A g raph ical re prese ntation of the frequency of use of the questio ning
strategy among the sample as a whole is p resented in Fi g u re 4 . 1 9 . The most
freque nt respo nse categ ories we re 'someti mes' (26.9%), ' rarely' (22.5%) and 'no
opport u n ity' ( 1 8. 2%) .
Fre q u e ncy of Use of Q uestio nin g
1 20
1 00
80
2: 8i
'0 60
]
E
:3
40
:822
Q9<
Q9<
20
No oppo rtunity
Never
Rarely
Sometimes
Questioning
Fig ure 4. 1 9
Often
Very often
1 35
A display of the f requency of use of co-operation by foreig n lan g u age learners is
p rese nted in Fig u re 4.2Q. The co-operation strategy was u sed eve n less that the
questioni ng strategy. The 'no opportu nity' response was the most frequent and
amou nted to 43. 5 percent of the responses. The next most frequent responses
we re 'someti m es' and ' rarely'.
Fre q u e n cy of U s e of Co - o p e ration
180
1 60
1 40
1 20
80
60
40
20
0
No opportunity
<:><>0000000<1
Never
Rarely
Sometimes
Co-operation
Fi g u re 4.20
Often
Very often
1 36
4.5.2 I n fl uences o n Social Strategy U s e
80
70
60
I!!
I)
'0 40
...
..8
30
10
0
1-----
Cloasroom
---
..___
_
_
_
Dlstanee
----1
Fig u re 4.2 1
The respo n se frequencies displayed i n Fig u re 4.21 reveal that t h e most frequent
respo nse by distance learners was t h at they did not have any opportu nity to use
the questioning strategy. The next most frequent response categories we re
'rare ly' and 'so meti mes'. The low p roportion of 'never' responses compared to 'no
opport u n ity' suggests that distance learners wou ld make use of social strategies
1 37
if they were avai lable. Fo r classroom learn e rs t h e m ost frequ e nt response was
'so meti m es' and most respo nses occu rred towards the higher e n d of the
respo nse scale.
The response for the use of co-operation as a strategy in lang uage learning is
displayed in Fig u re 4.22.
Freq u ency of Use of Co - o peration
Cl ass room a nd Dista nce Lea rners as Gro u ps
1 60
1 40
1 20
....
0
L
.8
E
::I
z
1 00
ao
eo
40
20
0
RIO
t----
Dlstance
<'I
----1
Fig u re 4.22
The patte rn is s i m i l ar to that for the questio n i ng strategy , but diffe re nces b etwee n
classroom and distance learne rs are more p ro no u nced. For distance learn e rs the
most frequent response was 'no opportu nity', and this was also the m ajo rity
respo nse. The most freque nt respo nses for classroom learners we re 'so meti mes'
and ' rare ly'. For both classroom and distance learne rs the ' n o opport u nity'
respo n se was m o re f reque nt than the ' never' response .
1 38
Freque ncy charts we re the n obtai ned fo r t h e use of questioni ng and co-ope ration
according to the four TL g roups : Fre nch , G e rman , Japanese and C h i n ese. Each
TL g roup s h owed very simi lar respo nses with respect to thei r use of the two
social strategi e s . The patterns of respo n se co rresponded close ly to those for the
sample as a whole (see Fig u re 4 . 1 9 and Fig u re 4.20). Thu s , learn e rs were not
diffe re ntiated i n their SSU according to TL.
4.5.3 Summary
Classroom and distance learne rs d iffe r i n te rms of their use of social strateg ies.
Respo ndi n g to questions about their use of the st rateg ies of q ue stioning and co
operatio n , distance learners noted that t h ey did not have the opportu nity to make
use of these strategies, particu larly wit h respect to the co-operati o n strategy. The
respo nses of classroom learn ers g e n e ral ly ranged between 'so metimes' and
' rarely'. Learn e rs of different Tls we re n ot fou nd to diffe r in thei r use of the social
strategies of questioning and co-op e rat i o n .
1 39
4.6 AFFECTIVE STRATEGY USE
4.6.1 F req uency of Affective Strategy U se
The affective response scale was ide ntical to that fo r the MSU measu res and the
CSU measu res , so it was possible to co m pute means and standard deviations
fo r each of the t h ree affective strateg ies. Th ese are listed i n Table 4. 1 9 .
Table 4 . 1 9
Freque ncy of Affective Strategy Use
( Means and Standard Deviations N=41 7)
Strategy
Mean
S.D.
S elf-talk
S e lf- rei nfo rce ment
Self-e ncou ragement
1 .6 1
2. 1 6
2.48
.95
1 . 22
1 . 20
U nderg radu ate fo reig n language learn e rs made infrequent use of affective
strategi es, that is, strateg ies which invo lved managing their fee lings about the
languag e learn i ng p rocess, with mean re spo nses co rrespo ndi ng closest to the
category of ' ra re ly'. The least used affective strategy invo lved tech niques to lowe r
anxiety (se lf-talk) while the most freque ntly used affective strategy was self
e ncourag e m e nt.
CVA was carried out to i nvestigate w h ether learners we re differe ntiated in their
ASU according to mode of study o r t h e TL. Results i ndicated t h at neither mode
of study n o r the TL was found to exert a sig nificant influe nce on ASU measu res.
1 40
4.7 SU M MA R Y
The resu lts f ro m the analysis o f t h e q u e sti onnai re data reveal t h at the pri ncipal
i nflue nces o n MSU are mode of study a nd the age of learne rs .
Distance l earn ers made g reater u s e o f metacog nitive strate g i e s than classroom
learne rs , m o st notably with regard to t he strategies of se lf-management and
advance o rg a nisatio n and, to a lesser e xtent, revisio n. Fou r sets of circu mstances
were ide ntifi ed in which the diffe re ntiatio n of classroom and distance learners on
MSU measu res became particu larly p ronou nced. These we re when the TL was
Japanese , when the proficie ncy level o f learners was
'8',
w h e n learners had no
'8',
and fo r learners of
Japanese. When learners had no p rior experie nce of the TL b efo re e n rolli ng in
a u nive rsity lang uage p rog ram me, distance learners made significantly g reater
use of self-manag ement than classroom learners . To sum marise , t h e n , the i m pact
of mode of study o n MSU was p ri mari ly associated with the se lf-manag e m ent
strategy and the use of advance o rganisati on.
As far as the influe nce of age o n MSU measu res was co nce rned , resu lts
i ndicated that l earners ove r thirty made significantly greater use of metacog nitive
strategies, particu larly of self-manage m e nt , than learne rs u nde r thi rty. This was
equal ly true fo r classroom and di stance learners.
The main i nfluence on CSU measu res was very clearly pri o r TL experience.
Learn e rs who had had such prior experie nce befo re e n ro l l i n g i n a u niversity
fo reig n langu ag e cou rse we re m aximally distingu ished from learne rs without such
prior experi e n ce in te rms of the i r g reater use of resourci n g and substitution and
t h e i r decreased use of translation both to and from English.
1 41
Mode of study exerted some i nflue nce o n CSU measures, but t h i s was less than
the i nfluence of p rior TL expe ri ence . Classroom and distance learn e rs cou ld be
co ntrasted i n te rms of a more frequ e nt use of resou rci ng , re petiti o n and transfer
by distance learners. The i nflue nce of m ode of study on C S U m e asures was
much less appare nt than the i nflue nce of mode of study on M S U measu res.
U nderg radu ate foreig n language learn e rs were co mparable in th e i r i nfreq uent use
of affective strateg ies, irrespective of t h e i r learn i ng context (e.g . , mode of study ,
TL) or i ndividual characte ristics (e.g . , ag e , g e nder, motivat i o n ) .
Resu lts fro m the ve rbal report study a re prese nted i n t h e followi ng ch apte r.
1 42
5. RESU LTS : VER BAL R E PORT STU DY
T h e early part of this chapter outli nes the methods u sed to analyse the ve rbal
re po rt data (section 5. 1 ) and apprai ses the effective n ess of the yoked subject
technique i n terms of p roductivity (section 5 . 2 ) . The classification of instances of
strategy use which were not part of the qu estion naire study is explai ned through
d efi nitions and ve rbati m extracts fro m the reports (section 5.3). The fi ndings are
p resented i n terms of the infl uence of mode of study o n strategy use (section 5 . 4)
and the i nfluence of the TL on strategy use (section 5.5) . The summary
u nderlines the mai n fi ndi ngs from the ve rbal report study (section 5 . 6 ) .
1 43
e xample , i n the fo l lowi n g e xtract the underli ning strategy, which is a form of note
taking , was m e ntioned twice but was recorded as a single i nstance of strategy
use.
1 FD
Some strategies re ported by learners did not easi ly fit i nto descri pti ons found i n
t h e literatu re , and i n t h ese cases new strategy n ames and defi nitions we re
developed to match t h e descri ptions i n the transcripts. Examples of these n ewly
ide ntified strategi es, such as time lapse and oth e r- rei nfo rcem ent are p rese nted
in section 5.3.
Fi rstly, the two rate rs t ransfo rmed the raw data i ndependently, ide ntifying and
classifyi ng instances of strategy use. Several descriptions of the use of each
strategy were extracted from the reports and put i nto a separate fi le. An e xample
fro m the p roblem ide ntification file fo l lows.
1 44
Problem Identification
Defi n ition :
Explicitly ide ntify i ng the central poi nt need i n g resoluti o n i n
a task o r identifying a n aspect of t h e task t h at hind e rs its
successfu l co m pleti on. O'Malley and Chamot ( 1 990 : 1 37)
E xamples:
1 FD
8FD
7J D
The verbs are usually the difficult point that I can 't get clear
in my mind. That's what lets me down. It is the verbs I get
wrong in my written work and in my conversation.
1 1 JD
E ach rater was then able to compare diffe re nt instance s of the u se of a particu lar
strategy type , and to ch eck that each i nstance was co n siste nt with the defi niti o n .
T h is was o n e way of e n s u ri ng i nt rarate r re liabi lity.
1 45
l ntrarater reliability was also assessed by both rate rs re-analysing the reports i n
a diffe re nt order usi ng fresh t ranscripts t o com pare the deg re e o f ag ree ment
which existed betwe e n the fi rst and seco nd an alyses. Each rate r then atte mpted
to reso lve any disparities betwee n t h e i r fi rst and seco nd analyses.
Up to this poi nt analysis of the verbal rep o rt data was carried out by the two
raters worki ng i nde pende ntly. The next stag e was fo r the two raters to co mpare
their a n alyses. Any differences i n classification were noted and ag ree ment was
reached through discussion. The m ajo r difficu lties re lated to dete rmi n i ng when
particu lar instances of strategy use were suffici ently frequ e nt and diffe rent fro m
the defi nitions which appeared i n the literatu re (Chamot et al. 1 987 ; Chamot et
al. 1 988a, 1 988b; Ellis and Si nclai r 1 989; O'Mal ley and Chamot 1 990 ; Oxfo rd
1 990) to warrant identification of a n ew strategy.
Final ly , descri ptive statistics (freque nci es and perce ntages) re lating to strategy
use w e re co mputed, and these were used to co mpare the range , type and
frequ e ncy of strategy use of classroom and distance learners, and of learners of
the two TL g roups (French and J apan ese).
'
The 37 ve rbal re ports we re analysed fol lowi ng the p rocedu res described i n
section 5 . 1 and were found t o co ntai n a total of 836 i n stance s of strategy use.
The foreign language learne rs i n this study related easi ly and readily to the yoked
subject technique and were able to talk about the tech niques they used to master
a fo reig n lang uage . The high n u m be r of instances of reported strategy use
i ndi cates that , in the context of the present study , the yoked su bject tech nique
was a p roductive m eans of o btai n i ng i nfo rmation about the strateg ies used by
foreig n language learners.
1 46
The rang e of strategies identified throug h the ve rbal reports exte nded beyond the
list used for the qu esti o n nai re data. This applied to all strategy g roupi ng s as
shown i n Tab le 5 . 1 .
Table 5 . 1
C o m parison o f the Range o f Strateg ies
Q uestionnai re and Ve rbal Report Data
Num ber of Strategy Types
Strategy G roupings
Questionnai re
Data
Yoked Subject
Data
Metacog nitive
Cog n itive
Social
Affective
10
17
2
3
24
28
3
4
Total
32
59
The increase in the ran g e of strategy types pertai ning to differe nt data collection
' procedures was particu larly pronounced fo r metacog nitive strateg ies. The number
of metacognitive st rategy types m o re t han dou bled fro m the questionnai re data
to the yoked su bject d ata.
1 47
The 32 strategies which formed part of the question nai re data were ide ntified and
defi n ed in sectio n 3 . 3. 1 . A further 27 strategies were added to that classification
sch e m e t h roug h analysis of the ve rbal repo rt data. This demonst rates that the
yoked su bject tech n ique can yield rich data re lated to questi o n s about how
stude nts manag e their lang uage learni ng. The additional strateg ies a re defi ned
i n the fo l l owi ng sectio n .
5 . 3 T H E CLASSIFICATION O F STRATEG I ES
5 . 3 . 1 Metaco g n itive Strategies
Fou rteen fu rther m etacog nitive strategies were ide ntified in the ve rbal report data.
Defi n iti o n s fo r these strategies appear in Table 5.2.
1 48
Table 5 . 2
Metacognitive Strategy Defi nitio ns
M etacognitive Strategy
Description
O rganisational Planning
Time Lapse
Analysing Needs
Comprehension Monitoring
P roduction Monitoring
Visual Monitoring
Strategy Monitoring
Plan Monitoring
Double-check M onitoring
P e rformance Evaluation
Ability Evaluation
Strategy Evaluation
1 49
Most of these strategies h ad bee n ide ntifi ed t h ro u g h the indicated sou rces :
O' Mal ley and Chamot ( 1 990) , Ellis and Si nclai r ( 1 989). Two strategy types which
did not appear to fit any of t h e previous classification schemes we re labelled time
lapse and seeki ng p ractice o pport u n ities.
Time Lapse
The use of the time lapse strategy i nvolved the deliberate structuri n g of t i m e to
allow fo r s paces to occ u r betwee n study sessions. This strategy was usually
e mployed i n three sets of circu m stances. The fi rst of these re lates to occasions
whe n l earners fou nd they were having difficu lty with a particu lar task o r aspect
of the TL.
2FD
The time lapse strategy was also used as a means of spacing either the amount
of learni n g , o r the le ngth of learning sessi o n s , when p ractising exte nsive liste ning
or learni ng new word s , fo r exam ple :
9J D
1 1 JD
I don 't try to learn too many kanji a t once - I just do a few
at a time, otherwise if I try to learn too many too quickly I
start forgetting the ones / learnt two o r three weeks earlier.
1 50
Ti m e lapse was also used i n a third way, close to a ki nd of m e ntal d rafting of
wo rk ove r ti m e , particularly fo r exte nsive writi ng.
1 FD
8J D
Seeking p ractice opportu nities invo lved eit h e r arranging extra activiti e s for the use
of the TL, or creati ng an e nvi ron ment which provided lang uag e i n put. Examples
of seeki ng p ractice o pportunities are :
1 1 FD
1 FD
I write out lists of key words I need to know and have them
dotted all over the house where I can find them
unexpectedly. The element of surprise helps me to
remember them.
1 51
9JD
The defi n itions p rovided in this section , together with the defi nitions of
metacognitive strateg ies in section 3 . 3. 1 co mprise the m etacogn itive classification
sch eme u sed in the analysis of the yoked subject data.
As shown i n Table 5 . 1 , the questi o n naire data co m prised 1 7 cog n itive strategy
type s and the verbal re port data co mprised 28. The qu esti o n naire CSU scale
d rew a disti nction between translatio n to E ng lish and t ranslatio n fro m English. I n
the ve rbal report classification scheme these diffe re nces we re su bsu med i nto o ne
catego ry for translatio n , becau se it was not always possible to t e l l whether the
translatio n was to or from E ng l i s h . The questi o n nai re CSU scale also
disti n g u i shed betwee n imagery and visual isatio n (see section 3 . 3 . 1 ) . H owever,
visualisation did not occ u r in the ve rbal re ports , tho u g h t h e re were five i nstances
'
Table 5 . 3 p rese nts descriptions o f cog nitive strategies which fo rmed part of the
ve rbal re port study, but which were not mentioned in section 3 . 3. 1 .
1 52
Table 5 . 3
Cog nitive Strategy Descriptions
Cog nitive Strategy
D escription
Note-taking
Writing Out
Listing
Noting Down
H ighlighting
U nderlining
Personal Elaboration
C reative Elaboration
R ead Aloud
Auditory Representation
Work in TL
Carrying out particu lar language activities in the TL i ncluding formulating ideas, note-taking, co mpiling word
lists.
Memorisation
Recombination
1 53
The note-taki ng strategy defi ned i n sectio n 3.3. 1 was fu rther diffe re ntiated i n the
cou rse of analysing t h e ve rbal report data. A n u mber of variati ons e m erged i n
terms of the nature , p u rpose , and exte nt of note-taki ng which made it n ecessary
to refi n e the single b road categ o ry used fo r the questionnaire scale . This was
also i m po rtant since n ote-taki ng proved to be the m ost frequent cog nitive strategy
fo r both classroom and distance learners.
A n u m b e r of e xtracts fro m the ve rbal rep o rts p rovide e xamples of the diffe rent
ki nds of note-taki n g strategies. The fo llowi ng re port contai ns exam ples of both
note-taki ng and u nd e rli n i ng :
7J D
The next e xtract co ntai ns instances of noti ng d own and writi ng out:
4J I
1 2 FD
E labo rati on can b e descri bed as the mental p rocess of relating n ew knowledge
to exi sting i nfo rmation already held in lo ng-te rm memory. The two furt h e r types
of e l abo ration (pe rsonal and creative) ide ntified in Table 5.3 appeared re latively
1 54
infrequ e ntly i n the reports. Personal e l aboration usually i nvo lved the reaction of
learners to mate rial i n te rms of their perso nal i nterest :
1 OFD
The o n e i n stance of creative e labo ration i nvo lved co nstructing a sto ry line as a
means of re membering the g rammatical behaviou r of particular words:
1 FD
6FD
4J D
Auditory represe ntation of wo rds or se nte nces was u sed as a means to memorise
or p ractise TL forms :
1 1 JD
1 55
3FI
2FC
The strateg ies discussed i n this section were reported re lative ly infreque ntly by
lang uag e learne rs , apart from the differe nt ki nds of note-taki ng strategies and the
read aloud strategy. The i nfluence of mode of study and the TL on CSU reports
wi l l be discussed i n sectio n s 5.4 and 5 . 5.
1 56
5 . 3 . 3 S ocial Strategies
Description
Other-rei nforcement
An e xample of h ow a learn e r u ses this social strategy to reinfo rce her learning
e ndeavou rs i s :
9FD
Self- motivation was a new affective strategy ide ntifi ed in the ve rbal reports. This
p roved to be the m ost frequently mentioned affective strategy. Se lf-m otivation can
'be d iffe re nti ated from the other affective strategies defi ned in sectio n 3 . 3 . 1
because it i nvolves a n e xplicit focus on the means learn ers u s e t o keep
t h e m se lves g o i n g , othe r than the arranging of rewards after an activity h as been
com pleted (self- rei nfo rce m e nt ) , beyo nd sayi n g positive t hings to o neself (se lf
e ncourag e m e nt) or apart fro m using tech niques i n o rd e r to fee l co mpete nt to
carry out a task (self-talk) . Th e self-motivatio n strategy is defi ned in Table 5 . 5 .
Table 5.5
Affective Strategy Defi nition
Affective Strategy
Description
Self-motivation
1 57
8JD
9JD
1 58
(self-monito ri n g , co m p re h e nsion monito ri n g , strategy evaluation) which se rve to
mo nito r and eval uate the use of a cog nitive strategy (note-taki n g ) .
9 FD
The cog nitive strategy of t ranslati on was frequently co mbi ned with m etacog nitive
strategies such as co m pre hension m o n itori ng , di rected atte nti o n and se lf
manag e m e nt as i n :
1 FD
I translate the words I don 't kno w first. I find that if I do that
it frees me up mentally to concentrate on the actual
comprehension of the material and to think about answering
the questions. If I have to stop and look up a lot of words
I forget what it is that I am supposed to be doing.
Tran slatio n i s combined with co mpre h e nsion mo nitori ng i n this s h o rter example :
8FD
. The metacog nitive strategy of revi sion is co mbi ned with b etween-parts
e laboration i n :
6FD
Two m etacog nitive strategies, namely comprehe nsion monitori n g and problem
ide ntification are used with two cog nitive strateg ies, u nderlining and i nfere nci n g ,
in:
4J D
1 59
A fi nal examp l e of the co mbining of metacog nitive and cog n itive strategies is
g ive n below w h e re strateg y eval uation is carried out together with g rouping and
re hearsal strategies:
7FD
learners . T h i s meant that fo r final co m parison s, there were 28 repo rts by distance
learne rs , but o n ly n i n e by classroom learne rs. The reasons fo r t h i s discre pancy
i n size , and s o me of the possi ble d rawbacks resulti ng from it have al ready been
presented (secti on 3.7 and secti on 3 . 8) . H ad the classroom sam ple been larg e r,
that is closer t o the size of the distance sample, the effects of i nd ividual variability
wou ld pro bably h ave b e e n more si m i lar fo r the two popu lations.
P reliminary analysis revealed that sign ificantly more descriptio n s of strategy use
appeared i n t h e reports of distance learn e rs co mpared to classro o m learn e rs. The
ve rbal rep o rts o btai ned from classro o m learners co ntai ned on ave rage 1 0 . 2
instances o f strategy use whi le those o btai ned from distance learn e rs co ntai n ed
o n ave rage 2 6 . 6 instances of strategy u se. There was no appreci able diffe rence
1 60
i n t h e l ength of reports of either g roup. A b reakdown of t h e mean n u mber of
strategies reported by classroom and distance learne rs is d i splayed i n Fig u re 5. 1 .
Mea n S c o res for Repo rted Strategy U s e
Classroom a nd Dista nce Lea rn e rs
16
14
Cll
12
10
tu
E
..c
:::1
c
c
::::::.;
::%
8
6
4
2
mm
;88
58
;xx
Meto
Meto
1-- ClcS8room -I
>XX<
Cog
Socio
Aff
r Distance --1
Fig u re 5. 1
While the n u mber of reports of cog nitive strategies was lowe r for classroom
learners than for distance learners, within the popu latio n of classroom learn e rs
1 61
cog n itive strategies accounted for more than h alf (52. 1 7 perce nt) of t h ei r strategy
use, while fo r distance learners the fig u re was closer to o n e third (36 .64 pe rce nt).
The p roportion of rep o rted SSU ws also hig h e r for classroom learn e rs (5.43
perce nt) than fo r distance learners (2.42 perce nt) . These resu lts are displayed i n
Fig u re 5.2.
t> 40
e
4D
a..
20
r- Classroo m ---1
Fi g u re 5. 2
Figure 5.2 also s h ows t h at distance learne rs u sed proportio n ately more
m etacognitive strategies and affective strategies than classroom learne rs . More
than h alf (56 .51 pe rcent) of the reported strategies of distance learn e rs re lated
to MSU and 4 . 43 percent related to ASU . Fo r classroo m learn e rs t hese
p ropo rtions were 3 9 . 1 3 percent fo r MSU and 3 . 26 pe rcent fo r A S U .
1 62
To sum marise t h e n , classroom learners as a g roup used p ropo rtio n ately m o re
cog n itive strategies than metacog nitive strategies i n their l anguag e learn i ng . They
also used m o re social strategies than affective strategies. The reve rse situation
was fou nd fo r d i stance learners who made more frequ e nt use of m etacog nitive
strategies than cog nitive strategies. They also used more affective strategies than
social strategies. When the two popu lati ons are com pared i n te rms of the
freque ncy of strategy use, classroom learn e rs used on ave rag e fewer
m etacognitive , cog nitive and affective strategies than distance learn e rs.
The results of the com parison of the reported strategy use of classroom and
distance learne rs wi l l now be presented in more detai l in terms of each of the fo u r
categories o f strategy use.
MSU re po rts were fi rstly classifi ed acco rd i ng to the three categories established
by Brown et al . ( 1 983) fo r plan ni ng , m o nitori ng and evaluati ng learning.
Fo r both classroom and distance learners , the majo rity of m etacog nitive
strateg ies re lated to pla n n i ng activities. Evaluation strategi es accou nted fo r 26.67
percent of t h e MSU of distance learn ers and 1 9 .44 percent of that of classroom
learners . Mo nitori ng strategies were the least mentioned g roup. They acco u nted
fo r al most 20 percent of the metacog nitive strategies used by distance learn e rs ,
but fo r o n ly eig ht percent of the M S U re po rts of classroo m learne rs .
1 63
When the mean n u m be r of instances of p lan n i n g , monitoring and evaluating
strateg i es was co m pared for classroom and distance l earners, the diffe re nces in
the amount of MSU between the two popu lati o n s were very evident. The least
repo rted categ o ry of M S U fo r distance learners, t h at of m o nitori n g , m atched the
most frequently reported dimension of MSU fo r classroom learners namely ,
plan n i ng . This is rep resented i n Fig ure 5.3.
6
r%
QQ<
Plonning
1:%(
Monitoring
Evoluotion
Plonning
<S<
Monitoring
QQQ<
Evo luotion
Fig u re 5 .3
The n ext stage of t h e analysis was to co m pare t h e strategy use of c lassro o m and
distance learners i n terms of the use of i ndividual m etacog n itive strategies.
Results of the analyses of the frequency of use of i ndividual m etacog nitive
strateg i es appear i n Table 5.6.
1 64
Tabl e 5.6
Freque ncy of Metacog nitive Strategy Use
Classroom and Distance Learne rs
M etacognitive Strategies
C l assroom
Learners
(N=9)
0/o
PLAN N I N G
Advance Organisation
Organisational Planning
Time Lapse
Directed Attention
Selective Attentio n
Self-management
Analysing Needs
Revision
Delayed P roduction
Prioritising
Seek P ractice Opp.
Subtotal
Distance
Learners
( N=28)
0/o
Total
N
2
4
26
2.78
8. 33
1 1 .1 1
2.78
8. 33
5.56
0 . 00
1 3 .89
2 . 78
5.56
1 1 .1 1
72.22
22
48
20
16
27
61
2
14
3
5
7
225
5 . 24
1 1 .43
4.76
3.81
6 .43
1 4 .52
.48
3.33
.71
1 .1 9
1 .67
53.57
23
51
24
17
30
63
2
19
4
7
11
251
5 . 04
1 1 .1 8
5.26
3. 73
6.58
1 3 .82
.44
4. 1 7
.88
1 . 54
2.41
55 .05
2
0
0
0
0
0
0
1
3
5 . 56
0 . 00
0.00
0 . 00
0 . 00
0 . 00
0 . 00
2.78
8. 34
39
7
1
1
10
2
1
22
83
9.29
1 .67
.24
.24
2.38
.48
.24
5 . 24
1 9 .76
41
7
1
1
10
2
1
23
86
8.99
1 . 54
.22
.22
2.19
.44
.22
5 . 04
1 8 .86
EVALUATION
Production Evaluation
Performance Evaluation
Abil ity Evaluation
Strategy Evaluation
Lang. R epertoire Evaluation
Subtotal
1
0
1
4
1
7
2.78
0 . 00
2.78
1 1 .1 1
2.78
1 9 .44
7
6
27
60
12
112
1 . 67
1 .43
6.43
1 4 .29
2.86
26.67
8
6
28
64
13
119
1 . 75
1 .32
6. 1 4
1 4.04
2.85
26. 1 0
Total
1 00.00
420
1 00.00
45 6
1 00.00
MONITO R I NG
Comprehension Monitoring
Production Monitoring
Auditory Monitoring
Visual Monitoring
Strategy Monitoring
Plan Mon ito ring
, Double-check Monitoring
Proble m Identification
Subtotal
1
3
4
1
3
2
0
5
1 65
For distance learners the fou r most frequent strateg ies we re se lf- manage ment,
strategy evaluation , o rganisational plan n i ng and compre h e n sion mo nito ri n g . This
g roup of strategies d raws on the t h re e mai n categ o ries of MSU
( plan ning,
Self-man ag e ment and strategy evalu ation together accou nted fo r 2 8 . 8 1 pe rcent
of MSU rep o rts for distance learn e rs. The fol lowi ng exce rpt illust rates the use of
self-manag e m e nt :
1 OFD
Strategy evaluatio n i n the fol lowi ng e xample relates to the time lapse strategy :
4J D
1 66
tasks, plan ning the sections of assi g n m e nts, plan n i ng to use particu lar vocabu lary
o r p l a n n i ng particular sentences. An e xample of o rganisational plan ni ng i s :
1 FD
C o m prehe nsi o n mo nitori ng was also mentioned reg u larly i n the repo rts and was
t h e m ost frequently used mo nitori ng strategy by distance learn e rs. lt was used
by l earners in two circu mstances: fi rstly to check up on their co m p re h e nsion of
t h e study g uide mate rial and seco ndly was combined with t ranslatio n as a way
of checki ng comp rehe nsio n . E xam ples of this are :
8J D
5FD
T h e least used strategies were other mo nitori ng activities (audito ry and visual
,mo nito ri n g , plan monitori n g and dou ble-check monitori n g ) as we l l as two plan n i ng
strategies (an alysi ng needs and de layed p roduction).
1 67
the mean u s e of the revision strategy reported by distance learn e rs (.50 ) . A
descri ptio n of revision p roduced by a cl assroom learn e r is :
2JC
When the m o re freque ntly used strategies of classroom and distance learners are
co mpared , o n l y strategy evaluatio n appears to be used freque ntly by both g roups.
One explan ati o n fo r this cou ld be that the yoked subject tech nique in requiri n g
learners to re port o n the strateg ies t h e y used also e licited m o re accou nts o f
strategy eval uati o n .
A s with the q u estion nai re data a very w i d e gap e m e rged betwee n c lassroom and
distance learners in the use of the self- management strategy. Classroom learners
repo rted two instances of this strategy, wh ile distance learn e rs repo rted 6 1
instances.
There we re also simi lar fi ndings betwee n the question nai re data and the ve rbal
re po rt d ata fo r the advance o rganisation strategy. I n Tabl e 5 . 6 , advance
'o rganisation is among the least frequently used plan n i n g strategies fo r classroom
learners , but are among the most freque ntly used plan n i n g strateg i es fo r distance
learners . An example of the use of advance o rganisation i s :
4FD
1 68
wh ich i nvo lves devi sing a way of wo rki ng t h roug h t h e diffe re nt section s of the
wo rkbook.
6FD
Summary
Table 5.7 prese nts the range and freque ncy of cog nitive strategies used by
classroom and distance learners . The strategies are listed i n order of freque ncy.
1 69
Table 5.7
F reque ncy of Cog nitive Strategy Use
Classroom and Distance Learne rs
0/o
N
Note-taking
Note-taking
Writing Out
Listing
Noting Down
Highlig hting
U nderlining
S ubtotal
Resou rcing
Repetitio n
Elaboration
B etween-Parts
Personal
I magery
World
C reative
Subtotal
Translation
R ead Aloud
Transfer
Co ntextualisatio n
Summarisation
Auditory Representation
Work i n TL
G rouping
l nferencing
M e mo risation
Rehearsal
R ecombination
Deduction
Substitution
Total
Distance
Learners
(N=28)
Classroom
Learners
(N=9)
Total
%
4
4
1
1
0
0
10
7
8
8. 33
8 .33
2.08
2.08
0 . 00
0 . 00
20.82
1 4 .85
1 6 .67
16
13
8
6
5
2
40
38
35
5.86
4.76
2.93
2.20
1 .83
0 . 73
1 8 .31
1 3 .92
1 2 .82
20
17
9
7
5
2
50
45
43
6 .23
5 .30
2.80
2. 1 8
1 .56
.62
1 8 .69
1 4 .02
1 3 .40
0
0
0
0
0
0
8
4
0
2
1
1
2
1
0
2
0
1
1
0
0 . 00
0.00
0.00
0.00
0.00
0.00
1 6 .67
8 .33
0 . 00
4. 1 7
2.08
2 . 08
4. 1 7
2 . 08
0 . 00
4. 1 7
0 . 00
2.08
2.08
0 . 00
29
7
5
5
1
47
27
18
12
9
6
6
4
4
5
3
4
2
2
1 0 .62
2.56
1 .83
1 .83
.37
1 7 .21
9.89
1 6 .59
4.40
3.30
2.20
2.20
1 .47
1 .47
1 .83
1 .1 0
1 . 47
0.73
0.37
0.73
29
7
5
5
1
47
35
22
12
11
7
7
6
5
5
5
4
3
2
2
9 . 03
2.18
1 .56
1 .56
.31
1 4 .64
1 0 .90
6.85
3 . 74
3 .43
2.18
2.18
1 .87
1 .56
1 .56
1 .56
1 . 25
0.93
0 . 62
0 . 62
48
1 00.00
273
1 00.00
321
1 00.00
1 70
The i nformatio n p rese nted in Table 5.7 reveals that class room and distance
learners we re co m parable i n their patterns of CS U . Among the m ost frequent
cog n itive strategies fo r both g roups were note-taki ng , resou rci n g , repetiti o n ,
t ranslation a n d reading aloud. W h e n the n ote-taki n g categ o ry was diffe re ntiated
fu rthe r, the most f requent strateg ies were note-taki ng and writi ng o ut regardless
of mode of study. Examples of t hese strategies were p resented i n section 5.3.2.
Resourci ng i nvolves refe rri ng to sources of information about the TL. These
sources can i nclude dictionaries, text books and prior wo rk. An e xample of
resou rci ng i s :
6JD
1 3FD
1 71
Thus, the m ost freque ntly used cog n itive strategies for classroo m and distance
learne rs we re strateg ies which did n ot i nvolve active i nvolve m e nt with the
learni ng material. There is one notabl e exce pti on to this t rend which also se rves
to diffe rentiate classroom l earn e rs fro m distance learne rs - the use of e labo rati on
strategies.
E labo rati o n strateg ies e ntai l a high level of active i nvolve ment with the mate rials ,
particu larly i n te rms of maki ng associations between the material and prior
knowledge, or betwee n diffe rent sections of the materials. Distance learners
made frequ e nt use of e labo rati on strategies, which together fo rmed the seco nd
most frequent g ro u p of strategies after note-taki ng. Within the e labo rati on
category, betwee n-parts elaboration was the most frequ e nt with 29 reported
instances of use. Classroom learn e rs did not report any instances of the use of
betwee n- parts elaboration. Betwee n-parts elabo rati on involves the learner re lating
3FD
7FD
The lowe r freque ncy strategies presented i n Table 5.7 were those which requi red
the learne r to e ng ag e actively with the TL materi als - strategies such as
deducti o n , substitutio n , i nfere nci n g and rehearsal . O n ly one categ o ry of CSU
1 72
i nvolvi ng active i nvolve m e nt with the T L was used frequently by learners . This
was the application of e laboration strategies to TL mate rials by distance learn e rs.
Social Strategies
Distance
Learners
(N=28)
Classroo m
Learners
(N=9)
/o
/o
Total
N
/o
Questioning
Co-operation
Othe r- Reinforce ment
4
1
0
80.00
20.00
0.00
11
3
4
61 . 1 1
1 6 . 67
22.22
15
4
4
65 .22
1 7 .39
1 7 .39
Total
1 00.00
18
1 00.00
23
1 00 .00
'Distance learners made p roportio nately less use of questio n i n g (6 1 . 1 perce nt)
and co-ope ration strategies (1 6.67 percent) than classroom learn e rs (80 .00
perce nt and 20.00 percent) . This fi nding is co nsistent with the resu lts from the
questio n n ai re study, in wh ich a large p ropo rtion of distance learners noted that
they had no opportu nity to use these strateg ies.
Howeve r, the resu lts also indicated that distance learn ers do ask for assistance
and support by contacti ng fe l low learners or by aski n g friends and family
members to p rovide e ncourag e me nt (other-rei nfo rce ment) . This strategy
appeare d to be stro n g ly associated with the desire to com plete the cou rse and
not to withdraw. Examples of othe r- re i n force ment we re p rovided i n secti o n 5.3.3.
Oth e r- rei nfo rcem e nt has been classified as a social strategy si nce it involves
1 73
co ntact with others , howeve r it clearly also h as a strong affective co mponent.
Classroom learn e rs did not report any i n stances of the use of the other
rei n fo rcement strategy.
F reque ncy data fo r the use of affective strategies by classroom and distance
learners are prese nted in Table 5.9.
Table 5.9
Freque ncy of Affective Strategy Use
Classroom and Distance Learn ers
Affective Strategies
Distance
Learners
( N=28)
Classroom
Learners
( N=9)
/o
;o
Total
N
/o
Self-motivatio n
Self-encou ragement
Self-reinforce ment
Self-talk
3
0
0
0
1 0 0 .00
0.00
0.00
0.00
16
11
5
1
48 .48
33 .33
1 5. 1 5
3.03
19
11
5
1
52 .78
30 .56
1 3 .89
2 .78
Total
1 0 0.00
33
1 0 0.00
36
1 0 0 . 00
Perusal of t h e fig u res i n Table 5.9 reveals that distance learn e rs use a wider
range of affective strategies than class room learn e rs, and that they also make
m o re frequent use of affective strateg ies than t h e i r classroom cou nte rparts . The
m ost frequent st rategy used by learners i rrespective of mode of study was se lf
m otivatio n , e xa mples of which were given in section 5.3.4. Se lf-e nco u ragement
was repo rted re lative ly frequently by distance learners and often co ntai ned some
e le ments simi lar to the self-management st rategy. An example of this strategy i s :
6 FD
Self- reinfo rce m e nt, i nvo lving arranging awards fo r o neself, and se lf-talk, involvi ng
ways of reducing anxiety to make o n e self fe el co m petent, w e re relatively
infrequent :
1 74
1 1 JD
?J D
The mai n fi ndings fro m the study co ncerning the influence of mode o n st rategy
use i n the verbal report data are reviewed i n section 5.6. The next section
prese nts findi ngs re lated to the influe nce of the TL on strategy use.
Learn e rs of French and Japanese were co mparable i n their patterns of use of the
fou r mai n strategy categories: metacognitive , cog nitive , social and affective. As
re prese nted i n Fig u re 5.4, the most frequently reported category of strategy use
fo r both g roups co ncerned metacog niti o n .
1 75
A Co m pa riso n of Cate g o ri e s of Strategy U s e
Fren ch a n d J a pa nese Lea rn e rs
60
50
'"
40
...
0
; 30
..
1:
0
L
:. 20
10
0
Mete
Cog
Socio Aff
French ---1
Fig u re
Mete Cog
Socio
Aff
r Japanese --1
5.4
For learn e rs of Fre nch MSU re ports accou nted fo r 55.74 perce nt of strategy use
and CSU reports accou nted fo r 36 . 8 1 perce nt. These figu res are close to those
for learn e rs of J apanese whose MSU re po rts accounted fo r 53.01 percent of
strategy u s e , and CSU reports acco u nted for 40.44 pe rcent. Both g roups of
learne rs re ported usi n g more affective strategies than social strategies.
1 76
Mea n S c o re s fo r R e p o rted Stra tegy U s e
French a n d J a pa nese Lea rn e rs
14
12
10
..
0
u
"
;a
4
2
0
1--- French
1--
Japanese
Nf
---1
Fig u re 5.5
Howeve r this d iffe re nce is re latively small co mpared to the dive rg e nce between
classroom and distance learners o n the mean sco res for MSU reports as shown
in Figu re 5 . 1 .
P relimi nary an alysis of the i nflue nce of the TL o n MSU reports was carri ed out
in terms of t h e th ree d i me nsio ns of metacog nitio n : plan n i ng , m o n itori ng and
evaluati o n . Fig u re 5.6 p resents a comparison of learners of French and Japanese
in terms of the i r use , on ave rage, of t h e se thre e dimensions.
1 77
Q/5
2
:x5<
0
Q/5
<SQ25
:&
Plannlng
t.1onltorfng
Evoluotlon
'9<X
Plonnlng
French ---;
Monltorlng
:88 :&
Evoluatlon
Fig u re 5 .6
'Learners of Fre nch reported , o n ave rag e , two more instances of plan ni ng
strategies and one more i n stance of evaluation strateg ies than learners of
Japanese. Learn e rs of French and Japanese were si m i lar in their freque ncy of
use of monitori n g . A com pari s o n of Figure 5 . 6 and Figure 5.3 reveals that the
i m pact of mode of study on the generic categ o ries of M S U , namely plan n i ng ,
mo nito ri ng and evaluatio n , was much g reater than the i m p act of t he TL.
F u rthe r co mparison of learn e rs of Fre nch and Japanese was carried out based
on thei r use of i ndividual m etacog nitive strategies. Frequency data are prese nted
in Table 5 . 1 0 .
1 78
Table 5 . 1 0
F requency of Metacognitive Strategy Use
Learners of F rench and Japanese
M etacognitive Strategies
French
Learners
(N=1 9)
N
PLAN N I N G
Advance O rganisation
Organisational P lanning
Time Lapse
Directed Attention
Selective Attention
Self-management
Analysi ng Needs
Revision
Delayed P roduction
Prioritising
Seek P ractice Opp.
Subtotal
MONITO R I NG
Comprehension Monitoring
Production Monitoring
Auditory M on itoring
Visual Monitoring
Strategy M o nitoring
, Plan Monito ring
Double-check Monitoring
Proble m Identification
Subtotal
EVALUAT I ON
Production Evaluation
Performance Evaluation
Ability Evaluation
Strategy Evalu ation
Lang. Repertoire Evaluation
Subtotal
Total
0/o
Japanese
Learners
(N=1 8)
%
Total
N
0/o
17
29
10
8
18
40
2
8
4
4
8
1 48
6.49
1 1 .07
3.82
3 . 05
6.87
1 5 .27
.76
3 . 05
1 . 53
1 . 53
3 . 05
56 .49
6
22
14
9
12
23
0
11
0
3
3
1 03
3.09
1 1 .34
7.22
4.64
6. 1 9
1 1 .86
.00
5.67
.00
1 .55
1 .55
53.09
23
51
24
17
30
63
2
19
4
7
11
25 1
5 . 04
1 1 .1 8
5 . 26
3 . 73
6.58
1 3 .82
.44
4. 1 7
.88
1 .54
2.41
55 .05
19
2
0
1
6
2
0
11
41
7.25
.76
0 . 00
.38
2.29
.76
0 . 00
4.20
1 5 .65
22
5
1
0
4
0
1
12
45
1 1 .34
2.58
.52
.00
2.06
.00
.52
6. 1 9
23.20
41
7
1
1
10
2
1
23
86
8. 99
1 . 54
.22
.22
2.19
.44
.22
5 . 04
1 8 .86
5
3
24
32
9
73
1 .9 1
1 .15
9.1 6
1 2 .21
3.44
27.86
3
3
4
32
4
46
1 .55
1 .55
2.06
1 6 .49
2.06
23.71
8
6
28
64
13
119
1 .75
1 .32
6. 1 4
1 4 .04
2.85
26 . 1 0
262
1 00.00
1 94 1 00.00
456
1 00 .00
( Due to rou nding the percent total does not always equal 1 00.)
1 79
Analysis of the i nformatio n i n Table 5 . 1 0 reveals that learners of F re nch and
Japanese were very s i m i l ar in thei r use of i ndividual m etacog nitive strategies. Fo r
both g roups the most f requent metacog nitive strategies were two plan n i ng
strategies and one evaluation strategy : se lf- m anag e m e nt , o rgan i sational pla n n i n g
a n d strategy evaluatio n .
Both
g roups also
re po rted
high
use
of t h e
The most notable differe n ce in MSU betwe e n the TL groups re lated to the ability
evaluati o n strategy. Fo r learners of French, abi l ity evaluati on is the fou rt h most
frequ e nt metacog n itive strategy (totalli ng 2 4 instances of strategy use) , but for
learners of Japanese this strategy is hard ly used at all (am o u nting to fo u r
reported i nstances o f st rategy use) . Abi lity evaluat i o n takes place a s the learn e r
judges his/he r abi lity to perfo rm a task. I n stances such a s t h e fol lowing we re
frequ e ntly reported by learners of French:
BFD
1 1 FD
For both g roups the least freque ntly used strategies were one plan ni ng strategy
(analysi n g needs) and several monitori ng strateg ies (audito ry m o nito ri ng , vi sual
monitori n g , plan m o n itori n g , dou ble-check monitori ng ) .
1 80
5.5.2 The I nfluence of the Target Lan g uage o n CSU Reports
Table 5. 1 1 presents f reque ncy data on the use of cog n itive strateg i es by learners
of Fre nch and J apanese.
Table 5 . 1 1
Frequency of Cognitive St rategy Use
Learners of French and Japanese
French
Learners
(N=1 9)
'
Note-taking
Note-taking
Writing Out
Listing
Noting Down
Highlighting
U nderlining
Subtotal
Resou rcing
Repetition
Elaboration
Between-Parts
Personal
I magery
World
C reative
Subtotal
Translation
Read Aloud
Transfer
Contextualisation
Summarisation
Auditory Represe ntatio n
Work i n TL
G rouping
l nferencing
M e morisation
Rehearsal
Recombi nation
Deduction
Substitution
Total
J apanese
Learners
(N=1 8)
0/o
Total
o/o
12
2
6
6
5
1
32
30
15
6. 94
1 .1 6
3 .47
3 .47
2 . 89
. 58
1 8 .51
1 7.34
8 . 67
8
15
3
1
0
1
28
15
28
5 .41
1 0. 1 4
2. 03
.68
0 . 00
. 68
1 8 .94
1 0.14
1 8 .92
20
17
9
7
5
2
50
45
43
6 . 23
5 . 30
2 . 80
2.18
1 .56
. 62
1 8 .69
1 4 .02
1 3 .40
24
7
3
2
1
37
12
10
6
6
3
5
5
4
2
0
4
1
0
1
1 3 .87
4.05
1 .73
1 .1 6
.58
2 1 .39
6.94
5 . 78
3.47
3 .47
1 .73
2.89
2.89
2.31
1 .1 6
0 . 00
2.31
.58
0.00
.58
5
0
2
3
0
8
23
12
6
5
4
2
1
1
3
5
0
2
2
1
3 .38
0.00
1 .35
2.03
0.00
6. 76
1 5 .54
8.1 1
4.05
3 .38
2 .70
1 .35
. 68
. 68
2.03
3. 38
0.00
1 .35
1 .35
. 68
29
7
5
5
1
47
35
22
12
11
7
7
6
5
5
5
4
3
2
2
9 . 03
2.18
1 .56
1 .56
.31
1 4 .64
1 0 .90
6.85
3 .74
3.43
2.18
2.18
1 .87
1 . 56
1 . 56
1 .56
1 . 25
0 . 93
0 . 62
0 . 62
1 73
1 00.00
1 48
1 00.00
32 1
1 0 0.00
1 81
Initial pe rusal of the fig u res in Table 5 . 1 1 m ay suggest that learne rs of French
and J apanese were co mparable in terms of frequent use of note-taki ng ,
resou rci n g and repetition. Howeve r, a more detai led study of t h e fig u res b ri ng s
to light diffe re nces i n t h e patterns of use o f these strategies. Fo r note-taki ng ,
learne rs of Japanese reported frequ e nt u se of the writi ng out strategy (1 5
instances) , while for l earners of Fre nch this strategy was ve ry i nfrequ e nt (two
i nstances). Writi ng out is a mechan ical procedure involvi ng re peated copyi ng i n
order to m e m o ri se material . T h e e xamples presented below show that for
learne rs of J apanese writi ng out was almost always li nked to the learni n g of kanji :
1 JD
3J D
9J D
I write out the kanji several times just to get the flow o f the
strokes. Learning the kanji is the most mind-numbing part
o f it all. Writing out the kanji helps me to concentrate on
learning them.
The m ost marked differe nces betwee n learn ers of Fre nch and J apanese relate
to the use of e labo ration and translatio n . Ove ral l , learners of F re n ch repo rted 37
i nstances of elaborat i o n , while learne rs of Japanese reported o nly eight such
instances. E l aborati o n , as discussed i n sectio n 5.3.2, takes place as the learne rs
make meani ngfu l con n ections betwee n new i nfo rmation and old i nfo rmatio n , o r
1 82
betwee n i nfo rmatio n i n differe nt parts of a text . I nstances of betwee n-parts
e laboration and pe rsonal e labo ration appeared 3 1 times fo r learners of Fre nch,
but o n ly five times for learne rs of Japanese. Thus, i n atte mpti ng to maste r the
F re nch languag e , learne rs made more active , mean i ngful associations with and
wit h i n the TL m ate rial than did their Japanese co u nte rparts .
Summary
To s u m marise , the n , more diffe re nces emerged betwe e n learn ers of French and
Japanese for CSU than fo r M S U . The most freque nt strategies used by learners
of F re nch and Japanese were note-taki n g , resou rci ng and repetitio n , though the
way i n_ which and the extent to which these strateg ies we re used varied accordi ng
to the TL g ro u p . Learners of Japanese made g reater use of the writing out and
re petition strategies than did learners of French . Ve ry clear diffe re nces e m e rged
between the TL g roups when the use of e laboration and t ran slation was
co nsidered. Learn e rs of Fre nch made frequent associations betwee n new and old
i nfo rmation and between diffe rent parts of the learning materials. Co nversely,
learners of Japan ese made hardly any use at all of e labo ration strategies.
I n stead , fo r learn e rs of Japanese atte mpts to i m p rove TL co mpetence were
c h a racte rised by the frequent use of translatio n .
1 83
5 . 5 . 3 The I nfluence of the TL on SSU Repo rts
Learners of Fre nc h and Japanese reported an a l most equal number of inst ances
of SSU , as displayed in Table 5. 1 2.
Table 5 . 1 2
Frequency of Social Strategy U se
Learners of French and Japanese
French
Learners
(N= 1 9)
Social Strategies
N
Questioning
Co-operation
Other-R einforce ment
Total
J apanese
Learners
(N= 1 8)
N
/o
Total
/o
5
3
4
41 .67
25.00
33 .33
10
1
0
90.91
9 . 09
.00
15
4
4
65.22
1 7 .39
1 7 .39
12
1 0 0.00
11
1 00 . 00
23
1 0 0 .00
1 84
Learners of Fre nch made more frequent u se of affective strateg ies t h an learners
of Japanese, and also used a g reate r range of affective strategi es.
Table 5 . 1 3
Frequency of Affective Strategy Use
Learners of French and Japanese
French
Learners
(N= 1 9)
Affective Strategies
N
/o
J apanese
Learners
(N=1 8)
/o
Total
/o
Self-motivation
Self-e ncou rageme nt
Self-reinforcement
Self-talk
10
8
4
43 .48
34.78
1 7 .39
4.35
9
3
1
0
69 .23
23 .08
7 .69
0.00
19
11
5
5 2 .78
3 0 .56
1 3 .89
2 .78
Total
23
1 0 0 . 00
13
1 00.00
36
1 00 .00
The fig u res i n Table 5 . 1 3 reveal that both g roups made al most equal u se of se lf
m otivati o n , but l earn e rs of Fre nch made g reater u se of se lf-e nco u rag e m e nt, se lf
rei nforce m e nt and se lf-talk than learn e rs of Japanese. The fo llowi n g excerpts
show how learn e rs of Fre nch used the strategies of se lf-e ncourag e m e nt and se lfre i nforce m e nt to manage their feelings about the learn i ng process.
8FD
1 1 FD
1 85
5.6 S U M MARY
The yoked s u bject tech nique proved to be a p roductive method for e liciti ng
repo rts of strategy use from u nderg raduate fo reign lang uage learners . Seve ral
n ew strateg i e s were ide ntifi ed which had not appeared in the strategy use
lite ratu re .
These
were
time
lapse
and
seeking
p ractice
o pportunities
As mentioned i n section 3.3.2, the verbal re port study was not as co m prehensive
i n scope as t h e question nai re study . T h ro u g h the yoked subject p rocedure two
variables were exami ned, namely mode of study and the TL , fo r t h e i r re lationshi p
to strategy choice.
I n the verbal report study variance i n MSU was more stro ng ly associated with
mode of study than with the TL. Distance l earners made more freq u e nt use of
m etacog nitive strategies and employed a wider range of m etacog n itive strategies
than classro o m l earn e rs. Furth ermo re , distance learners used all t h ree aspects
Of metacog n ition (planni n g , mo nito ri n g , evaluation) to ove rsee t h e i r learning
e ndeavo u rs while classroo m learners confi n ed their metacog nitive co ntrol mostly
to plan ning activities. Distance learne rs m ade g reater use of s e lf-m anage ment
and advance o rganisatio n than classroom learn e rs.
The TL was not fou nd to play an obvious ro le i n metacog n itive strategy choice
by lang uage learners in diffe rent TL g ro u ps . Learne rs of French and Japanese
we re co m parable in their MSU , thoug h learners of French used slig htly more
planning and evaluation strategies than learners of Japanese.
The mai n i nfluen ce on CSU was very clearly the TL rat h e r than m o d e of study.
Learners of J apanese re lied on strateg ies which pertai ned to the more
m echanical aspects of lang uage learni ng , i n particular repetiti o n , t ranslation and
1 86
writi ng out. A major characte ri stic of the CSU of learn e rs of Fre nch was thei r
attempts to make active , meani ngful associations with and withi n the TL m aterial,
in particular throu g h the frequent use of e labo rati o n strateg ies.
Mode of study was not found to exert a marked i nflue nce on C S U . Classroom
and distance learne rs made reg u lar use of note-taki n g , re petition and resou rci ng.
One notable exception to this comparable patt e rn was the freq u e nt use of
elabo ration strateg i e s by distance learn e rs. This st rategy did not appear i n the
repo rts of classroo m learners.
There was a stro n g e r relationship betwee n mode of study and SSU than betwe en
the TL and SS U. Classroom learners m ade p ropo rtio nately g reate r use of the
social st rategies of q uestioning and co-operation t h an distance learn e rs , a fi nding
co nsistent with the q u estionnai re study. Reports of the use of other- rei nfo rcement
appeared i n the yoked subject data, a strategy which had not be e n m e ntioned in
the learni ng strategy lite ratu re. I n stances of the use of this strategy were co nfined
to distance learne rs . Learne rs of Fre nch and J apanese we re si mi lar i n the
frequency with which they used social strategies, t h oug h learners of Fre nch used
a wide r range of social strategies than learn e rs of Japanese.
Variance in AS U was re lated to both mode of study and the TL. Distance learners
made wide r and m o re frequent use of affective strategies t han classroom
learners . Learne rs of F re nch reported more i n stances of ASU than learn e rs of
Japanese.
A discussion of the resu lts fro m the question nai re study and the verbal report
study are prese nted i n the fo l lowing chapte r. A synthesis of the mai n fi ndings
fro m the cu rre nt study i s prese nted in C hapte r 8 .
1 87
6. DISCUSSION
The early part of this chapte r exami nes fi ndings co nce rni ng the identifi cation and
classification of strategies t h roug h the m etacog nitive, cog nitive , social and
affective model of strategy use (sectio n 6 . 1 ). The remai nder of the chapte r is
devoted to a d i scussion of the resu lts of the study o rganised i n six sectio n s , each
of which relates to one of the research questions p rese nted i n secti o n 3 . 3 . 2 . The
final section of t h e chapter (section 6 . 8) d raws together the various strands of the
fi ndi ngs to form a co herent picture of factors wh ich cont ri bute to the ki nds of
strategies learne rs deploy to deve lop foreign lang uage ski lls.
The strategy classification sch eme used in the cu rrent study was based on a
disti nction betwee n metacog nitive, cognitive and socio-affective strategies and is
s i m i lar to oth e r learning strategy models in the fie lds of educational psychology
and cog n itive psychology. lt proved to be an effective means thro u g h wh ich to
e xamine the strateg ies used by u nderg raduate fo reig n lang uag e learn e rs . I n
particular, t h e m odel was valuable i n servi ng to clarify how diffe re nt categories
of strategy use ( e .g . , metacog nitive , social) are i nfluenced by particu lar l earner
c haracte ristics or aspects of the language learn i ng context. One mod ification to
t h e mode l , n a m e ly the separation of the socio-affective g rouping i nto two distinct
categories, was i ntroduced to reflect the fact that SSU and ASU re late to quite
diffe re nt u n d e rlyi ng co nstructs.
1 88
The time lapse strategy i nvo lved the spaci ng of learni n g sessio ns i n o rder to
faci litate the co mprehe nsio n , memorisation o r p roduction of the TL. The use of
time lapse was re lated to three ci rcumstances : when e ncou nteri n g difficu lties,
when l earning TL fo rms , and as a fo rm of me ntal d rafting of mate rial , particu larly
for extensive writing . The oth e r new metacog n itive strategy was seeki ng p ractice
opportunities which was used by learners to make arrangements fo r using the TL
beyond the opportu nities provi ded by the cou rse.
The analysis of the ve rbal reports also resu lted in the i nclusion of n ew instances
of social and affective strateg ies in the taxo n o my of strategy use. Other
rei nfo rce ment was identified as a social strategy and proved to be a characteristic
of distance learne rs (see section 6.3. 3). This strategy involved so liciting h e l p f rom
oth ers , not specifically in relation to language tasks, but to gai n t h e n ecessary
e ncouragement to persevere i n the face of difficulties. The ASU of di stance
learne rs was also characterised by a p reviously u nidentified affecti ve strategy,
namely se lf- motivatio n , i nvolving re minding o neself of the advantages of
co nti nui ng to learn t h e foreig n languag e (see section 6.3.4).
I n atte m pti ng to classify instances of strategy use in t his study, t h e re were two
mai n findi ngs which warrant fu rther discussio n . T hese co ncern questions of the
bou ndary betwee n m etacog nitive and cog nitive strateg ies, and the fact that
strategies are used i n particu lar combinations. The re mai nder of t h i s section is
1 89
devoted to a consideration of these i ssues.
In Chapte r 2, in a discussio n of the researc h i nto m etacog niti o n , it was noted that
atte mpts to delin eate metacognitive strategies f rom cog nitive strategies h ave met
with some difficu lties. O'Mal ley and Chamot ( 1 990 :99) poi nt to t h e fact that there
are 'diffe re nces i n o pinion about what co nstitutes a metacognitive o r cog nitive
strategy' and C o h e n (1 9 9 1 : 1 34) notes that 'th e re is sti l l confu sio n ove r where to
d raw the line betwee n metacog nitive and cog n itive strateg ies' . W h i le the diffuse
bou ndaries betwee n the two mai n strategy g roupings are acknowledged in the
lite ratu re , specific e xam ples of classificatio n d ifficulties rare ly appear, and the
natu re of such difficu lties tends not to be pu rsued. Two difficu lties which arose
in the cu rre nt study i n re lation to t h e m etacog n itive/cognitive disti nctio n will now
be considered.
A second example of the diffuse bou ndaries between metacog n itive and cog nitive
strategy use arose i n re latio n to o n e of t h e n ewly ide ntified strategies, namely
work i n the TL. This strategy was classified as a cog nitive strategy si nce it was
applied to specific tasks as a means of com pleting them to best effect. However,
i n the verbal report data this strategy appeared at times to be used as an
1 90
executive thi n ki ng ski l l rather than as an i nteg ral part of task pe rfo rmance , thus
blu rri ng the disti n ctio n between t h e fu nctio n s of cog nitive and m etacognitive
strategies.
These e xamples poi nt to the fact that what is m etacognitive can not always be
read i ly se parated fro m what is cog nitive. lt i s i m portant for researc h e rs usi ng the
m etacognitive , cog nitive, social and affective model of strategy use to u nde rstand
that such difficulties may arise especially w h e n attempting to classify particular
instances of strategy use which appear in ve rbal report data. lt wo u ld also be
useful if there was more e xtended discussion i n the lite ratu re of the ki nds of
p roblems which can a rise when atte m pting to mai ntai n a disti nction betwee n MSU
and C S U .
In a rece nt review of studies atte m pting to detai l the precise ways i n which use
is m ade of specific strategies O'Malley and C hamot (1 990) e m p h asise t h e n eed
to study fu rther t h e eo-occu rrence of particul ar strategies. Furthermore, inst ruction
in g e n e ral learn i n g st rategies now attem pts to i ncorporate the fact that l earners
use a series of strategies rather than a sing le o ne when eng aged i n a particu lar
lang uag e task (Pali n csar and Brown 1 984; G raham, Harris and S awyer 1 987 ;
Ogle 1 987). The fi ndi ngs of this study poi nt to t h e need to go beyond t h e tallyi ng
of the u se of i nd ividual strategies, to consi d e r in more detai l h ow particular
strategy co mbinati o n s are used i n the performance of diffe re nt languag e tasks.
1 91
The re mai nder of t h e chapte r i s devoted to a d i scussio n of the fi ndi ngs which are
exami ned and i nterpreted acco rdi ng to the six research questions o utl i ned i n
section 3.3.2.
Research Questio n 1
The u nderg raduate foreign languag e learne rs i n this study were characterised by
the f requent use of metacog nitive and cog n itive strategies and t h e re lative ly
i nfreque nt use of social and affective strategies. This patte rn was comparable to
that of other learn e rs whose strategy use has b e e n i nvestigated by means of the
m etacog nitive , cog nitive, socio-affective model (O'Malley et al. 1 985a, 1 985b ;
Chamot et al. 1 98 7 ; Chamot et al. 1 988a, 1 988b).
In t h e question nai re study the most frequ ently used metacog nitive strategies were
advance organisatio n , self-monito ri ng and se lf-evaluation. These strategies
together e nco mpass the thre e dimensions of m etacog nitive contro l of learning
namely , plan ning , m o nitori ng and evaluatio n . The fi ndings from the much smal ler
verbal re port study carried out with 200-level learners of French and Japan ese
p rovi ded a simi lar picture of the most frequent m etacognitive strategies with one
e xception . Self- manag ement emerged as the most freque ntly u sed strategy
ahead of the oth e r t h ree planning , monito ri n g and evaluation strateg ies. Two
possi ble explanatio n s for this can be p roposed.
Firstly , the frequent use of se lf-management was found i n both studies to be the
mai n characte ristic differenti ating distance learne rs from classroo m learn e rs in
terms of MSU. Thus the recu rrent reporti ng of se lf-manag e m e nt can be partly
attri buted to the hig her proportio n of distance learners i n t h e ve rbal report
sample. Secondly, the nature of the yoked subject procedu re in the verbal repo rt
1 92
study may well h ave i nfluenced learn ers to focus o n t h e techniques by which they
learn best, and the ways in which they manage their l earn i ng co nditi ons p rocesses which are di rectly related to the self-management category of strategy
use.
Plan n i ng was the most freque ntly u sed dimension of MSU i n the questionnai re
study and the ve rbal report study. This finding is also co nsistent with a n u mber
of oth e r studies carried out with h i g h school language stude nts (e.g . , Ch amot and
KUppe r 1 989). However, com pared to patte rns of strategy use in prior research ,
the monitori ng and evaluating d i m e nsions of strategy use we re we l l-deve loped
among t h e learn e rs in the cu rre nt study. This can p robably be attri buted to the
ag e and leve l of abi lity of learn e rs in this study as we ll as to the de mands of the
distance learning co ntext.
I n the qu estion nai re study and t h e ve rbal re port study resou rci ng , repetiti on ,
translation and either e labo ratio n o r transfe r we re among t h e most frequently
used cog n itive strategies. At the same time a nu mber of i m portant diffe rences
between the patterns of CSU o btai n ed t h rough the two studies emerged.
The ove rall pattern of reporting in the question nai re study i n vo lved the frequent
use of key strategies such as i nfe re ncing and substitutio n . In the verbal report
study , on the othe r hand , learners appeared to i nclude more mechanical
strategies such as note-taki n g and translatio n in the i r cog nitive strategy
repertoire . These diffe re nt trends m ay be re lated to the i nfluence of the data
co l lection i nstru me nts on the repo rts of strategy use .
Since the questio n nai re inst ru m e nt requi red learners to respond to questions
about particular instances of strategy use, it i s possible t h at subjects were
influenced to react in te rms of what t h ey thoug ht they should do, rather than what
1 93
they actually did, to learn the TL. The yoked su bject proced u re , o n the other
hand, was a p roductive task with no specific p rompts, and t h i s task may have
provided m o re accu rate revelations of the strategies learne rs actually used.
In spite of t h e se obse rved diffe re nces in the freque ncy of CSU reported in the
questio n n ai re study as opposed to the ve rbal report study, t h e re was overall
co nsiste ncy i n the frequent use of resourcing , repetition, elaboratio n , translation
and elaborati o n o r transfe r in the two studies.
The most f requ e ntly used affecti ve strategy was self-encou rage me nt, which is
used when learners make positive state me nts to themse lves in o rder to fee l
competent to carry o ut particular tasks. Other means of affective co ntro l were
used less frequently, and agai n , t h e o rganisational demands of other means of
affective co ntro l (such as arranging rewards fo r o neself) may have prohi bited thei r
use.
1 94
6.3 THE I N FLU E N C E OF MODE O F STU DY
Research Question 2
The findi ngs fro m t h is study revealed that mode of study stro n g ly i nflue nced the
MSU and the S S U of u nderg raduate foreign language learne rs .
The results from this study also suggest that o nce distance learn e rs have
deve loped ways of managing their learn i n g , their inte ract i o n s with the TL
m aterials , as described i n terms of their C S U , are not markedly diffe rent from
t hose of thei r classroom cou nte rparts. That is, i n the context of p rivate study , both
g roups of learn e rs use comparatively similar processes in wo rki n g with the TL
materials to d eve lop fo reign lang uage ski lls. Fou r exceptio ns to this comparable
p attern are exami ned in section 6 . 3.2.
1 95
For the social categories of strategy use, mode of study was found to i nte ract
with the o ppo rtu nities learn e rs had to deploy such strategies. While distance
learners had li mited opportunities to use t h e more conventional social strateg i es
of questio n i ng and co-operation , they exhi bited particu l ar resou rceful ness i n their
u se of a further dimension of SSU , name ly oth e r-reinfo rce ment (see secti on
6 .3.3).
Fi nal ly, in terms of ASU , b oth classroom and distance learn e rs made re latively
infrequent use of th ese strategies. Classroom learners de ployed social strategies
more ofte n than affective strateg i es. Distance l earners, on the other h and ,
resorted more to affective means to manage their learni ng than to tech niques
involvi ng social contact. Distance learn ers also re ported using a wider range of
affective strategies than classroom learne rs .
G ive n then that this is the gen eral pictu re f o r t h e use of the fo u r mai n strategy
u se categ o ries by classro o m and distance l earners , we can now co nside r which
i ndividual strategies are i nflue nced by m ode of study. Furthermore , we can
s pecu late as to the reasons why particu lar strategies are deployed diffe re ntly by
l earne rs in classroom and non-classroo m e nvi ro n m e nts.
I n both the question nai re study and the verbal report study it was evident that the
single most i mpo rtant strategy which s e rved t o diffe re ntiate classroom and
distance l earners was se lf-manage ment. Distance learners are clearly set apart
from classroo m learners i n the extent to which they m ake use of this strategy. An
esse ntial p re requisite for the use of self- manag e m e nt is for the i ndividual l earner
to know how s/he learns best. The use of self- management also requi res the
learner to have the necessary procedu ral skills to set up these opti mal learning
conditions. We can now reflect o n the reason s why variance in the use of se lf
management is so close ly associated with m ode of study.
1 96
i n that it co mprises t h e two d i m e nsions of m etacog nition outli ned i n Chapte r 2 :
knowledge of cog niti o n and control of cog n iti o n . The oth e r metacog n itive
strategies are concerned, fo r t h e m ost part, with co ntrol of cog niti o n . As such ,
they exercise t h e executive d i me n sion of m etacog nition rather than the self
knowledge dimension.
aspects of
metacog nition, it is a powe rful i ndex of the metacog nitive knowledge and ski l ls
of learners. Th us, w h e n one considers the stro ng relationship between mode of
study and metacog niti o n , it is not surp rising that se lf- management e me rg e s as
the p ri ncipal strategy to set classroom learners apart from distance learne rs .
The use of advance o rg anisatio n i nvolves previewi ng mate rial to obtai n a g e n e ral
idea of tasks , while o rganisational plann i ng invo lves generati ng a plan for the
execution of the tasks. In the verbal report study, the MSU of distance learners
was found to be c haracte ri sed by the frequent applicati on of these strategies.
Looki ng t h roug h m ateri al and plan n i ng how to wo rk with it appears to be more
i mportant for distance l earners than classroom learners.
1 97
8FD
/ look through the whole dossier first because I'm keen to see what
it is about and to get some idea of what I'll be expected to work
with, what I'm expected to master. I like having an idea of what is
coming up and also of where I am going. So then I decide what I
am going to study first, what next and so on. I decide what I will
leave until last too. That's usually the hardest part. Also any essays
I leave until last.
I n the question nai re study , the use of advance o rganisati on was fou nd to be
seco nd o n ly to se lf- management i n separating classroom and distance learn ers.
There was no questionnaire item relating to t h e use of org anisational plan n i ng ,
but, h ad such an ite m been included i n the study , it is expected that distance
learne rs wo u ld have been shown to exceed classroom learne rs i n t h e i r use of this
aspect of m etacognition. In sum mary , as far as plan n i ng strateg i es were
conce rned, classroom and distance learners were differe nti ated i n their use of
self- management , advance o rganisation and o rganisational planni n g .
l t was also found that di stance learners made much g reater u s e of t h e mo nito ri ng
and evaluation dimensions of metacog nition than classroom learne rs. I n terms of
monito ri ng , distance learners we re particu larly co ncerned with co m pre he nsion
monito ri ng , that is, c hecki ng up o n thei r understanding of the TL, and p roblem
ide ntificati o n , which e ntails ide ntifyi ng the aspects of a task which are p reve nti ng
its successfu l co mpletio n .
Firstly, di stance learn e rs are not able to reg u late the deg ree of co m pl exity of the
materi al p resented to them. They are not necessarily provided with mate rial at a
1 98
leve l appropriate to t h e i r own powers of u nderstanding and so it is possible that
they h ave to decide for the mselves wh ethe r the materi al is wit h i n t hei r
co m preh e nsion abi lities. Lang uage teachers , o n the other hand , i n face-to-face
classroom teach i n g , are generally se nsitive to t h e l evel of u nde rstanding of thei r
learners and make atte mpts to present m aterial which is with i n the l earners'
g rasp. Secondly, d i stance learners do not h ave a teache r to check u p o n thei r
co m preh e nsion t h roug h the regu lar posi ng of q u e stions. Distance learn e rs must
make efforts to m o n itor thei r u nderstandi n g fo r t h e mse lves. Classroom l earners ,
on the other han d , do not have the sam e n e ed t o ask the mse lves 'How a m I
going i n all of this?' and ' Have I got the rig ht idea?' .
I n the verbal re port study one of the m etacog n itive strategies most frequ e ntly
mentioned by classroom learne rs was revisio n , while in the qu estio n n ai re study
the re latively i nfreq u e nt use of revision by classroom learn e rs was fou nd to
co ntri bute sig nificantly to the separation of learners acco rding to mode of study.
This discrepancy can be related to effects fro m t h e co nte nt of the i n structions fo r
the ve rbal report p rocedure (see Appe ndix D ) .
In the questi o n n ai re study, classro o m learn e rs used revi sion sig nificantly less
often than classro o m learners. The reaso n s fo r this can be related to t h e fact
that , i n the cou rse of regu lar languag e sessi o n s , classroom learners are p robably
exposed to a wid e r range of previously l earnt forms than di stance learners.
1 99
Distance learne rs, o n the othe r h and, h ave to arrange opportu nities for
the mselves to review prior wo rk in order to g ai n the repeated exposure n ecessary
to retai n l ang u ag e forms.
Results fro m t h e question nai re study and the verbal re port study i ndicated that
t h e re we re some diffe re nces betwee n classroom and distance learners o n CSU
m e asures, t h ough these were re latively m i n o r compared with the very evident
differences o n MSU measures. Through the yoked subject p rocedu re classroom
and distance learners repo rted usi ng similar cog nitive strategies as they e ngaged
with the TL i rrespective of mode of study. However there were four exceptions
to this comparable pattern re lating to the use of e laboratio n , transfer, resourcing
and repetitio n .
Distance learn e rs reported maki ng frequent use of e labo ratio n whi le classroom
learners did not repo rt any instances of the use of this strategy. The use of
e laboration i nvo lves maki ng meani ngful co nnecti ons betwee n parts of the learn i ng
m aterials, o r re lating n ew info rm ation to prior knowledg e and to personal
experience. Transfe r strategies are similar to e laboration strategies in that
learners use previ ously acquired knowledg e of the language to faci litate the
completion of a task. O' Mal ley, Ch amot and Walke r ( 1 987) also note the close
con nectio n s between e laboratio n and transfe r, and co nside r that e laboration can
be co nsidered to be a superordinate catego ry for transfer. Distance learners were
characterised by a m uch g reate r use of e laboration and transfer than classroom
learne rs .
lt is clear that both elaboration and transfe r require the learne r to e ngag e actively
with the TL material and to process it at quite a deep level . E labo ration has long
been considered to be a sig nificant learn i n g strategy because of the be nefits for
compre h e nsi o n and retention which have been demonstrated to acco m pany its
u se (e.g. Reder 1 980 ; Wei nste i n and Maye r 1 986). One explanation for the
g reater use of e l aboration and transfer by distance learne rs is that they must
200
i nteg rate the course mate rials fo r the mselves si nce there is no teache r to situate
e ach learn i n g t ask in wide r co ntexts, o r to re late the material to what they already
know.
Resou rcing and repetition are diffe re nt in n ature from elabo ratio n and transfer.
Resourcing i nvolves usi ng alte rnative sou rces of informati on about the TL while
repetition i nvo lves repeati ng wo rds or p h rases i n the course of pe rfo rming a TL
task. Distance learners would appear to m ake g reater use of resourcing than
classroom learn e rs si nce they can not readi ly approach a teac h e r w h e n t h ey strike
d ifficu lties, and so must re ly on other sou rces fo r assistance. Repetiti o n is also
u sed more often by distance learne rs. The reaso ns for this a re not e nti re ly clear,
but one explanation could be simi lar to that provided for the i ncreased use of
revision by d i stance learners, namely that distance learners a re not exposed to
the ki nds of repetition of lang uage ite ms which occur in a classro o m e nviro n ment
and which a re usefu l fo r the rete nti o n of TL ite ms.
Distance learn e rs are very co nscious of the fact that they are p u rsuing their
studies i n re l ative iso lation co mpared to t h e i r on-campus cou nte rparts. Attempts
h ave been made to mini mise this isolatio n , as outlined i n section 3. 1 , and to
p rovide distance learne rs with opportu niti es fo r co ntact with tuto rs and fel low
learners. This b ackg rou nd is i m po rtant to u nderstand the s o mewhat conflicti ng
respo nses m ade by distance learn e rs in t h e questionnai re study and the ve rbal
repo rt study about their SSU .
I n the qu esti o n n ai re study the most frequent response by distance learners when
asked about t h e i r use of questio n i ng and co-ope ration strategies was that they
had 'no opport u nity' to use these strategies. Classroom learners u sed them
'someti mes' or ' rarely'. Howeve r, i nsights f rom the ve rbal report study revealed
that distance learners did make use of these social strateg ies and t h at they also
co ntacted oth e rs to elicit e ncou rag e m e nt to conti nue with the cou rse. This
strategy was te rmed 'other-rei nfo rce ment'. lt was classified as a social strategy
20 1
i n this study though it also has a stro ng affective compone nt. N o i n stances of the
use of other-rei nfo rcement we re fou nd in the ve rbal reports of classroom learners.
One reason for the lack of consiste ncy betwee n the two studies in reports of SSU
co u ld lie in the fact that di stance learne rs are very aware of the isolated co ntext
in which they study, and therefore , i n t h e closed response format of the
qu estion nai re, readily opted for the 'no opportu nity' catego ry . Howeve r, when
asked about the strategies they used in a more ope n-ended co ntext, their
resou rceful ness and atte mpts at netwo rki n g in order to ove rcom e t h e i r isolated
study context were very evide nt.
lt appears then that distance learners do app ly social strateg ies to support their
learni ng endeavou rs but that their opportu n ities are more li mited than those of
classroom learne rs . Howeve r, distance learn e rs are disti nctive i n their use of a
further dimension of SSU , namely othe r-re i n force ment.
I n the verbal report study distance learn e rs used a wider range of affective
strategies than classroom learne rs and also u sed them more freque ntly than their
classroom cou nte rparts. Distance learners u sed affective strategies to e ncou rag e
the mselves, to reduce anxiety , to reassu re t h e mselves they cou ld tackle the work
and to motivate themse lves (see sectio n s 5.4.4 and 5.5.4. for e xam p les). The
m eans classroo m learners used for affective manag ement of learn i n g were all
re lated to self- motivation.
202
strategies than affective strategies (see Fig u re 5. 2).
I n the questio n n ai re study both classro o m and distance learn e rs repo rted
i nfrequent use of affective strategies. The t hree ASU ite ms did n ot serve to reveal
the ki nds of diffe re nces which we re evident in the verbal re port study. lt is
possible that t h e q uestion nai re ite ms did n ot tap appropriate aspects of AS U . lt
is also possible that more carefu l ly worded questio n nai re ite ms d rawn from the
descri ptions in the verbal report study wou ld have e licited a diffe rent respo nse.
Research Questio n 3
Four sets of ci rcumstances we re ide ntifi ed in the question nai re study i n which the
g ap betwee n classroom and distance learners on MSU measu res became
particu larly p ronou nced . These ci rcu m stances were whe n learn e rs had no p rior
TL experi e nce, when the TL was J apanese , when the profici e n cy level of l earners
was 'B', and w h e n learners were at the 200- leve l of study.
Fo r learne rs who had no pri o r TL experi ence before e n ro l l i n g in the u nive rsity
203
langu ag e cou rse , the se lf-manag e ment strategy was of p ri m e i mportance in
setting learne rs apart according to m ode of study. lt appears that classroom
learners with no p ri o r TL experie nce had not developed the sam e u nderstandi ng
of their learn i ng p rocesses, or of ways of establishing these p rocesses, as thei r
distance cou nterparts. One explanati on for this is that classroo m l earners with no
pri o r TL expe rie nce a re able to re mai n depe ndent on the teache r to estab lish
their learni ng conditi ons, whi le distance learn e rs , with a similar backg rou n d , must
re ly o n thei r own e ndeavou rs if they are to p roceed. In such ci rcu m stances
distance learn e rs develop se lf-manag e m e nt strategies to manage their language
learni n g .
Fi rstly, with regard to the i nflue nce of the TL, it is co nsidered to be very difficu lt
to study Japanese at a distance. This is acknowledged both by lecturers and by
e n rolled students. I n order to keep u p with t h e m aterial it is n ecessary to spend
at least two hou rs d ai ly wo rki ng with the TL, according to cou rse guidelines. A
larg e nu mber of learners do not com plete the course because of t hese d e m ands.
So, if distance learners are to succeed they must be very o rg anised and must
plan their p rog ress carefu lly through the course. Advance o rganisation is a
p ri mary plan n i n g strategy.
At the 200-level of lang uage study learne rs are often immediately exposed to a
less co ntrolled lang uag e e nvi ron ment and a re expected to deve lop a much wider
204
range of ski lls in language forms and lang uag e use. The teac h e r is less likely to
nomi nate e xactly what has to be l earnt, and more responsi bility is placed on the
learn e r to master aspects of the T L. For the distance learner, the more open
ended e xposu re to the TL, which begins at the 200-leve l , also requi res carefu l
plan n i ng of the learni n g process.
'
8' ,
were m axi mally diffe rentiated on M S U measures. One explanation fo r this lies in
what we al ready know about metacog nitive co ntro l as one of the keys to
successfu l lang uag e learni ng. One would expect that hig h-achi evi ng learn ers, that
is 'A' learners, would have highly d eve loped metacog nitive abi lities i rrespective
of mode of study. lt also seems that at the p roficie ncy leve l of 'C' classroom and
distance lang uag e learners are relatively simi lar in the u se they make of
metacog nitive strategies. Howeve r, betwee n the highly successfu l learners and
the l ess successfu l learners there is a g roup who diffe r sig nificantly o n how they
manag e their learn i ng , acco rding to whether they are classroom learners or
distance learners. lt is possi ble that at this level of ability classroom learn ers sti l l
do n ot n eed to be self-di rected to make prog ress, w h e reas their di stance
cou nte rparts have to make a g re ater effo rt to di rect thei r l earning i n order to
succeed.
In these three sets of ci rcu mstances, the use of advance o rg ani satio n was the
mai n strategy associated with the separation of classroom and distance learners.
I n o rder to direct the i r learning , distance learn e rs fi nd it n ecessary to preview
material to see h ow it is organi sed and to dete rmi n e what is required of them
befo re t hey can p roceed. The advance o rganisati on strategy appears to be critical
fo r distance learn e rs to enable them to e nter i nto the appro p riate mental frame
to work o n particu lar language tasks. Classroom learners d o not appear to use
this strategy to the same exte nt in the ci rcu mstances outlined, and one can
assu me that they re ly on the d i rection given i n the classro o m for an ove rview of
learni n g tasks and fo r an u nderstanding of what is requ i red.
205
6.5 THE TARG ET LANGUAG E AN D STRATEGY USE
Research Questio n 4
What role does the TL play in the strategies learners choose to use
(metacognitive, cognitive, social, affective) to improve their TL competence ?
I n the previous section the co ntribution of o n e o f the TLs in this study , namely
Japanese , was discussed in relation to the varia nce of classroom and distance
learners on MSU measures. In this sectio n the re lationship betwee n particu lar
TLs and the ki nds of strategies reported by learners wi ll be co nsidered.
The T L was not fou nd to i nflue nce the freque n cy with which learners of different
TLs used the g e n e ric categories of strategy u s e : metacog nitive , cog nitive, social
and affective . Howeve r, the TL did exert an i nfluence on the ki nds of cog nitive
strategies learners used. Re latively m i nor diffe re nces i n metacog nitive , social and
affective strategy use we re also detected between learners of diffe re nt TLs. The
discussion of resu lts will be limited to the mai n area where the TL was found to
i mpact on strategy use, that is thro u g h the cog nitive strateg i es learn e rs use as
they i nte ract with the TL mate ri als. The clearest fi ndi ngs in this regard emerged
through the verbal re port study, and the exam i n ati on of the relationship between
the TL and CSU will make refere nce to these resu lts.
The three main categories of CSU fo r learn e rs of French and J apanese were
note-taki ng, resou rci ng and repetiti on. H oweve r the two TL g roups differed in the
extent to which repetition was u sed and in the ki nds of note-taki ng st rategies
which were deployed. An analysis of the sub-categories of the note-taki ng
strategies revealed that learners of J apane s e re lied heavily o n the w riting-out
strategy , that i s , mechanical copyi n g (see section 5.5.2 for e xamples) .
Deployment of the writi ng-out strategy was clearly li nked to atte m pts to learn
kanji , and as such was di rectly attri butable to a structural featu re of the TL.
Learners of Fre nch we re characteri sed by t h e use of elabo ratio n and atte mpts to
carry out tasks i n the TL without recou rse to English. Learners of J apanese , on
the other hand , relied heavi ly on translation as a means of developing
206
compete nce in the TL. Thus, l earne rs of Fre nch made active , m eani ngfu l
associ ations with and within the TL mate ri als while learners of J apanese
atte m pted to relate Japanese forms to English forms and vice-ve rsa. Learne rs of
Japanese repo rted twice as many i nstances of repetition as learners of Fre nch.
lt is i mportant to conside r whether these diffe re nces i n strategy use are also
attri butable to the d e m ands placed on learne rs by t he particular structure of the
TL, or whether there are other i nteNe ning factors such as cou rse methodology
and the level of advancement in the TL of learn e rs of Japanese compared to
learne rs of Fre nch.
lt is a particularly com p l ex task to atte mpt to characte rise the lang uage teaching
approach of a cou rse in terms of
One m ust co nsider a range of facto rs such as course materials, textbooks , the
nat u re of classro o m i nteractions and the beliefs of the lectu re r o r teacher. Such
a detai led analysis is beyo nd the scope of the cu rrent study. However, an
i nformal comparison of the methodology of the F rench and Japanese cou rses at
the 200- leve l was u ndertake n .
207
One can conclude that the diffe rences i n CSU betwee n learn e rs of French and
J apanese we re o n ly partly attributable to the i nfluence of the TL. The language
teach i ng m ethodo logies i mpl icit i n the materials also appeared to affect t h e ki nds
of strategies l earne rs used as they i nteracted with the TL in atte m pts to deve lop
their lang uag e ski lls.
A simi lar i nterpretation of the perceived diffe re nces i n strategy use betwe en
diffe re nt TL g ro u ps was made in an early study by Po litze r ( 1 9 83) . H e sugg ested
a co m plex i nte raction between language teachi ng methods and the diffe rent
learning be havi o u rs of students of French , German and S pa n i s h . S i m i larly in a
recent longitud i nal study of strategy use , carried out by C ha m ot and KOpper
( 1 989) , the i nfluence of course objectives on strategy use was n oted, even
though their study was confined to learners on one TL, S pa n i s h . Chamot and
KOpper ( 1 989 : 1 7 ) obse rved that 'student s i n classrooms e m ph asi sing g rammar
apparently p refe rred strategies such as deduction and t ranslatio n , whereas
students i n classrooms focusi ng on p roficie ncy te nded towards strategies such
as infere nci n g and substitution'.
208
Fre nch study , learne rs are re latively fluent in the ext e n sive use of languag e ski l ls.
Lectu re rs in Japanese o bserved that such flue n cy could not be expected at the
200-level of study. Fu rthermore , learners m aj ori ng i n both F rench and J apanese
indicated that thei r ski lls in J apanese were at a ve ry much less advanced level
than the level at which they cou ld pe rform in F re nch.
There is also the q uestion of the rate at which learners can be e xpected to
prog ress i n the learn ing of differe nt Tls. Fo r example, the US Foreign S e rvices
Institute charts suggest that learners of Chi nese and of Japanese requ i re twice
as much time as learners of French and G e rman to reach the same level of
mastery of languag e ski l ls. This, it is expected, would impact on the strategies
learners of Fre nch and Japanese were able t o use afte r an equal n u mber of
hours of learning the TL. Howeve r, it should also be noted that the oste nsibly
slower rate of prog ress for the learn i ng of J ap.anese remai ns open to question
and can not be separated from the effects of teachi ng approaches o r course
m ethodo logy (see , for example, Marriott and Yamada (1 991 ) )
209
6 . 6 T H E I N FLU E NCE OF THE LEAR N I NG CONTEXT
Research Question 5
Do aspects of the language learning context (namely level of study and language
use opportunities) affect the metacognitive and cognitive strategies employed by
undergraduate foreign language learners ?
No clear pattern was fou n d in the study fo r a re lati onship betwee n level of study
and strategy choice. The reasons for this probably re late to the fact that each
leve l of study does not represe nt a si mi lar level of advance m e nt across diffe rent
TL g roups. This poi nt was i l lustrated with regard to learn ers of French and
J apanese at the 200-level i n section 6.5. Both level of study and TL experience
are ways of measuri ng d u ration of study and a bette r indicato r of the effects of
the duratio n of study o n strategy choice proved to be pri o r co ntact with the TL
rather than cou rse leve l . P rior experie nce in learning the TL befo re e n rolling i n
t h e u nive rsity cou rses was fou nd t o have a g reate r i mpact o n cog nitive strategy
choice than the level of stu dy. Resu lts relati ng to the effect of p rior TL experi ence
on strategy choice are di scussed i n section 6 . 7 . 2.
An investigation of the i nte raction of language use o pportu niti es with strategy
choice was i ncluded i n this study to dete rmi n e whether functional p ractice
opportu niti es, beyo nd those provided by the cou rse, had an effect o n the
strategies learne rs chose to use wit h i n the cou rse. I n the questio n nai re sample
1 3.7% of learn e rs reported that they had some opportunities to p ractise with
native speake rs , though the frequ e ncy with which they took advantage of these
opportu nities was not determined. Languag e use opportu nities were not fou nd to
influence the strateg ies learners used eithe r i n plan n i ng thei r learning or whe n
e ng ag i n g with the T L materi als. A recent model o f second-lang uage acquisition
p roposed by Gard n e r and Macl ntyre ( 1 992) e mphasises the i nfluence of formal
and info rmal co ntexts on the student's level of ach ieve ment, which in tu rn is seen
to i nfluence strategy use. Fi ndings from this study revealed that learners who
21 0
h ave access t o i nformal learn i ng co ntexts used strategies i n fo rmal co ntexts
which we re si m i lar to those used by l earners whose learn i n g o p po rt nities were
limited to m o re formal e nvi ronments. More research on strategy use i n fo rmal and
info rmal learn i n g e nvi ronments is required.
Research Q u estion 6
Fi ndi ngs fro m this study provide evidence fo r the influence of age and p rior TL
experie nce o n strategy choice by fo reig n language learne rs . The re lati onship
betwee n age and MSU is discussed i n section 6 . 7.1 . The i nflue nce of prior TL
e xpe rie nce o n CSU is exam ined in sectio n 6 . 7 . 2 . Proficie ncy was fo und to have
some effect on MS U , but this was not ve ry marked. A discussi o n of th ese resu lts
appears i n section 6 . 7.3. Metacognitive and cog nitive strategy choice were
e xami n ed i n re lation to ge nder, motivatio n and language l earn i ng experie nce of
learne rs , but n o re lationship was found between these vari ables. A comparison
of these res u lts with findings fro m p revious studies is p resented i n section 6.7.4.
An o rdered co mparison of the strategy use of lang uage learn e rs whose .age
range cove red a period of five decades revealed that there was a p rog ressive
increase i n M S U from you nger to older ag e g roups i rrespective of mode of study.
In particu lar learners ove r thirty were found to diffe r sig nificantly f ro m learn ers
u nder t h i rty i n their frequency of use of particu lar metacog n itive strateg ies.
The increase i n the use of metacog n itio n with age and e xperi e n ce has been
demonst rated i n studies of the acquisitio n of fi rst languag e readi n g ski l ls by
21 1
schoo l-age children (e.g. Mye rs and Pari s 1 978; Canney and Wi n o g rad 1 979 ;
Kobasigawa, Ransom and Holland , 1 980 ; Forrest-Pressley and Waile r 1 984) . The
pattern of g re ater metacognitive contro l in learn i ng among m atu re- ag e stude nts
was also we l l e stablished a decade ago ( E ntwistle and Ramsde n 1 982; Watki ns
1 983; Watki ns and Hatti e 1 985), and the fi ndi n g s have been co n sistent over
several studies but h ave not bee n verified specifically i n re latio n to second
lang uag e learni ng .
A n u mbe r o f o bse rvations have been made about the use o f lang u age learni ng
strategies i n re lation to the age of learn e rs. For example, O'Mal ley et al.
( 1 985a:35) noted that high-school ESL stude nts we re able to make ' e xte nsive use
of m etacognitive strategi es' and they concluded that 'conside rable reflection on
the acquisition and fu nction of language was occurri ng'. Howeve r t h e i r study did
n ot i nclude a d i rect comparison of hig h -school students with you n g e r learne rs. In
reviewi ng E h rman and Oxford's (1 989) study, Oxford (1 989 :238) n otes that the
adu lt lang uag e learners in their study 'see med to use somewhat more
sophisticated lang uage learning strategies than did you n g e r learne rs in other
studies' . Ag ai n , these observations are i m pressionistic and are d rawn fro m loose
co m pariso ns m ade across diffe re nt studies. I n the cu rre nt study the investigation
of variance i n lang uag e learning strategy use due to age is based o n an o rdered
co m pariso n of the reported be haviours of learn e rs whose age range cove rs a
span of almost five decades.
Findings revealed that five metacog nitive strateg i es we re closely associated with
t h e i ncreased use of metacog nitio n acco rdi ng to matu rity of l earners: se lf
m anag e m e nt, advance o rganisatio n , de layed p roductio n , revision and self
m o nitori n g . These strategies tog ether e ncompassed the plan n i n g and monitori ng
di m e nsions of metacog nitio n . The metacognitive strategy most closely related to
t h e age of learners was se lf-manag e m e nt .
21 2
are avai lable. S e lf-manag e me nt thus requi res of the learn e r a g ood deal of self
knowledge and also particu lar o rgani sational ski l ls to p rovide such optimal
learni ng conditions. lt is apparent then that older learne rs h ave a deeper
u nderstandi n g of their learni ng p rocesses and requirements.
9JO
Although the se lf-management strategy was i ncluded i n the studies carried out
by O'Malley , Chamot and othe rs , the ro l e of this strategy was not explored in
relati on to age .
Olde r learn e rs were fou nd not o nly t o use more se lf-manag e m ent to co ntrol thei r
languag e l earni ng but also showed an i ncreased use of advance o rganisation,
de layed productio n , revision and self- mo nitori ng. That is, older learne rs showed
a g reater p refere nce for g ai n i ng an ove rview of the learni ng mate rial before
begi n ning particu lar tasks , for postpo n i n g speaki ng , for reviewing material
reg u larly and fo r m o nito ri ng their perfo rmance in the TL. A n u mber of
explanations can be put fo rward fo r t h e deve lopment of these fou r strateg ies
among more m at u re learn e rs.
The increase in the use of delayed p roductio n with age , which was found i n this
21 3
study, h as oft e n also been obse rved by language teachers. They note that the
older students fe el less comfortable with the o ral aspects of lang u ag e use and
are not so ready as you nger learners to e n g age in tasks requ i ri n g speaki ng in the
TL. This m ay be due to earlier exposu re among older learners to m o re traditi onal
forms of lang u age teaching with little e mphasis on speaki n g skills. lt is also
possible t h at o lder learners display a g reate r co nce rn fo r accu racy in their
languag e use , which can work ag ainst the develo pment of o ral fluen cy.
lt is not i m m ediately obvious why se lf- mon itori ng should incre ase with age . Se lf
monito ri n g has bee n identified as a co re l earning strategy (C hamot and Ku pper
1 989) and i s co nside red critical for prog ress in lang uage learni ng . lt may be that
olde r learne rs have a g reate r aware n ess of the i mportance of monitori ng
perfo rmance i n the TL as an ingredi e nt i n the developme nt of TL co mpete nce ,
and so d e ploy this strategy frequently.
Revision was used 'rarely' by learn e rs u nder 20 and 'ofte n' by learne rs ove r 60.
One reaso n fo r the i ncreased use of revi sion among o lder learn e rs cou ld be due
to a decli n e i n powe rs of memory with age. Skehan (1 986) in a study which
attempted to e stablish learner types e m pi rical ly found that older l earners had less
i m p ressive m emory fu nctions than a you nger g roup who were m o re readily able
to ass i m i l ate larg e amou nts of TL m aterial. The comments fro m older learn e rs i n
t h e cu rre nt study indicate a n aware ness o f t h e i mpo rtance o f revision fo r the m
in o rd e r to retai n TL forms. Older learne rs also fou nd it h e lpful to gai n an
ove rview of the topic and o rganisation of materials before actually begi n n i ng to
wo rk with t h e m .
21 4
acquired i n real-life than in the class room'. H e goes on to arg u e t h at the fu rther
the learn e r i s away from the classro o m i n time, t he m o re likely s/he is to use
m etacog n itive approaches to studyi n g .
One powerfu l i n dicator o f t h e i ncrease i n use o f metacog n iti o n with age and
experie nce i n this study was fou nd to be the use of se lf- manag e m e nt. Self
manag e ment deve lops with expe rience and with co nfidence i n one's own
experi e n ce as the measure of how best to p roceed. The i ncreased control of the
lang uag e learn i ng process exhibited by more matu re learn e rs appears to be the
result of se lf-knowledge obtai ned t h roug h t rial-and-e rro r and also of an ongoing
co ncern to discove r the most app ropri ate lang uage learn i ng strategies for
themse lves.
In Chapte r 3 it was noted that a l most 70% of the sample had had some
experie nce with the TL before e n rolling in the cou rse. Fu rt h e rm o re , it was evident
that among the learne rs who p rog ressed to higher leve ls of language study, an
i ncreasing p roportion of them had had prior TL learn i n g experie nce before
e m barki ng o n u n ive rsity language study. Such prior TL e xperi e nce was fou nd to
be the mai n i nflu ence on the CSU of learn ers. That is, the m eans learners used
to i nte ract with the TL were depe nd e nt on the presence o r abse nce of prior
learning experie nces in the TL.
Out of the list of 1 6 cog nitive strategies, resou rci ng , su bstituti o n and translatio n
both to and f rom E nglish were ide ntified a s the key strateg i e s i n disti ng uishi ng
betwee n learners on the basis of pri o r TL experience. Th e use of t h ese strategies
fo rmed a n i nteresting pattern . Learners with p rior T L e x p e rie nce made
sig nificantly g reate r use of substitutio n and resourci ng, strateg ies which requ i re
a certai n facility with the TL and an abi lity to seek alte rnatives. Substitution
i nvolves selecting alte rnative approaches, wo rds or phrases to accomplish tasks ,
and resou rci ng i nvolves usi ng alte rnative sources of i nformation about the TL.
Both these strategies requi re resou rcefulness on the part of t h e language learne r
21 5
and the abi lity t o use alte rnatives. lt is n ot su rprisi ng that these strategies are
more readily used by learners who have a l ready had co ntact with t h e TL si nce
their p ri o r experi e nce wou ld not o n ly h ave e nabled them to develop a reserve of
TL fo rms, but i n all probability would also have exposed t h e m to a number of
sou rces (textbooks, dictionaries, tapes , workbooks) from which to learn the TL.
2JD
3FD
Translatio n requires less active t ran sfo rmatio n of the TL than substitution and is
g e nerally accepted as a hig h ly i nefficient strategy for language learni ng (O'Malley
et al . 1 985a:39). Research conducted by McG roarty ( 1 988) i nto university
learners of J apanese and Spanish fou nd that 'conscious co m parison of the new
lang uag e with the native language' was co nsistently n eg ative with respect to
learning outcomes (McG roarty and Oxfo rd 1 990:72). However, in the cu rrent
study learn e rs with little experi e nce of the TL felt that t rans lation was an
i mpe rative , and a n ecessary stag e t h rough which they m ust pass before a more
auto matic cont ro l of the TL cou ld be d eveloped.
Cohen and Aphek ( 1 981 ) found that certain strategies (e.g. co ntextualization)
were difficult for beginning leve l stud e nts to use because t h ey presu me some
level of p roficie ncy. This explanation wou ld apply equal ly well t o the disti nctive
21 6
use of substitutio n by m o re experie nced learners i n this study. Furth e rm o re , many
sou rces about the TL such as text books and g ram mar books can p robably o n ly
be accessed with ease when the learn e r has a particu lar l evel of ski l l i n the TL.
In the questi o n naire study so m e effects were fou nd fo r the i nfl ue nce of
profici e ncy o n M S U , though these effects were m uch less p ronou nced than those
fo r mode of study and the age of learn e rs. Certai n ly, the use of m etacog nitive
co ntrol i n learn i ng has been linked with success in a numbe r of studies (O'Mal ley
et al. 1 985a, 1 985b ; Wenden 1 986a, 1 986b, 1 987a, 1 9 87b ; Du ran 1 987; Carrell
1 989; Carre l l et al. 1 989). Howeve r, pronou nced diffe re nces i n M S U between
effective and less effective learn e rs did n ot e merge in this study. The reaso ns fo r
t h is cou ld be related to the high level of educational expertise of the tertiary- leve l
learne rs , c o mpared to the hig h-school learn e rs in other studies. As Skehan
( 1 989) notes, cautio n is required when atte m pting to link profici ency to strategy
use in stu di es co nducted with learn ers who are al ready, in terms of the g e neral
populati o n , a select g roup.
Gender
Results f ro m the qu estion nai re study did not reveal sig nificant diffe rences i n
eith er M S U o r C S U between m e n and wo men u nderg radu ate learners. The
research of Oxfo rd and eo-worke rs usi ng the S I LL questionnai re ( E h rman and
Oxfo rd 1 989 ; Oxfo rd and Nyikos 1 989) revealed that wo men u se strategies more
often than m e n and also deploy a wider range of strategies t han m e n . The
opposite fi ndi ng s e m e rged fro m a study carried out by Tran ( 1 9 88) who found
that Vietnam ese men used more strategies than a com parable sample of
Vietnamese wo m e n . This study did n ot i nvestigate g ende r diffe re n ces in strategy
use t h roug h the ve rbal report p rocedu re nor in relation to social and affective
strategy use. Fu rther investigatio n of g ender diffe re nces i n strategy use throug h
these m eans is requ i red.
21 7
Language Learning Experience
The expectatio n that the learni ng of foreig n languages is m ade easi e r once the
learner has al ready had experi e nce of fo reign lang uage learn i ng is based on the
ass u m ption that the learne r wi l l al ready h ave deve loped a reperto i re of learning
strategies which can be applied to the n ew task. Howeve r, the resu lts of this
stu dy reveal that the learni ng of other languag es is not a p redo m i n ant i nflue nce
on the m etacog nitive and cog nitive strategies learners depl oy i n learning a new
fo reign lang uage. Of far greater i mportance is the e xtent of experi e nce in learning
the TL, and this i nflue nces t he ki nds of cog nitive strategi e s which are used (see
section 6 . 7 . 2) . Th us, while learne rs m ay have developed a particu lar strategic
repertoi re fo r foreign language learn i n g , it appears that these st rateg ies can not
be auto matically applied to a new foreig n languag e . lt is o n ly w h e n learners have
reached a particular level of proficie ncy i n the TL that certain cog nitive strategies
can be deployed.
Mo tivation
An i m portant c riticism of the Oxford and Nyikos ( 1 989) study, put fo rward in
Chapter 2 (section 2.5.9) was that operati onal defi n itio ns of m otivation were not
give n , neithe r were the precise m eans fo r measu ri ng this co mplex co nstruct.
The Rost and Ross (1 991 ) study on Learner use of strategies in interaction:
21 8
typology and teachability does not m easu re the motivation of lang uage learne rs
n either does motivation fig u re i n t h e discussi o n of the fi ndi ngs. The findings do
suggest , h oweve r, that the teac h i n g of particu lar questio n i n g strategies can
i nflue nce t h e i r subsequ e nt use. These fi ndi ngs relate to the resu lts of studies
carried out i nto the relationship betwee n motivation and strategy t rai ni ng (e.g
O'Malley et al. 1 985b; McCombs 1 98 8 ; Pari s 1 988). H oweve r, they do not
add ress d i rectly the question of the relationship between motivatio n and strategy
use indepe ndent of strategy inst ruction.
The negative findi ngs of this study with respect to the influe nce of motivation on
strategy use are consistent with those of an earlier study carri ed out by Politze r
(u ndated) reported by Oxford ( 1 990 ) . Politze r studied t h e lang uage learning
strategies of O riental and Hispanic g raduate students learning E ng lish and fo und
that they we re i nstru mentally rather than i nteg rative ly motivated to learn the
languag e . I n st rume ntal motivation accou nted fo r course g ai ns b ut little evide nce
existed fo r a link betwee n the strateg ies used and motivatio nal o rie ntati o n .
What i s clear from the fi ndi ngs o f these studies is that more carefu l research is
required befo re either the level of m otivati onal i nte nsity , or the motivational
o rie ntation of languag e learne rs can be re lated to strategy choice.
6.8 S U M MARY
Fro m this study it is evident that the abse nce of face-to-face classroom instruction
has particular effects o n the metacog n itive and social strategies used by di stance
lang uag e learners. We have an i mage of distance learners who use all
d i m e nsions of metacog nitive co nt ro l (plan n i ng , mo nitori n g and evaluation) to
manage t h e i r learning p rocesses. In cont rast, the u se of metacognitive strategies
by c lassroo m learne rs i s concerned m ostly with plan n i n g . lt i s also appare nt that
distance learners have thought about t h e ways they learn best, and h ave devised
ways of p roviding for themse lves the particular learning co nditi ons which they
co nsider to b e o pti mal. Classroom learn e rs and d istance learn e rs a re re lative ly
similar i n thei r u se of cognitive strategies, e xcept for the markedly g reater use of
21 9
e labo ration strateg ies by distance learners.
Fi ndi ngs fro m the curre nt study also revealed that the extent to which learners
use social as opposed to affective strategies is re lated to their mode of study.
Distance learners use affective means to support their learn i ng m o re than social
means, while the reverse pattern is the case fo r classroom learn e rs. In the face
of more l i m ited opportunities fo r SSU distance learners have fou nd ways to solicit
fro m oth e rs the e ncou ragement necessary fo r the conti n u ation of their learning
e ndeavo u rs .
The cog n itive strategies learners used were fou nd to be stro ngly i nfluenced by
the prese nce or absence of prior experi e nce with the TL. N ovice learners we re
not able to de p loy higher-order strategies such as substitution to the same extent
as learners who had al ready learnt something of the TL. N ovice learners we re
also more re liant on translation.
220
7. CONCLUSION
The conclusions d rawn in this ch apte r are diffe re ntiated fi rstly with regard to the
theoretical and methodo logical i mplications of the study (section 7 . 1 and section
7.2). The n a number of practical applications of the study are p ro posed (section
7.3), and in the fi nal section the i m p lications fo r furthe r researc h d i rections in the
field of lang uag e learning strateg ies are discussed .
1.
2.
3.
The m etacog nitive , cog nitive , social and affective strategy use model is a
co m mo n ly used framewo rk for research into learn i n g p rocesses in
educatio nal psycho logy and cog nitive psychology. Users of this model in
applied linguistics research can e xpect to benefit f ro m d eve lopments in
these re lated discipli nes.
221
4.
learning
(e.g
Gardner and
Lambert
1 972 ;
Mclaug hlin
1 978 ;
Mclaug h li n , Rossman and Mcleod 1 983 ; Gardne r 1 985) did not i nclude
languag e learning st rateg ies. More rece ntly Skehan (1 989) and Gardner and
Mac l ntyre ( 1 992, 1 993) have represe nted strategy use t h roug h two models.
Skehan 's ( 1 989) model of i nflue nces o n lang uage learn i n g is rudi me ntary i n
e xplanato ry te rms but i s an atte mpt t o provid e a framewo rk t h rough which to
investigate the ro le played by a range of individual differe nces in the language
learning p rocess. A more recent model by Gard n e r and Macl ntyre (1 992, 1 993)
is a revisi o n of Gardner's (1 985) socio-educatio nal model of seco nd lang uage
acquisitio n . H e re lang uage learn i ng strateg ies are represented as cog nitive
variables i n the acquisition process along with i ntel ligence and lang uag e aptitude.
While t h e re are some difficu lties with this theo retical formu lati o n , particu larly i n
t e rms o f the relationship between m otivation a n d strategy u s e a s discussed i n
Chapter 6 , t h e model represents a ti m e ly attempt at p roviding a co nceptual
i nteg ration of i ndividual characte ristics in seco nd lang uag e learn i n g . lt is hoped
that the refi n e m ent of cu rre nt mode ls of second lang uage acqu isiti o n wi l l co nti nue
and that fu rt h e r attempts wi ll be made to inco rporate learni ng strategies and the
i nflue nces on l earning strategy use i nto such models.
222
7.2 M ETHODOLOG I CAL I MPLI CATIONS
This study co nstituted a conve rgent assessme nt of the influe nce of mode of study
and the TL on metacog nitive , cog nitive , social and affective strategy use. Throug h
the questionnaire data, variance i n strategy use was re lated to a n u mbe r of
learner characte ri stics and to aspects of the language learn i n g co ntext. The
m ethodological implications of the study are presented below.
1.
2.
I n languag e learning strategy research i t i s also critical that the means for
collecti ng data on strategy use a re carefu lly descri bed becau se of the
appare nt i nflue nce of data co llection procedures on reports of strategy use
(section 3.4. 1 ). I n order to be able to i nterpret accu rately the resu lts of
particular studies, and in o rder to be able to assess their g e n e ralisabi lity ,
it is i m perative that careful descri ptions of research desi g n s are avai lable.
Futu re research wi l l benefit fro m the i nclusion of more precise descri ptions
of m ethodo log ical approaches to the i nvestigation of i nflue nces o n strategy
use.
3.
223
can not accu rately attribute reports of strategy use to i ndividual learners.
Th us this procedure excludes the possi bility of investigating the
re lati onship betwee n particu lar strategies and i ndividual characteristics. lt
is simply u n m anageable for most researchers to conduct individual
i nte rvi ews with a larg e sam ple of learne rs.
4.
5.
Whi le we know that diffe re nt d ata co llection methods i nflue nce strategy
use reports, we have not as yet c larified exactly how such methods i mpact
o n strategy use reports. Carefu l co mpariso ns need to be made betwee n
resu lts obtai ned fro m a si ngle sam ple using diffe re nt data collection
procedures, and more precise conclusions need to be drawn concerning
the ways in which particu lar procedures are like ly to i mpact on the reports
which are obtai ned. This is n ecessary if we are to o btai n more focused
and more re liable i nsig hts i nto the strategy use of learne rs. Such
co m parisons between data co l lection p rocedures wi l l , of course, o n ly be
possible if more multiple m easu reme nts of strategy use are obtai ned i n
studies o f learn i n g strategies.
224
strategy research. In particular, the methodological implications of this study
u nderli n e the i mportance of multi m ethod assessment of strategy use and poi nt
to the need for the conti nued explo ration of diffe re nt ki nds of ve rbal report
techniques. Fu rthe rmo re , attention n eeds to be given to the deve lopment of ways
of co llecti n g co nve rgent data o n strategy use with large g roups of learne rs , i n
o rder t o p rovide more valid measu res o f i nflue nces o n lang uage learning
strategies.
The fo l lowi ng practical reco mme ndations are put fo rward based o n the findings
of the c u rre nt study. They are te ntative reco mmendations si nce the study did not
e ncompass the dimension of strategy training, n either did it seek to develop
specific g u idelines fo r lang uag e learners and teachers. H oweve r, the study did
demonstrate the i m po rtance of m etacog nitive ski lls both in language learn i n g , and
in the deve lopment of learn ers' se lf-knowledge of how to learn. A pri mary g oal
fo r lang uage i nstruction is to avoid the creation of what B ru n e r ( 1 966:53)
descri bed as 'a fo rm of mastery that is co nti ngent upon the pe rpetual presence
of the teacher' and to e nable the learner to become se lf-sufficient. The
sugg estio n s outlined be low are proposed as i mportant conside rations if languag e
learners are t o b e se lf-directed and if they are t o learn 'how to do for the mselves
what teachers typically do fo r them in the classroom' (We n d e n 1 985:7) .
1.
2.
225
3.
Language learners should be p rovided with e xperi e nces which allow them
to assess their strategy use. I n additi on they should b e e ncou raged to
e xp and thei r strateg ic re pe rtoi res.
4.
5.
6.
7.
Distance education language learn i ng mate rials should i nco rpo rate
e l e m ents of learner trai n i ng which are approp riate to l earners who do not
h ave access to reg ular face-to-face i nte ractions to support their learning
e n deavou rs .
8.
U ntil such time as distance lang uag e cou rses incorpo rate strategy trai ning
wit h i n their mate rials, distance languag e learne rs should be e ncou raged
use strategy t rai n i ng cou rses ( e . g . Rubi n and Tho mpson 1 982; Brown
Azarowicz, Stan nard and G o ldin 1 986; Ellis and Si nclai r 1 989 ; Wi lling
226
1 98 9 ; White 1 992) as part of the process of deve lo pi ng thei r lang uage
learning ski l ls.
9.
The i n si g hts from this study rai se a n u mber of fu rther questio n s i ncluding : Are
there similarities betwee n the strategy use of distance learn e rs and the strategy
use of learners in othe r autonomous lang uage learni ng e nvi ronments? Do
learn e rs of Chi nese also rely on more mechanical cog nitive strateg ies to master
the TL writing syste m ? How are strategies such as e labo rati o n actually used by
distance learners in re lation to particular language tasks? Is it possible to identify
fu rther ki nds of social and affective strategies? Is social and affective strategy use
re lated to fu rthe r aspects of the lang uag e learni ng co ntext , o r to learner
characte ristics?
N u m e rous u nanswered questions re mai n , and the final part of this chapte r
suggests specific ave nues fo r fu rther research into languag e learning strategies.
But firstly , the mo re g e neral implicatio n s of the resu lts of this study are p rese nted
below.
1.
227
2.
3.
4.
There is a need in strategy use studies for the deve lopm e nt of more valid
m e asures of i ndividual learn e r variables, in particular fo r measu res of
m otivatio n .
5.
6.
7.
228
and
learner vari ables i nevitably ove rlap and i nte ract with others
(d'Ang lejan and Re naud 1 985) . As Larsen-Free man and Lo ng (1 991 ) point
out, this suggests that we are not getting a t rue measu re of a facto r if we
isolate it fro m othe r factors . Thus more co mplex m u ltivariate statistical
tech niques should be used to e xamine the re latio nships betwee n strategy
choice and i ndividual variables .
The fi ndi n g s from the study also sugg est a nu m ber o f specific ave nues for fu rther
research i nto lang uage learni ng strategies and these fo rm the concluding section
of this chapter.
1.
2.
M o re detai led study is required of the social and affective dime nsions of
strategy use. The existi ng categories of social strategy use (e . g .
qu estioning) a n d of affective strategy use (e .g . se lf-talk) need t o b e
refi ned. I nflue nces on t h e u s e of social and affective strategies n eed to be
fu rt h e r explo red. In particu lar, the ro le played by gender and motivation in
the u se of social and affective strategies is worthy of atte nti o n .
3.
The fi ndi ngs from this study a re based on learners' se lf-descri ptions of
perfo rmance. These se lf-descriptions need to be u lti mate ly validated
agai n st the learn e rs' actual pe rformance on lang uag e tasks in naturalistic
setti ngs. Th us, for distance learners their distinctive use of se lf
management, advance o rganisation and monitori ng should be further
229
i nvestig ated , possibly t h rough i ntrospective tech niques, as the l earners
wo rk on TL tasks i n their h abitual learn i ng context.
4.
More research is needed to ide ntify how key strategies such as self
monito ri n g or elaboratio n are used in relation to particu lar tasks e . g .
writi n g , listeni ng , reading . A stro ng precedent has e m e rged f o r this type
of analysis in cog nitive psychology (e .g. Alexander and Judy 1 989 ; Pe rki ns
and Salomon 1 989).
5.
6.
The i nflu ence of a new TL writing syste m (sy llabic o r logog raphic) on the
strategies learn ers use requ ires fu rther study. We have little u nderstanding
of how a new ki nd of writing system im pacts o n the way learners atte mpt
to engag e with the TL materials. The re latio nship betwee n particular TL
writing systems and cog n itive strategy use should be explo red.
7.
The ro le of diffe rent ki nds of p rior TL experie nce (e.g. formal school-based
i nstructio n , i nformal learning in the TL cou ntry, i nfo rmal learn i ng with fam i ly
members or friends) i n the choice of cog nitive learning strategies is a
p romising area fo r fu rt h e r research.
8.
The re lationship betwee n prior TL experie nce and strategy use needs to
be investigated using longitudi nal comparisons of strategy use. Th rough
such research it wou ld be possi ble to compare the fi ndings fro m this study,
co nce rning the effects of prior TL experie nce o n strategy choice, with
those based on the changes in the strategy use of i ndividuals over time.
230
The final chapter provides a s u m m ary of the key aspects , and limitations, of this
study.
231
8. S U M MARY
The p ri m ary i ntent of this study was to e xamine the re latio nship between mode
of study and the use of languag e learni ng strategies by fo reig n lang uage
learners . In additio n , a co mparison was made of the sig nificance of mode of
study relative to oth e r influe nces fro m the lang uage learn i n g context and from
learner characteristics. The opening secti o n of the study situated the i nvestigation
of language l e arn ing strategies in the more gen eral co ntext of the deve lopment
of research i nto language acqu isition processes. The defi n i n g characteristics of
the distance mode of study, as o ne of the language learni ng co ntexts in the
cu rre nt study , we re then i ntroduced. Recent co mmentari es o n the g rowth of
lang uag e learning at a distance and on our li mited u nderstanding of the
processes of learn i ng in this co ntext we re presented . Fu rthermore , the co ntext
of distance language learn i ng was established as an autono mous setting fo r
lang uag e l earn ing and the i m portance for the distance learner of acqui ri ng the
means fo r learning in the abse nce of teacher di rection was e m phasised. Th e
i m petus fo r the cu rre nt project was seen to be de rived fro m the field of language
learning strategy research togethe r with the evident need fo r research i nto the
lang uag e learn ing behaviou rs of learn e rs in no n-classroo m setti ngs.
The early studies of lang uag e learn i n g strategies were exami ned and the
i m portance of i nsig hts fro m the fields of educatio nal psycho logy and cog nitive
psychology
i nto the
metacog n itive
processes was
acknowledg ed. The range of criteria used to classify strategy use in the early
studies was considered and issues re lati ng to the ide ntificatio n , defi nition and
classification of i nstances of strategy use were raised. Specific studies of
i nfluences o n strategy use were then discussed to establi sh the cu rrent state of
knowledge i n the field, to reveal g aps in our u nderstandi ng , and as yet
u nexplored ave nues fo r researc h . The research p roblem to be investigated in this
study was t h e n outli ned as an exami n ation of the contri butio n of the lang uag e
232
learning
co ntext and
strategy choice
of
The setting for the study was a dual-mode tertiary i n stitution offe ri ng
u nderg raduate foreig n language cou rses (Fre nch , G erman , Japan ese , Chi nese)
throu g h a classroom prog ramme and a distance p rogramme. The study took
place with an i ntact g roup of learners , and partici pation was vo lu ntary. Subjects
fo r the ve rbal report study (N=37) were a subsam ple of the learners who
partici pated i n the qu estionnaire study (N=41 7). A descri pti on was m ade of the
characte ristics of the qu esti on naire sample and of their distribution accord i ng to
ag e , g e nde r, TL (Fre nch , German ,Japanese, Chi nese ) , level of study and p rior
TL experie nce.
Ope ratio n al defi nitions for the fou r variables re lating to the learn i ng context
(m ode,TL, leve l of study , language use oppo rtu n iti es) were give n . In additio n ,
operati o n al defi niti ons for the six variables relating to learn e r characte ristics (ag e ,
ge nder, language learning experie nce , pri o r experie nce learni ng t h e TL,
motivation p rofici e ncy) were presented. Reaso ns fo r the choice of a model of
strategy use which was derived orig i n ally fro m work in cog nitive psychology,
namely the metacog nitive , cog nitive , socio-affective mode l , were presented. F rom
this model the strategy use variables to be i nvestigated i n the study we re
obtai ned.
The six research qu estions fo r the study we re described. lt was e mphasised that
the most i n-depth parts of the study co ncerned the relationship between strategy
choice and both mode of study and the TL. lt was also emphasised that the
investigati on of the re lationship betwee n other variables and strategy choice was
to be li m ited to the m etacog nitive and cog nitive di mensions of strategy use.
The research design used to examine influe nces o n strategy choice was d e rived
fro m p revious studies in which the advantages and lim itatio n s of various d ata
collectio n tech niques had been trial led. Two inst ru ments were used to gathe r the
233
data fo r the study, namely a s elf-report questi o n nai re and a ve rbal report
procedu re known as the yoked subject techniqu e . The advantages and limitations
of these instru m e nts we re descri bed. For the questi o n nai re procedu re attempts
to li mit the social desirabi lity influence on su bjects were outlined. A detai led
co nsid e ration of the limitatio n s of ve rbal report p rocedu res was made , and
specific g uideli nes were fo l lowed in an attem pt to co u nteract some of the
pote nti al weaknesses of the p rocedu re . These steps i ncluded the use of a warm
up p hase, the provi sion of learn i ng materi als in re lation to which learners reported
on strategy use, and the collecti o n of converg e nt data on strategy use.
Pi lot studies we re carried out to test both the p ractical aspects of administe ri ng
the i n st ruments and the clarity of instructions and questions. I n the case of the
question nai re procedure the pi lot study was used to determine the i nte rnal
co n siste ncy re liabi lity of the strategy use scales. A m o re g e n eral aim of the pi lot
study was to t rial the viabi lity of t h e procedu res. On the basis of the fi ndi ngs fro m
t h e pilot study a n u m ber o f modificati ons were i ntroduced t o the wo rding o f some
of the questi ons and to the freque ncy respo nse scale. I n addition, the socio
affective scale was separated into two scales, thus yielding high i nte rnal
co n siste ncy reliability scores. The ve rbal transcri pts o btai ned through the yoked
subject pi lot study we re used to deve lop procedu res for the coding of instances
of strategy use and for the trai ni ng of an assi stant rater.
The p rocedu res used to ad m i nister the qu esti o n n ai re instru m e nt and the ve rbal
report instrument i n the mai n study were carefu l ly described. I n particular, care
was t ake n to e nsure that the procedu res were as s i m i lar as possible for
classroom and distance learne rs . The methods for p rocessing the questionnai re
d ata and the verbal report d ata were detai led. I n the co ncluding sections of the
m ethodology chapter the criteri a of re liability and validity we re applied to the
research desig n , and li mitatio n s of the study were conceded.
In o rder to analyse the relationship betwee n the strategy use vari ables and the
learn i ng context o r learner characte ristic variables, canonical variate analysis was
234
app lied to the question nai re d ata. The use of this type of m u ltivariate an alysis
was e x p lai ned. In additio n , Du ncan's multiple range test was used to locate the
sig n ificant diffe re nces betwee n differe nt age g roups i n term s of their M S U .
1.
The p redomi nant i nflue nce on MSU is the mode of study of languag e
learne rs . Classroo m
2.
3.
235
4.
5.
6.
7.
Learn e rs under thirty are sig nificantly diffe rent from learners ove r thi rty i n
terms o f their frequency o f M S U .
8.
9.
1 0.
S oci al and affective strategies are used i nfrequ e ntly by u nderg raduate
fo reign language learners.
11.
236
1 2.
1 3.
The predominant i nflu e nce on CSU i s p ri o r experi e nce with the TL.
Learne rs who have h ad such e xpe rie nce befo re enrolli ng in a u niversity
language cou rse make g reater use of resourci n g and su bstituti o n , and less
use of translati o n , than do learne rs who have not had such prior
experie nce.
1 4.
The need to maste r a n ew writi ng syste m i nflu e nces the ki nds of cog nitive
strategies learners of J apanese use , i n particu lar the marked appl icatio n
o f t h e writi ng out strategy.
1 5.
Further diffe re nces i n CSU betwe e n learn e rs of diffe rent Tls , which
i nvolved the increased use of t ranslation by learners of Japanese and the
frequent use of elabo ratio n by learners of Fre nch , can n ot be automatically
attri buted to the i nfluence of the TL per se . Language teaching
m ethodologies and the hours take n to attai n a particu lar level of mastery
i n different Tls also appear to contribute to the ki nds of cog nitive
strateg ies learn ers use.
1 6.
The li mitations of this research have bee n acknowledged at appropriate poi nts
t h roug hout the thesis. These limitations are su m marised be low and should be
born e in mind in any i nterpretation of the fi ndi ngs fro m the study.
237
1.
2.
3.
There was no accu rate measure of durati on of study i n the research. Leve l
o f study did not reflect the deg ree of mastery of the TL in a co nsiste nt way
across diffe re nt TL g roups . P rior TL expe rience proved to be a better
i ndicato r of the du ration of TL study than the level of study variable.
4.
5.
6.
Findi ngs fro m the ve rbal report study can not be seen as p roviding a
co mpre h e nsive accou nt of the strategic re pertoi re of each learne r si nce
learne rs p robably re ported on a particu lar subset of strateg ies they were
co nscious of at the ti m e of the procedu re . I n additio n , subjects may have
varied in thei r abi lity to talk about thei r strategy use , and thus it is possible
that their reports were re latively i nco m plete com pared to their actual
strategy u se.
238
7.
The study did not i nclude a check of self-reports ag ai nst learn e rs'
performance o n particu lar languag e tasks. This remai ns a hig h ly desirable
s o u rce of co nfi rmation about the re liabi lity o f se lf-report data.
The theoretical and m et hodological i m plications of the study have been o utli ned
in C hapte r 7. The p ractical reco m me ndations fro m the study were o ri e ntated
towards the need fo r learn e r trai ni ng to be e m bedded in languag e learning
p rog ram m es for both c lassroom and distance learners. Several ave n ues for
fu rthe r research were also put forward .
Finally, the findings fro m t h i s study have permitted n u m e rous i nsig hts into
variance i n lang uage learning strategy use d u e to mode of study , the ag e of
learn e rs , prior TL experie nce and, to a lesser e xtent, the TL. Such fi ndi ngs
provide a strong e r fou ndation fo r theory construction conce rning i nflue nces on
diffe re nt dimensio ns of strategy use, and , m o re generally, co nce rning the
co ntri butio n of lang uage learning strategies to the process of seco nd lang uage
acquisiti o n .
239
APPENDIX A
SAMPLE PI LOT Q U ESTIONNAI R E (G ER M A N)
LANG UAG E LEARNING STRATEG I ES
The followi ng q u esti on nai re is part of a study looki ng i nto how i nternal and
extramu ral stude nts go about learni ng a lang uag e in private study. So, you wi l l
b e asked about y o u r own strategies f o r learn ing German o u t of class - that is, the
tech niques you use when you are studyi ng at free poi nts d u ri n g the day , in the
eve n i n g or whe n ever.
You a re asked to respond to a series of qustions in te rms of how often (if at all)
you e ng ag e in particu lar learn i ng behaviours. There are no rig ht or wrong
answe rs , no good or bad strategies. You should simply choose the response that
best d escribes you r own out-of-class language learni ng experie nce.
The q uesti ons relate to how ofte n you do something . Fo r e xample
Before you start a task h ow ofte n do you look thro u g h it to get a g e neral
idea of h ow it is o rgan ised and what it is about?
usually
5
often
4
3
someti mes
rare ly
2
never
1
Some of the questions i n Part 3 also ask whether you have the opportu nity to use
certai n strategies. For example
How often do you work tog ether with yo u r fel low learn e rs to so lve a
p roble m , p ractise co nversations, check over a task . . . ?
usually
5
often
4
sometimes
3
rarely
2
neve r
1
no opportu nity
0
240
Re member to answe r i n terms of what you actually do, not i n terms of what you
thi n k you should do.
Th ese questi ons are not a part of the assessme nt of the co u rse , and all the data
co llected wi l l be kept co mpletely co nfidential .
P lease add any com me nts to the qu estion nai re co ncerning ite m s o r i n st ructions
wh ich are n ot clear, o r about you r use of some of the strategi es.
241
Part 1
How often do you repeat German wo rds, phrases o r se nte nces - out loud
or to yourself?
u sually
5
4
often
so m eti m e s 3
rarely
2
1
n ever
2.
D o you use sources o f i nfo rmation about German such as dictio naries,
other textbooks?
u sually
5
4
often
someti mes 3
2
rare ly
n ever
1
3.
4.
a s they occur w h e n
5.
5
4
3
2
1
5
4
3
2
1
242
6.
7.
How often do you m ake a me ntal pictu re of the spelling of a word you
want to re m e m ber, or of the object itself?
usually
5
often
4
sometimes 3
rarely
2
1
neve r
8.
9.
1 0.
243
11.
sentence as part of
5
4
3
2
1
1 2.
1 3.
When you work with German how ofte n do you translate i nto Eng lish or
fi nd you rse lf relati ng the German wo rds to Eng lish wo rds?
usually
5
often
4
so meti mes 3
rarely
2
n ever
1
1 4.
1 5.
Do you use other parts of the se nte nce o r passage to fig u re out the
m ean i ng of unfamiliar language ite ms?
usually
5
often
4
someti mes 3
rare ly
2
n ever
1
244
1 6.
1 7.
245
Part 2
2.
3.
When you stu dy do you decide i n advance to focus on specific tasks and
to ignore i rre l evant distractio ns?
usually
5
often
4
so metimes 3
rare ly
2
neve r
1
4.
246
5.
6.
When you have difficu lties in completi ng a language task how often do you
ide ntify aspects of the task which are hi nderi ng you r p rogress - for
e xample , difficult structures, u nfami liar acce nt , advanced vocabulary?
usually
5
often
4
so meti mes 3
2
rare ly
1
n eve r
7.
While you are doi ng a lang uage task do you mo nito r your use of lang uag e ,
co rrecting i f n ecessary your p ro n u nciati o n , g rammar, style , etc. ?
usually
5
often
4
sometimes 3
2
rarely
1
n eve r
8.
Do you check through your work when you have fi nished a task?
usual ly
5
often
4
sometimes 3
rarely
2
1
n eve r
247
9.
1 0.
248
Part 3
Th ese questi o n s ask you about the e xtent to which you i nteract with othe rs in
learning G e rman - and also about h ow often you wo rk with you r own emotions
and attitudes towards languag e learn i ng . Remember there i s no sense in which
you should be doing any of these t h i ngs.
1.
speake r fo r
5
4
3
2
1
0
2.
How o fte n do you wo rk toget h e r with your fel low learn e rs to solve a
p rob le m , p ractise co nversatio n s , ch eck over a task . . . ?
usually
5
4
often
so meti mes
3
rarely
2
1
neve r
no opportu n ity
0
3.
249
4.
5.
250
Fi n al ly, please a n swe r the followi ng questio ns i ncluding detai ls about your
p revi ous language l earn i ng experie nce . This i nfo rmation is req u e sted to help us
to analyse and i nte rpret the data we col lect .
I n o rder t o e n s u re confidentiality all i nform ati o n wi ll be coded - a t that point your
n ame wi ll disappear and be replaced by a n u m ber!
1.
Name
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
Age
_
_
_
_
_
_
3.
Sex
_
_
_
_
_
_
_
4.
5.
_
_
_
_
_
Where?
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
7.
H ave you learnt any othe r languages apart from English a nd German?
(Yes/No)
___
Which ?
W h e re ?
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
251
8.
How i mportant to you are the fo llowi ng reasons for learni ng German ?
Please respond to each possi ble reaso n .
I a m studying G e rman
(a )
(b )
5
4
3
2
1
5
4
3
2
1
5
4
3
2
1
(c)
(d )
(e )
252
(f)
(g)
9.
because it wi l l
e m pl oy m e nt.
be
u sefu l
for
my
present
e xt re mely i mportant
very i mportant
i m portant
not so i m po rtant
not i mpo rtant at all
5
4
3
2
1
5
4
3
2
1
(o r
futu re)
(a)
(b)
253
APPENDIX 8
S AMPLE QU ESTIO NNAI R E (C H I N ESE) : MAI N STU DY
LANGUAG E LEAR N I NG STRATEGI ES
Introd uction
The fol lowi ng question nai re is part of a study looki ng i nto h ow i nternal and
extramu ral stude nts go about learni ng a langu age in private study. Research
shows that i n g en e ral i ndividuals learn i n diffe rent ways and the techniques that
wo rk for so m e people do not necessari ly work fo r others. We are i nterested i n the
range of strategies that you , as a lan g u age learn er, use. S o , you will be asked
about you r own strategies fo r learn i ng Chinese out of class - that is, the
tech niques you use when you are study i ng at free poi nts d u ri n g the day , in the
evening or whe never.
A number of strategies are presented , some of which you m ay h ave tried , or
which you m ay have abandoned becau s e they are not effective fo r you. please
respo nd to e ach question in terms of how often (if at all) you engage in particular
learni ng be haviou rs. Fo r each question you should ci rcle the re sponse that best
descri bes you r own out-of-class language learning experi e nce. Fo r exam ple :
Do you put words i nto g roups as a way of re memberi ng them?
ve ry often
ofte n
someti mes
rare ly
neve r
5
4
3
Re member to answe r i n terms of what you actually do, not i n t e rm s of what you
think you s h o u ld do. Feel free to add any fu rther co m me nts you like.
These questi ons are not a part of the assessment of the cou rs e , and all the data
collected wi l l be kept co mpletely co nfide ntial.
254
Part 1
H ow often do you re peat C h i nese wo rds, phrases o r sente nces - out loud
or to yourse lf?
5
ve ry ofte n
4
often
someti mes 3
rarely
2
1
n eve r
2.
3.
4.
255
5.
u s i n g the languag e
5
4
3
2
1
6.
7.
8.
W h e n you are liste ning o r readi ng how often do you t ry to see i n you r
m i nd what is happe ning to clarify the meani ng ?
ve ry ofte n
5
ofte n
4
sometimes 3
rarely
2
n eve r
1
9.
256
1 o.
11.
1 2.
How often d o you make a m e ntal o r written sum mary of key languag e
ite ms o r info rmation contained i n languag e tasks?
ve ry often
5
often
4
sometimes 3
rarely
2
n eve r
1
1 3.
Wh en you work with Chi nese how ofte n d o you t ranslate i nto E n g lish o r
find you rself re lating t h e Chinese word s to Eng lish words?
very ofte n
5
often
4
so meti mes 3
rare ly
2
never
1
1 4.
257
1 5.
Do you use othe r parts of the se ntence o r passage to fig u re out the
mean i ng of u nfami liar languag e ite ms?
5
ve ry ofte n
4
often
sometimes 3
2
rare ly
1
n eve r
1 6.
1 7.
258
Part 2
B efore you start a task how ofte n do you look through it t o g et a g e neral
idea of how it is organised and what it is about?
ve ry often
5
4
often
sometimes 3
rarely
2
neve r
1
2.
3.
4.
Do you prefe r t o listen to an item seve ral ti mes befo re you use it i n
speaki ng?
ve ry ofte n
5
often
4
someti mes 3
rarely
2
neve r
1
259
5.
6.
W h e n you have difficu lties i n co mpleti ng a lang uag e task how often do you
ide ntify aspects of the task which are hi ndering yo u r p rogress - for
example , difficult structu res , u nfami liar accent, advanced vocabulary?
ve ry often
5
ofte n
4
sometimes 3
rare ly
2
1
n eve r
7.
Whi le you are doi ng a lang uage task do you ch eck on you r use of
languag e , co rrecting if necessary your pronu nciation, g rammar, style , etc. ?
ve ry often
5
often
4
so m eti mes 3
2
rare ly
n eve r
1
8.
Do you ch eck through . your work when you have fi ni shed a task?
ve ry often
5
often
4
someti mes 3
2
rare ly
n eve r
1
9.
260
1 0.
5
4
3
2
1
261
Part 3
These questi o n s ask you about the e xtent to which you i nteract with oth ers i n
learning C h i nese - and also about how often you wo rk with y o u r own e motions
and attitudes towards language learning .
Some of the questions i n this section re late both to how ofte n you d o so mething
and also whether you have the opportu nity o r not. So if you co u ld wo rk with your
peers but d o not do so, you wou ld respond to the qu estio n as fol lows:
How often do you wo rk toget h e r with you r fel low learn e rs to so lve a
p rob le m , practise conve rsati o n s , check ove r a task . ?
5
ve ry often
often
4
sometimes
3
rarely
2
n ever
GD
no opportu nity
0
. .
If you do n ot have the opportu nity to wo rk with other learn e rs you should
respo nd :
How o ften do you wo rk toget h e r with you r fel low learn e rs to so lve a
p rob le m , p ractise co nversations, check over a task . . . ?
ve ry ofte n
5
ofte n
4
sometimes
3
rare ly
2
never
1
no opportu n ity
@
Remember we are not sugg esti ng that you should o r should n ot be d o i ng any of
these things, s o please answer in terms of you r own learn i n g h abits.
262
1.
speaker fo r
5
4
3
2
1
0
2.
3.
4.
5.
How o ften do you e ncourage yourself i n lang uage learn i ng - fo r examp le,
by say i ng positive things to yourself to g ive you rse lf m o re confide nce ?
ve ry often
5
often
4
sometimes 3
2
rare ly
1
never
263
Part 4
Fi nally, please answe r the fo llowi ng qu estions i ncluding d etai ls about your
p revi ous language learn ing experie nce. This info rm ation is requested to help us
to analyse and i nterpret the data we co llect.
I n o rder to e n s u re confidentiality all i nfo rmatio n wi ll be coded - at that poi nt your
name wi ll disappear and be replaced by a n u mber!
1.
Name
( S u rname and initials)
2.
Age
u nder 2 1
2 1 -30
3 1 -40
4 1 -50
5 1 -69
ove r 60
1
2
3
4
5
6
3.
Sex
Male
Fe male
1
2
4.
Mothe r To n g ue (s)
E ng lish
Other
P lease specify
1
2
5.
_
_
_
_
_
_
_
School
I n a C h i n ese-speaking cou ntry
Other
P lease specify
1
2
3
_
_
_
_
_
_
_
_
_
_
How lo n g ?
< o n e year
o n e year
two years
t h ree years
fo u r years
five years
> five years
0
1
2
3
4
5
6
264
6.
1
2
3
4
5
_
_
_
_
_
_
_
_
_
_
7.
Have you learnt any othe r languag e s apart from yo u r m oth e r tongue and
Chinese?
1
Yes
No
2
If so , which?
Fre nch
1
Ge rman
2
J apanese
3
Maori
4
Italian
5
Latin
6
Other
8
P lease specify
265
8.
How i m po rtant to you are the followi n g reasons for learn i n g Chi n ese?
Please respond to each possi ble reaso n .
I am studying Chi nese
(a)
(b)
5
4
3
2
1
5
4
3
2
1
5
4
3
2
1
(c)
(d)
(e)
266
(f)
(g)
9.
because it wi ll
e m ploy m e nt.
be
usefu l
fo r
my
p resent
extremely i mportant
very i mportant
i mportant
not so i m portant
not i mportant at all
5
4
3
2
1
5
4
3
2
1
(o r
futu re)
(a)
(b)
How i mport ant is it fo r you to co nti nue to study Chin ese once you
have fi nished this language paper?
extrem e ly i mportant
5
4
very i mpo rtant
3
important
not so i mportant
2
not i m po rtant at all
1
,
\\
4
MASSEV
Facsimile
U N I V ERS ITY
APPENDIX C
LETTER TO QU ESTION NAI RE SUBJECTS
Private Bag
Palmerston North
New Zealand
Telephone 0-6-356 9099
-
0r l y .
You rs sincere ly ,
Cynthia J . Wh ite
Lecture r
Li nguistics a n d Second Lang uage Teaching Secti o n
Departme nt o f Modern Languages
268
APPENDIX D
SAMPLE YOKED S U BJ ECT I NSTRU CTIONS: DISTANCE LEARN E RS
LEARNING JAPANESE
I magi n e that you have been asked this questi on by someon e who is p lan n i ng to
take a 200-leve l J apanese cou rse. They're i nte rested in what you do when you
study Japanese e xtramurally. They u nderstand about study g u ides, assig n m e nts
and so on - they want to know about yo u r particular strategies fo r learni ng .
I m agine that they are with you , so you can refe r to the section of the wo rkbook
to give them actual exam ples of how you go about things. R e m e m be r they aren't
aski ng you to actually do the exercises - they want you to g ive t h e m ideas about
you r ways of learni ng J apanese .
269
LEAR N I N G FRENCH
270
APPENDIX E
SAMPLE TRANSCRIPT O F YOKED SU BJ ECT V E R B A L R E PORT
DISTANCE LEARNER OF J APANESE
With e ach wo rkbook, which is laid out i n a simi lar patte rn to all the others , the
fi rst thing I do is to learn the kanji for the lesso n. I might n ot at that stag e learn
t h e m perfectly , but I go through e ach of the 28 kanji . I read each o n e , I draw it
i n my m i n d , I d raw it without a p e n in my hand but usi ng my fi ngers and the n I
actu ally draw it o n paper. W h e n I fee l that I am more fam i liar with its shape I
draw the kanji with all its diffe rent mean i ng s and o nce I have done t hat several
times, whe n I am co nfident that I can read any of those kanji at random and
know at least o n e of its meani n g s , its basic meaning , I turn to the vocabu lary list
of that lesso n and I learn the vocabulary. Once ag ai n , I m i g ht not h ave the time
to learn the vocabu lary abso lutely pe rfectly so that if I look to the E n g lish words
I cou ld not say what all of t h e m m eant i n Japanese but I do spe nd e noug h time
o n the vocabu lary list to be almost at that stag e , one or two words might slip. So
far that is, learn the kanji first , learn the vocabulary seco nd and at that poi nt I feel
I am ready to turn to the wo rkbook key e xamples and I read them and listen to
t h e m on the tape at the same ti m e . I use my walkman , I fi nd that is essenti al,
and with the key e xamples that are given with each u nit I check what they mean.
If it i s n ot i m mediately obvious to me what they mean the n I go over them u ntil
I can see as far as possi ble how each word of the J apanese se ntence relates to
the E ng lish se ntence and while I am doi ng that I listen to the se nte nce as we l l ,
I play it back seve ral times a n d once I have wo rked o u t what the se ntence
means, someti m e s I might u nd e rstand i m m ediately what it m e ans and sometimes
it might take a while, I liste n to the se nte nce and I repeat it and I repeat it u ntil
I can say it without looki ng at it. That means that I can d evelop a Japanese
i nto n ation and a deg ree of fl ue ncy. It's n ot real flue ncy because obviously I can't
produce any se ntence that I want to , but it certai n ly helps. lt forms a very solid
basis fo r produci ng sentences t h at I pe rsonally want to p roduce.
While
am
271
i n the marg i n . lt is either a question m ark to say I don't u nd e rstand this or I
specify what it i s I don't u nde rstand and I don't let it bother m e too m uch u n less
it affects m y u nderstanding of the whole passag e . If I fou nd that it did then I
wou ld have t o write to Massey U nive rsity and say what does this mean. I find
that soo n e r or late r the problem that I h ave got in that place is cleared up
because I see m any othe r structu res that a re similar. So , I think that the re is a
tendency a m o n gst fi rst time learn e rs of a foreign lang uage to be very
discourag ed because they don't u n derstand eve rything that is p re se nted to them
and I thi n k t h at if you have bee n t h rough somethi ng seve ral times and there is
just a little t i n y bit that you don't u nderstand then it is quite a l rig ht to just leave
it alone and you wi l l probably understand it three workbooks later because it will
become a lot cleare r. If I were a real ly good student, and last year I was , I wou ld
g o through t h e vocabu lary list at the end of a u nit and I wou ld also look at my key
e xamples i n E n g lish and I wou ld t ry to put them i nto Japanese. I did a lot of that
last year fo r t h e 1 00- level stag e and I haven't had time fo r that this year but I
know it would be a good idea if I did h ave the time. So to su m u p really I think
the key things a re fi rst of all to p re pare y o u rself for lesso n by learn i n g the kanji
and learning t h e vocabulary for that l esso n , then by seei ng the vocabu lary in
se ntences you become more fami liar with each wo rd so you learn its appl ication
better. By repeating as many times as possi ble se nte nces and phrases in
J apanese you d o develop flue ncy . You develop a feeling fo r the way a se ntence
is structu red and that i n turn helps you r readi ng.
272
APPENDIX E
SAMPLE TRANSCR I PT OF YOKED SU BJ ECT V E R B A L R EPORT
DISTANCE LEAR NER OF FRENCH
I wo rk systematically through the study gui de from the beg i n n i n g rig ht to the end.
T h e fi rst t h i n g I do is read e ach section carefully. I spend a long t i m e wo rki ng
with the dictio nary looki ng up all the words t h at I don't know , and there's a lot
of them. I find that if I do that first it frees me up m e ntal ly to concentrate on the
actual comprehension of the mate ri al and to t h i n k about answe ri ng the questions.
If I have to stop and look u p a lot of wo rds I fo rget what i t is that I am su pposed
to be doi n g . So that's the fi rst step. By doi ng that it also g ives me an idea of the
range of the who le u nit. lt also shows me any p h rases o r ideas that I can use i n
o n e sectio n , o r which I can t ransfe r t o another o n e . I f I d o n 't do that then I don't
h ave the fu l l pictu re and it means that I do n 't h ave t h e vocabulary and the
backg round to answer the questio n s as fu lly as I fee l I sho uld be able to. So the
fi rst thi ng to do is to read the whole lot throug h with a dictio nary without actually
d o i n g any of the tasks. Then sometimes I use diffe re nt methods. Sometimes I
start back at the beg i n n i ng agai n , worki ng t h roug h each secti o n , read i ng the
backg round i nformation and completi ng the sectio n of the assig n m e nt that re lates
to that u n it, for example g ram mar. Someti mes if I ' m p ushed for ti m e I do the
section that appeals to m e fi rst because it g ives me a psychological boost. I find
the g rammar section the easi est. Because I'm t rai ned in other lang uages in the
o ld fashio n ed method I fee l most co mfo rtable l earn i n g g rammar and applying the
ru les. lt gives m e a se n se of safety. If you do it carefully you 've g ot a fai r chance
of w i n n i ng . So I do the g ram mar section to give me a boost someti mes. I leave
t h e hardest sections till last. I find the essay writing the most p roblematic, so I
wait. I let those sections stew and I do them almost subco n sciously. I have them
ticki ng ove r g e ntly fo r most of the fou r weeks that I h ave to co mplete the
assi g n me nt . When I do the i n dividual sections I do t h e m fro m the beg i n n i ng to
the e nd. I don't jump around and do the bits that look easiest fi rst. As far as
revision i s co ncerned I don't g o back systematically and revise each sing le u nit
273
that we've d o n e half a dozen times d u ring the course of the year. I go back and
revise those sections of previous u n its which are re levant to what I ' m doing at the
moment - fo r example g ram mar, li n ki ng words, ce rtai n sectio n s of vocabul ary
which prove usefu l , but I don't g o b ack and learn the vocab off by heart. I t ry and
re member it each time. When I ' m actu ally wo rki ng with the m aterial I don't make
too many e xtra notes other than looki n g up the vocabu lary. S o m etimes if there's
a cross refe rence which can be u sefu l in other sections I go back and find it
ag ai n and u nderline it i n handouts but I do n 't actually w rite out notes. I do make
notes fo r g ram mar. I try to co mpile an exerci se book of g ra m m ar ru les because
I fi nd it's h e lpfu l to write those down i n my own way , to add to the notes in the
g rammar books, so meti mes to rearrange the o rder, to give othe r examples, to
u nderli n e t hose things i n my own writing to rei nforce them i n my m i nd visually.
I learn visually better than I learn oral ly. If I read a sentence I can re member it,
If I hear it I forget it. I have a co lleague who is doing stag e t h ree French. We
practice speaki ng French but I d o n 't read Fre nch se nte nces aloud to myse lf. I
have bits of pape r all over the house label ling the objects t h at I ' m su pposed to
know the vocabulary for. I write out key phrases and g ram matical se ntences and
idiomatic phrases which I want to use or i nclude in essays and I have the m
dotted all ove r the place where I can find them u nexpectedly. The elem ent of
surprise h e lps me to remember. I h ave fou nd it most effective . Another way of
learning vocab which I've always found good is to put the wo rds tog ether i n a
series of p ictu res. I do n 't tran slate Fre nch se nte nces i nto E ng lish because it
i nte rru pts t he thought process. O n e of the things I like about this particular
Fre nch cou rse is that it doesn't deal i n di rect translatio n which is the method I 've
always used previously, i n previous courses. Wh en I was at school we learnt
French that way and co nsequ e ntly I didn't know very much Fre nch at the end of
it. This year I've fou nd I've retai n ed a lot more because I've fou nd we've actually
had to use the wo rds in co ntexts to co mplete certai n tasks and that is much more
effective. I tend not to repeat wo rd s to re meber them. As I said I have a visual
me mory and I write them down and stick them all ove r the p lace rather than
saying t h e m ove r to myself. What do I do if I g et stuck? Try to d o the easy things
fi rst to build up co nfide nce . I d o n 't real ly g et stuck ve ry ofte n . So meti mes it's
274
difficu lt to i m ag i n e what a lectu re r's i nte rpretation of a questi o n might be and you
worry whether your i nterpretation is t h e same . For example it's difficu lt to pick out
e xactly what the key ideas are so m etimes. Do I encou rage myself to kee p g o i ng ?
I fi nd t h e longer I carry o n t h e less liable I a m t o stop because all the previous
wo rk that I've done wi ll count fo r nothing. Sheer panic and te rro r kee ps me going
probably - I 've got a li mited ti me to d o something. From a positive point of view,
sheer e scapis m . The vai n hope that one day all this wi ll mean g etting out of
l nvercarg i l l and g oi ng to see the big wide wo rld.
275
B I B LI OGRAPHY
Alexander, P.A. and J . E . Judy ( 1 9 89) The i nte ractio n of d o m ai n-specific and
strategic knowledge in academic perfo rmance. Review of Educational
Research 58 :375-404.
Ande rso n , J . R. ( 1 983) The architecture of cognition. Cam b ridg e , Mass. , Harvard
U nive rsity P ress.
Ande rso n , J . R. ( 1 985) Cognitive psychology and its implications. 2nd ed. N ew
York, Freeman.
Bake r, L. and A.L. B rown ( 1 984) Metacognitive ski l ls and read i n g . In R. Barr,
M . L. Kami l and P. Mose nthal (eds) Handbook of reading research. New
York, Lo n g m an : 353-394.
Bates, A.W. (ed. ) ( 1 990) Media and technology in European distance education.
M i lto n Keynes, EADTU and the Ope n U nive rsity.
276
Bi alystock, E. ( 1 978) A theo retical model of second languag e learn i n g . Language
Learning 28 :69-83.
Bialystock, E. and M. Froh lich ( 1 978) Variables of classro o m achi eve ment i n
seco nd lang u age learning. Modern Language Journal 6 2 :327-335.
B rown , A.L. ( 1 978) K nowi ng whe n , where , and how to re member: A p roblem in
metacognitio n . In Resources in Education. Phoenix, AZ, O ryx P ress.
Research
and
instruction.
Washingto n ,
DC,
N ational
Reading
277
B rown , A. L. a n d AS. Pali ncsar ( 1 982) I nd uci ng strategic learning fro m texts by
means of i nfo rmed, self-co ntro l t rai n i ng . Topics in Learning and Learning
Disabilities 2 , 1 : 1 - 1 7.
Carmichael's manual of child psychology. Vol . 1 . New Yo rk, Wi ley :77- 1 66.
TESOL Quarterly
26 , 4 :629-664.
B rown , J . S. and R. R. Burton ( 1 978) Diag n ostic models fo r p roced u ral bugs in
basic m athe matical ski l ls. Cognitive Science 2 : 1 55-1 9 2 .
research
perspective.
Rochester,
New
York,
N ational
Reading
Calve rt, J. ( 1 989) Instructional desig n fo r di stance learning. In K.A. J o hnson and
L.J . F oa ( eds) Instructional design. New alternatives for effective education
278
Can ney, G . and P. Wi nograd ( 1 979) Schemata for reading and reading
Carre l l , P . ( 1 989) M etacognitive awa re ness and seco nd lang uag e readi ng.
Carre l l , P . L . , B . G . Pharis and J . C. Li berto ( 1 989) Metacog nitive strategy t rai ning
fo r ESL reading. TESOL Quarterly 23,4 :647-678.
The
Chamot , A. U . (1 987) The learn i ng strategies of ESL stude nts. In A . Wenden and
J . Rubin (eds) Learner strategies in language learning. E n g lewood Cliffs,
N ew J e rsey , P re ntice Hall :71 -83.
279
Chamot , A. U . and J . M . O' Mall ey ( 1 987) The cog n itive acade mic lang uag e
learning approach : A b ridg e to the m ai nst ream .
TESOL Quarterly
2 1 ,2 :227-249 .
strategies in foreign language instruction: the third year and final report.
Mclean , Va. , I nte rstate Research Associ ates.
C hesterfield , R. and K . B . C hesterfi e ld (1 985) Natu ral order in chi ldre n 's use of
second language learning strategies. Applied Linguistics 6 , 1 :45-59.
C lyde , A . , H . Crowther, W. Patchi ng , I. Putt and R . Store ( 1 983) How stude nts
use dist 'ince teac h i ng materials: an institutio nal study. Distance Education
4 , 1 :4-27.
Cohe n , A. D. ( 1 983) I nt rospectio n about seco nd lang uage learn ing. Studia Anglica
280
Cohe n , A . D. (1 984) Studying seco nd- language learni ng strateg i e s : How do we
g et t h e i nformation? Applied Linguistics 5 , 2 : 1 0 1 - 1 1 2 .
Co h e n , A . D . (1 99 1 ) Feedback o n writi ng. The use of verbal report. Studies in
Co h e n ,
A D.
Coh e n ,
A D.
Research insights for the classroom. Cam bridge, Cam b ridg e U niversity
P ress : 1 55- 1 77.
281
Dansereau , D . F. (1 978) The development of a learn ing strategies cu rriculum. In
H . F. O' N ei l , J r. (ed . ) Learning strategies. N ew York, Acad e m i c P re ss : 1 -29.
Dicki nso n , L. and D. Carve r (1 9 80) Learn i ng how to learn : Steps towards self
directi o n in foreign language learn i n g in schoo ls. English Language
Dodds, A. E . and J .A. Lawrence ( 1 983) Heu ristics fo r plan n i n g u n ive rsity study at
a distance. Distance Education 4 , 1 :40-52.
E h rman , M. and R. Oxfo rd (1 989) Effects of sex diffe rence , care e r choice , and
psychological type on adu lt lang u age learning strategies.
Language Journal 73 ,1 : 1 - 1 3.
Modern
282
Eisenberg , E . and T. Dowsett ( 1 990) Student d ro p-out from a d i stance education
p roject cou rse : a new method of analysis . Distance Education 1 1 ,2: 231 253.
Text 8 :295-325.
E ricsso n , K.A. and H .A. Simon ( 1 980) Ve rbal repo rts on d ata. Psychological
E ricsso n , K.A. and H . A. Simon ( 1 984) Protocol analysis. Cambridg e , MA, MIT
P ress/Bradfo rd .
E ricsso n , K.A. and H .A. Simon ( 1 987) Ve rbal reports on thinki n g . In C. Faerch
and G. Kaspe r (eds) Introspection in second language research. Clevedon,
U . K. , M u ltiling ual Matte rs :5-23.
Faerch ,
C.
and
G.
Kaspe r
(eds)
( 1 983)
Strategies
in
interlanguage
283
Fae rch , C. and G. Kaspe r (1 987) Fro m p roduct to process- introspective m ethods
in seco nd language research. In C . Fae rch and G. Kaspe r (eds)
Behaviour 2 : 1 21 - 1 35.
Gag ne, E . D . ( 1 985) The cognitive psychology of school /earning. Bosto n , Mass . ,
Little, B rown.
Gal loway, V. and A. Labarca (1 990) Fro m stude nt to learner: Style , p rocess, and
strategy. I n D .W. Bi rckbichler (ed.) New perspectives and new directions
Gard n e r, R.C. ( 1 985) Social psychology and second language learning: The role
284
G ardn e r, R.C . , R . N . Lalonde and R. Moorcroft ( 1 985) The ro le of attitudes and
motivation in seco nd lang u ag e learn i n g : Correlatio n al and experi mental
considerations. Language Learning 35,2 :207-227.
1:
G ardn e r, R . C . and P.D. Macl ntyre ( 1 993) A stude nt's contri bution to
second-language learni ng . Part
11:
Teaching 2 6 , 1 :1 - 1 1 .
Garner, R. ( 1 988a) Verbal- report data o n cog nitive and metacog n itive strateg ies.
I n C . E . Wei n stein, E .T. Goetz and P.A. Alexande r (eds) Learning and
Garner, R., S . Wag o ner and T. S mith ( 1 983) Externalisi ng questi o n -an swe ri ng
strategies of good and poor co mpre h enders . Reading Research Quarterly
1 8 :439-447.
Garne r, R. and P .A. Alexander (1 989) Metacog nition : Answered and u nanswered
questions. Educational Psychologist 2 4 : 1 43- 1 58.
285
Hallgarte n , K. and B. Rostwo rowska ( 1 985) (Pi lot ed n . ) Learning for autonomy
learner training materials for ESL and literacy groups in adult education.
Londo n , ALBS U , I ndependent Learning P roject .
Education 6 , 1 :79-90.
Hose nfeld, C . ( 1 976) Learn i ng about learn i ng : Discoveri ng our stude nts'
strateg ies. Foreign Language Annals 9 : 1 1 7- 1 29.
Hosenfeld, C . ( 1 977) A pre l i m i n ary i nvestig atio n of the read i ng strateg ies of
successfu l and non-successfu l lang uage learners. System 5 : 1 1 0-1 23.
286
Hosenfeld , C. (1 984). Case studies of ni nth g rade readers. In J . C. Alders o n and
A . H. U rq u hart (eds) Reading in a foreign language. Lo ndo n , Lo n g m an : 23 1 249.
H uang , X-H . and M. van Nae rssen (1 987) Learning strategies fo r o ral
co m mu nication. Applied Linguistics 8 ,3 :287-307.
J acobs, J. and S . G . Paris ( 1 9 87) Chi ldren's m etacog nition about readi ng : Issues
i n d efi n iti o n , meas u re me nt and instructio n . Educational Psychologist
22 :255-278.
Kai l , R.V. J r. and J. Bisanz (1 982) Cog nitive strateg ies. In C . R . Puff (ed.)
287
Knig ht, G . ( 1 987) Distance education i n mathematics. Studies in Mathematics
Lo ng, M.H. ( 1 983) Inside the black box: Methodo log ical issues i n class roo m
research on lang uage learn i n g . I n H .W. Se lige r and M. Lo ng (eds)
Mang u bhai , F. (1 991 ) The p rocessing behaviou rs of adult seco nd-lang uage
learne rs and thei r relationship to second- languag e proficiency. Applied
Linguistics 1 2 ,3:268-298.
Marland, P., W. Patchi ng , I. Putt and R. Sto re ( 1 984) Learn i n g fro m distance
teachi ng materials : a study of stude nts' mediati n g responses. Distance
288
Marland, P . , W. Patchi ng , I. P utt and R. Putt (1 990) Distance learners'
i nte ractio n s with text while studyi n g . Distance Education 1 1 , 1 :71 -91 .
Mayer, R.E. ( 1 988) Learning strateg i e s : an ove rview. In C . E . Wei nstein, E.T.
Goetz and P.A. Alexande r (eds) Learning and study strategies. New York,
Acad e m i c P ress : 1 1 -22.
Japanese (CLEAR Tech nical Report No 1 0). Los Angeles, U nive rsity of
Califo rni a, Ce nter for Lang uage Education and Researc h .
McG roarty , M . a n d R. Oxfo rd ( 1 990) An i ntroduction and two re lated studi es. I n
A . M . Podilla, H . M. Fai rchild and C . M . Valadez (eds) Foreign language
289
McLaug h l i n , B. (1 978) Th e monito r mode l : Some methodolog ical co nside rati ons.
Miyacke , N. and D.A. No rman ( 1 979) To ask a questio n , one must know e noug h
to know what is n ot known. Journal of Verbal Learning and Verbal
Behaviour 1 8 :357-364.
Mosenthal, J . H . ( 1 987) Learn i ng from discussi o n : Requireme nts and co nstrai nts
o n class room i n st ruction in reading com pre hensi o n strateg ies. In J . E .
Reade nce and
R . S.
Baldwi n
(eds)
290
Myers , M. and S . G . Paris ( 1 978) C hi ld re n's metacog nitive knowledge about
read i n g . Journal of Educational Psychology 70 :680-690.
Nai m a n , N . , M. Fro h lich , H . H . Ste rn and A. Todesco ( 1 978) The good language
missionaries. Published by F riendship P ress for the N atio nal Cou nci l of
C h u rches in the U SA.
N isbett , R . E . and T.D. Wilson ( 1 977) Te lling more than w e can know: Verbal
reports on mental p rocesses. Psychological Review 8 4 :231 -259.
291
O'Malley, J . M . , A.U. Chamot, G . Stewner-Manzanares, L. KOppe r, and R . P.
Russo ( 1 985a) Learning strategies used by beg i n n i ng and i nte rmedi ate
E S L stude nts. Language Learning 35 , 1 : 2 1 -46.
Oxfo rd , R. ( 1 989) Use of language learn i ng strateg ies: A synthesis of studies with
i mplications fo r strategy t rai n i n g . System 1 7,2 :235-247.
292
Oxfo rd , R. ( 1 990) Language learning strategies: What every teacher should
Oxfo rd , R . and
D.
Oxford, R . and M. Nyi kos (1 989) Variables affecti ng choice o f language learni ng
strategies by u nive rsity students. Modern Language Journa/ 73,3 :29 1 -298.
Pali ncsar,
A.S.
and
A.L.
B rown
(1 984)
Reci procal
teaching
of
P aris, S . G . ( 1 988) Fusi ng ski l l and wi l l : the i nteg ration of cog n itive and
m otivational psychology. Pape r prese nted at the a n n ual meeti ng of the
Ame rican Educational Research Associati o n , N ew Orleans, April 1 988.
Pei, M. ( 1 966) How to learn languages and what languages to learn. New York,
Harper and Row.
293
Perki ns, D . N . and G. Salomon ( 1 989) Are cog n itive ski lls co ntext-bound?
Acquisition 6 , 1 :54-68.
Politzer, R . (u ndated) Motivatio n , lang uag e learni ng behaviou r, and ach i evement
i n an i ntensive E S L course. U n published man uscri pt . Stanfo rd U niversity ,
Stanford , CA.
Ram i rez, A. G . (1 986) Lang uage learning strategies used by ado lescents studying
F re nch in N ew Yo rk schools. Foreign Language Annals 1 9 , 2 : 1 31 - 1 4 1 .
294
Ram say , R . M . G . ( 1 9 80) Language-learn i n g approach styles of adult m u ltili nguals
and successful language learners. A nnals of the New York Academy of
Sciences 75:73-96.
in
Research 50 :5-53.
Rencher, A. C. ( 1 992) I nterpretatio n of can o nical discrimi nant functi o n s , cano nical
variates and princi pal co m po n e nts. The American Statistician 46,3 :21 7225.
Sciences 1 7: 1 2-22.
Rowntree ,
D.
Rubi n , J . ( 1 9 75) What the 'good lang u age learner' can teac h us. TESOL
Applied Linguistics 1 1 , 2 : 1 1 7- 1 3 1 .
Rubi n , J . ( 1 987) Learner strategies: theo retical assu mptio n s , researc h history and
typology. In A. Wenden and J . Rubin (eds) Learner strategies in language
295
Rubi n , J. and I. Tho m pso n ( 1 982) How to be a more successful language
Sch u man n , J . H . ( 1 976) Seco nd languag e acquisitio n : G etti ng a more g lobal look
at the l earne r. Language Learning, Speci al Issue No. 4 : 1 5-28.
Schu man n , J . H . ( 1 978) The accu ltu ration model for seco nd lang uage acquisiti o n .
I n R . C . G i n g ras (ed . ) Second language acquisition and foreign language
Seli nke r, L. and J . Lam e nde lla ( 1 976) Two perspectives o n fossilization i n
i nte rlang uage learn i n g . lnterlanguage Studies Bulletin 3 : 1 44- 1 91 .
Skehan , P. ( 1 986) Cluste r analysis and the ide ntificatio n of learn e r types. I n V.
Cook (ed.) Experimental approaches to second language acquisition. New
York, Macm i llan :81 -94.
296
S parks , E . E . (1 988) Carl : A case study of executive co ntro l in a g ifted adolesce nt
read e r. In J . E . Reade nce , R. Scott Baldwi n , J . B . Kanopak and P . R.
O' Keefe (eds) Dialogues in literacy research. C hicag o , I L , The National
Read i n g Confe rence , l nc. :208-2 1 7.
Ste rn , H . H . ( 1 975) W hat can we learn fro m the good language learner? Canadian
Svan e s , B.
Sweet, H . ( 1 899) The practical study of language: A guide for teachers and
Taro n e , E. (1 983) Some thoug hts on the notion of com m u nication strategy. I n C .
Faerch and G . Kaspe r (eds.) Strategies in interlanguage communication.
New York, Lo ng man : 6 1 -74.
297
Thompso n , G. and A . Knox ( 1 987) Desig n i n g for dive rsity : are fie ld-dependent
learne rs less suited to distance educatio n prog ram s of i nstruction?
Tran , T.V. (1 988) Sex diffe rences i n E ng lish language accu ltu ration and
learning strategies among Vietnamese adu lts aged 4 0 and over i n
the U nited States. Sex Roles 1 9 :747-58.
Tyacke , M. and D . Mende lsoh n (1 986) Stude nt n eeds : Cog nitive as well as
co m m u n icative . TESOL Canada Journal S peci al Issue 1 : 1 7 1 - 1 83.
Wag ner, E . D. ( 1 990) Instructi o nal desi g n and deve lo p me nt : Continge ncy
manag e m ent for distance educatio n . In M . G . Moore (ed . ) Contemporary
Wag n e r, R . K . and R.J. Ste rnberg ( 1 9 85) Practical intellige nce i n real-world
pursu its : The role of tacit knowledge. Journal of Personality and Social
Psychology 49 :436-458.
Watki ns, D. ( 1 983) Depth of p rocessing and the quality of learn i ng outcomes.
Watki n s ,
D.
298
Wei n ste i n , C . E . and V . L. U nd e rwood ( 1 983) Learn i ng strategies: the how of
l earni n g . I n J .W. Seg al , S . F. Chipman and R. G laser (eds) Relating
Wei nste i n , C . E . and R.E. May e r ( 1 986) The teachi ng of learning strategies. I n
M . C. Wittrock (ed . ) Handbook o f research o n teaching. 3 rd ed. New York,
Macmi l lan :31 5-327.
Wenden, A. (1 986a) What d o L2 learners know about the i r lang uage learning?
A seco nd look at retrospective accou nts. Applied Linguistics 7,2 : 1 86-205.
Wenden, A. ( 1 986b) Helping lang uag e learne rs think about their learning. EL T
Journal 40 , 1 :3- 1 2.
Wende n , A. ( 1 987a) Conceptual backg rou nd and uti lity . I n A. Wenden and J .
Rubi n (eds) Learner strategies in language learning. E ng lewood Cliffs, NJ ,
P re ntice-Hal l :3- 1 3.
Wende n , A. ( 1 987b) Metacog nition : An e xpanded view on the cog nitive abilities
of L2 learne rs. Language Learning 37,4 :573-597.
We rtsch , J.V. ( 1 977) Metacog nition and adu lt-child i nte ractio n . I n Resources in
299
W hite , C.J . ( 1 992) A guide to strategies for language learners. Massey
U n ive rsity, CUES.
language ability and language behaviour. New York, Acad e m ic P ress :203228.