Anda di halaman 1dari 26

Perceptual & Motor Skills: Perception

2015, 120, 2, 576-590. Perceptual & Motor Skills 2015

EFFECTS OF AUDIO-VISUAL AIDS ON FOREIGN LANGUAGE TEST


ANXIETY, READING AND LISTENING COMPREHENSION, AND
RETENTION IN EFL LEARNERS1, 2
SHU-PING LEE

SHIN-DA LEE

Department of Foreign Languages and


Literature, Asia University, Taichung, Taiwan

Department of Physical Therapy


Graduate Institute of Rehabilitation Science,
China Medical University, Taichung, Taiwan

YUAN-LIN LIAO AND AN-CHI WANG


Department of Foreign Languages and Literature, Asia University, Taichung, Taiwan
Summary.This study examined the eects of audio-visual aids on anxiety,
comprehension test scores, and retention in reading and listening to short stories
in English as a Foreign Language (EFL) classrooms. Reading and listening tests,
general and test anxiety, and retention were measured in English-major college students in an experimental group with audio-visual aids (n = 83) and a control group
without audio-visual aids (n = 94) with similar general English proficiency. Lower
reading test anxiety, unchanged reading comprehension scores, and better reading
short-term and long-term retention after four weeks were evident in the audiovisual group relative to the control group. In addition, lower listening test anxiety,
higher listening comprehension scores, and unchanged short-term and long-term
retention were found in the audiovisual group relative to the control group after the
intervention. Audio-visual aids may help to reduce EFL learners' listening test anxiety and enhance their listening comprehension scores without facilitating retention
of such materials. Although audio-visual aids did not increase reading comprehension scores, they helped reduce EFL learners' reading test anxiety and facilitated
retention of reading materials.

Traditionally, the use of visual aids on the overhead projector has


been proved to facilitate foreign language listening. Mueller's study (1980)
showed that visual contextual cues on the overhead projector after listening have positive eects (Cohen's d = 0.63) on foreign language listening
comprehension. Because of the advances in technology and the emphasis
on communicative functions in foreign language teaching, the application
of audio-visual aids in foreign language classes has attracted practitioners attention (Garrett, 1989; Carol, 1994; Herron, York, Cole, & Linden,
1998; Canning-Wilson, 2000; akir, 2006). Utilizing authentic videos is beAddress correspondence to Shu-Ping Lee, Department of Foreign Languages and Literature, Asia University, WuFeng, Taichung, Taiwan or e-mail (shuping@asia.edu.tw).
2
The study was supported by Grants 102-asia-18 and CMU99-ASIA-03 from Asia University
and China Medical University. We would like to thank Dr. Tim Williams for proofreading the
manuscript. Also, many thanks are due to the reviewers, whose critical comments not only
helped us with the revision but also sharpened our knowledge in related fields.
1

DOI 10.2466/24.PMS.120v14x2

16-PMS_Lee_140120.indd 576

ISSN 0031-5125

15/04/15 6:36 PM

AUDIO-VISUAL AIDS, EFL ANXIETY

577

lieved to help facilitate, as well as stimulate, the understanding of both


foreign language and culture because authentic videos help create real
world-oriented scenarios in the classroom (Secules, Herron, & Tomasello,
1992; Herron, Morris, Secules, & Curtis, 1995; Wagner, 2008). Herron &
Hanley (1992) showed that videos can ease foreign language comprehension and enhance retention (Cohen's d = 0.45) because of the student-oriented contents. Many related studies have focused on the eectiveness of
videos in foreign language listening (Wagner, 2008). Carol's study (1994)
suggested that the use of video texts enhances students receptive skills.
Visual cues were also found to improve learners general comprehension
(Baltova, 1994). Herron, Hanley, and Cole (1995) indicated that video instruction with pictures can help facilitate foreign language retention and
extensive listening. Rubin (1990, 1994) also emphasized the appropriate
selection of videos and showed that watching dramas on video improved
students foreign language listening comprehension compared with students who were not exposed to videos for their listening training (Cohen's
d = 0.4). Arthur (1999) indicated that videos not only oer a visual reinforcement of a foreign language but also lower the listener's anxiety.
Most studies of audio-visual aids have neglected their eects on reading comprehension. Both reading and listening comprehension are complex processes that involve factors such as linguistic knowledge or cognitive processing skills (Omaggio, 1979). Cognitive processing includes
interpreting information from visual and auditory systems. Reading and
listening are interpretive skills in which readers and listeners actively construct understanding from information input. In order to receive benefits
for further language use, the relevant schemata formed from received information must be retained long enough for eective cognitive re-activation. The visual cues in audio-visual aids for foreign language reading
and listening comprehension (Canning-Wilson, 2000) could help retain received information long enough for further use.
Many studies have investigated the role of input in second language
acquisition (Ellis, 1981; Sharwood, 1993; Mayer, 2003). Mayer's study (2003)
pointed out that input processes such as reading and listening begin with
the information entering the eyes and ears and being briefly stored in the
visual and auditory neural systems. Audio-visual aids for eective learning
were suggested by Mayer's cognitive theory of multimedia learning (2003),
which indicated that input was divided into two channels: words and pictures. The words could be printed or spoken, whereas the pictures could be
static, such as photos or maps, or dynamic, such as videos. Prior to Mayer's study, Snyder and Golon's study (1988) showed that the group with audio-visual aids produced significantly better listening comprehension test
performance (r = .13). Secules, et al. (1992) reported that French classes using

16-PMS_Lee_140120.indd 577

15/04/15 6:36 PM

578

S-P. LEE, ET AL.

videotapes had significantly higher mean listening comprehension scores


(Cohen's d = 0.85). Although Secules, et al. (1992) indicated that learners with
video input had higher listening comprehension scores, no significant difference was found on writing ability or reading comprehension. In order to
gain empirical data concerning the eects of divergent input (printed or spoken words and dynamic pictures) on reading and listening comprehension
based on Mayer's hypothesis (2003) of eective learning by using audio-visual aids as input, the present study compared the eectiveness of foreign
language interpretive skills by giving one group audio-visual aids (selected
films from YouTube) while the other group did not have these aids.
Audio-visual aids also have been suggested to reduce listening anxiety (Snyder & Golon, 1988; Secules, et al., 1992; Arthur, 1999). Anxiety
is a prevalent phenomenon connected to foreign language learning (MacIntyre & Gardner, 1994a), and comprises trait anxiety, state anxiety (Spielberger, Gorsuch, & Lushene, 1970; Spielberger, 1983), and situation-specific anxiety (MacIntyre & Gardner, 1994a, 1994b). Although Kleinmann
(1977) noted that foreign language anxiety might not be necessarily harmful, anxiety has been shown to have a negative eect on foreign language
learning (Horwitz, Horwitz, & Cope, 1986; Young, 1986, 1992; MacIntyre
& Gardner, 1991; Phillips, 1992). Anxiety could also aect the functions of
working memory and attention, leading to lower academic performance
(Aronen, Vuontela, Steenar, Salmi, & Carlson, 2005). Experiencing learning anxiety could become an aective filter (Krashen, 1985) that prevents
certain information from being retained for processing. Interpretive anxiety could obstruct information processing and hence impede learning
(Chastain, 1975; MacIntyre & Gardner, 1989).
Horwitz, et al. (1986) developed the Foreign Language Classroom
Anxiety Scale (FLCAS) to measure foreign language anxiety as a situationspecific anxiety comprising communication apprehension, fear of negative
evaluation, and test anxiety. Foreign language reading anxiety is claimed
to be related to but distinct from foreign language anxiety (Saito, Horwitz,
& Garza, 1999; Sellers, 2000; Ahmad, Al-Shboul, Nordin, Rahman, Burhan,
& Madarsha, 2013); the sources of foreign language reading anxiety have
been investigated (Bekta-etinkaya, 2011; Ahmad, et al., 2013). Saito, et
al. (1999) developed a Foreign Language Reading Anxiety Scale (FLRAS),
which elicited students' self-reports of anxiety about various aspects of
reading and their perceptions of reading diculties in their foreign language classes. Their study showed that foreign language reading anxiety
was negatively correlated with grades in foreign language courses, consistent with Sellers (2000), who indicated that foreign language reading anxiety influenced readers concentration on a reading task and their comprehension of the reading passage. Tsai and Li (2012) reported that foreign

16-PMS_Lee_140120.indd 578

15/04/15 6:36 PM

AUDIO-VISUAL AIDS, EFL ANXIETY

579

language reading proficiency was negatively related to test anxiety and


foreign language reading anxiety (r = .41, p < .01).
With regard to listening anxiety, Elkhafaifi's (2005a) study of Arabic
in a foreign language classroom situation indicated that foreign language
anxiety and foreign language listening anxiety were separate but related
phenomena. Elkhafaifi discovered that higher listening anxiety resulted in
lower listening comprehension grades (r = .65, p < .01). Mills, Pajares, and
Herron (2006) also found that listening anxiety was negatively related to
listening proficiency (r = .34, p < .01). Zhang (2013) applied structural equation modeling to investigate the possible causal relationship between foreign language listening anxiety and listening performance; foreign language
listening anxiety could aect foreign language listening performance, with
a significant coecient of .20 in Test 1 and .03 in Test 2 (p < .01). However, computer-based immediate feedback in English listening comprehension tests did not lower the debilitating eects of anxiety, but enhanced students intrapersonal eustress-like anxiety (Cohen's d = 0.63) and probably
improved their attention during listening tests (Cohen's d = 0.45) (Lee, Su, &
Lee, 2012). Furthermore, some studies have obtained evidence that learners'
listening anxiety varied according to their ability in foreign language listening (Elkhafaifi, 2005b; Mills, et al., 2006; Chang, 2008). To reduce students'
anxiety, Arnold (2000) used relaxation and visualization exercises as two
stages to investigate whether they could help advanced-level students deal
with their anxiety in listening comprehension tests. According to Arnold's
results, the students who had visualization-relaxation training achieved
better scores on listening comprehension (Cohen's d = 1.10) than students
who did not have training. However, in their study investigating how various forms of listening support aected the listening anxiety of EFL learners,
Chang and Read (2008) indicated that relaxation practice may reduce anxiety, but it may not be appropriate when taking a formal test.
Although some studies reported the use of audio-visual aids on foreign
language listening comprehension for facilitating retention (Herron & Hanley, 1992; Herron, et al., 1995) and reducing anxiety (Snyder & Golon, 1988;
Secules, et al., 1992; Arthur, 1999), the claims in favor of using audio-visual
aids to facilitate foreign language interpretive skills (reading and listening),
retention, and interpretive anxieties are not well supported. This study differentiated general foreign language anxiety from foreign language test anxiety and investigated the influence of audio-visual aids on test anxiety, test
scores, and retention in foreign language reading and listening.
Research goal 1. To assess whether audio-visual aids improve students' reading scores, reduce reading test anxiety, and reinforce
retention of information through reading.

16-PMS_Lee_140120.indd 579

15/04/15 6:36 PM

580

S-P. LEE, ET AL.

Research goal 2. To assess whether audio-visual aids improve students' listening scores and reduce listening test anxiety, as well
as reinforce short-term and long-term retention of information
through listening.
METHOD
Participants
A total of 177 English-major college students in a university in Taiwan
ranging in age from 20 to 22 years participated in this study (control 73 women, 21 men; experimental 63 women, 20 men). The participants included
sophomore (36 control, 27 experimental), junior (36 control, 33 experimental), and senior (22 control, 23 experimental) students who took English literary stories as a required course, which was taught by the same instructor.
There was no dierence in the learners general English proficiency between
the control and experimental groups on average grades in required English
courses (control M = 79.88, SD = 11.46; experimental M = 81.41, SD = 10.05;
t = 0.90, p = .37). Participant groups did not dier on sex (control 73 women, 21
men; experimental 63 women, 20 men) or age (control M = 20.6 yr., SD = 1.6;
experimental M = 20.6 yr., SD = 1.3; t = 0.08, p = .94). After being approved by
the Medical Research Ethics Committee of Asia University, the procedure
was described to all participants without informing them what kind of aids
would be provided in the content they were given. All participants signed a
written informed consent before their participation. The experimental group
was given audio-visual aids, and the control group was not.
Measures
All the variables measured by instruments in this study were self-report surveys, except that the reading and listening comprehension tests
were measured through multiple-choice content questions. The original
questionnaires were applied and back-translation methods were supplemented in translating survey items from English to Chinese except the
State-Trait Anxiety Inventory.
General anxiety and general foreign language anxiety.Refers to the anxiety level measured 2 weeks prior to the test and includes the STAIT (general anxiety), the FLCAS (general foreign language classroom anxiety), the
FLRAS (general foreign language reading anxiety), and the FLLAS (general foreign language listening anxiety).
Trait anxiety.The Trait Anxiety Scale (STAIT) was used from the
State-Trait Anxiety Inventory (STAI; Chung & Long, 1984; Chinese version
with permission). The STAIT consists of 20 items, each accompanied by
a 4-point scale with anchors 1 = Rarely and 4 = Always. Items 1, 6, 7, 10, 13,
16, and 19 were reverse scored. Scores could range from 20 to 80. Higher
scores indicate greater anxiety. Cronbach's in the current sample was .85.

16-PMS_Lee_140120.indd 580

15/04/15 6:36 PM

AUDIO-VISUAL AIDS, EFL ANXIETY

581

Foreign language classroom anxiety.The Foreign Language Classroom


Anxiety Scale (FLCAS) was designed by Horwitz, et al. (1986) to assess students anxiety in Spanish classes, and was adapted for English classes by
replacing the term foreign language used in the original FLCAS with
English. It consists of 33 statements, rated on a 5-point Likert scale with
anchors 1: Strongly disagree and 5: Strongly agree. Items 2, 5, 8, 11, 14, 18, 22,
25, 28, and 32 were reverse scored. Horwitz, et al. estimated the internal consistency reliability at .93 (n = 75) and test-retest reliability over eight weeks
at .83. The possible score range of the FLCAS is 33 to 165. Higher scores indicated higher foreign language classroom anxiety. Cronbach's in the current sample was .79.
Foreign language reading anxiety.The Foreign Language Reading
Anxiety Scale (FLRAS) was developed by Saito, et al. (1999). The words
French, Russian, Japanese in the original scale were replaced with the
word English in this study. The scale consists of 20 items; the participants recorded a response on a 5-point Likert scale with anchors 1: Strongly disagree and 5: Strongly agree. The theoretical range of scores is 20 to
100. Saito, et al. (1999) reported internal consistency at .86 (N = 383). Cronbach's in the current sample was .73.
Foreign language listening anxiety.The Foreign Language Listening
Anxiety Scale (FLLAS) was developed by Elkhafaifi (2005a). Elkhafaifi
modified the FLRAS for a survey of Arabic listening anxiety, in which the
word reading in the FLRAS was replaced with the word listening and
the words French, Russian, and Japanese were replaced by Arabic. In
the current study, the word Arabic was replaced with the word English. The possible range of scores is 20 to 100. Elkhafaifi reported internal consistency at .96 (N = 233). Cronbach's in the current sample was .71.
Reading and listening comprehension tests.The two texts that were used
in the study were selected by two literature instructors who had experience
in teaching the same English course and included The Unicorn in the Garden for the reading test and The Giving Tree for the listening test.3 Both texts
are intermediate level, the participants average proficiency. The criteria for

The former is a fable written by James Thurber (1940) about a husband and wife. The
husband wakes the wife up and tells her that he saw a unicorn in the garden eating roses.
However, the wife believes that her husband is crazy and therefore calls the police and the
psychiatrist. When they arrive, they ask the husband if he saw a unicorn in the garden. He
replies, The unicorn is a mythical beast. They take the wife away. The second text, by Shel
Silverstein (1964), tells a story about giving and receiving between a tree and a boy. In his
childhood, the boy likes to play on the tree. He comes to the tree to eat her apples, swing
from her branches, or slides down her trunk. The tree is happy to have the boy. When the
boy grows older, he wants more from the tree: he sells her apples for money, cuts down her
branches for a house, and cuts her trunk for a boat. The tree is nothing but a stump, which
provides the boy (now an older man) a place to rest, and the tree is happy.
3

16-PMS_Lee_140120.indd 581

15/04/15 6:36 PM

582

S-P. LEE, ET AL.

text selections were based on learners interests, level of diculty, minimal


requirements of background knowledge, and appropriate syntactic complexities. All students in required English courses were asked to take the
reading and listening tests, but only those who signed the written informed
consent were involved in the surveys and included in this study. For each
comprehension test, 15 multiple-choice content questions in English with
bottom-up processes of comprehension about the two stories were provided, with the total score being 100. Two suitable videos from YouTube.com,4
which show complete animated stories of these two texts, were selected
and evaluated by the instructors. The researchers overhauled the audiovisual aids to ensure there were no captions for the comprehension test
and to ensure that the dialogues followed the short story text word-forword for the listening comprehension test. The reading and listening comprehension tests took place in computer labs. Cronbach's s in this sample
were .92 and .93, respectively, for reading and listening.
State anxiety.Reading and listening state anxiety (State Anxiety Scale,
STAIS) was measured immediately after the reading and listening comprehension tests, respectively. Each item in the STAIS subscales was rated on a 4-point scale with anchors 1: Rarely and 4: Always. Items 1, 2, 5, 8,
10, 11, 15, 16, 19, and 20 were reverse scored. Scores could range from 20 to
80, and higher scores indicate greater anxiety.
Reading test anxiety and listening test anxiety.The test anxiety score was
defined as total score from the State Anxiety Scale (STAIS) minus the total score on the Trait Anxiety Scale (STAIT). This indicates increased anxiety during the reading or listening test: pre-test general trait anxiety (two
weeks prior to the test) compared to state anxiety for the reading or listening test. Cronbach's in this sample was .83 for reading, .83 for listening.
Short-term and long-term retention of information.Short-term retention
refers to a post-test retention recall and long-term retention refers to a 4-week
memory recall without prior notification. One open-ended question was designed for each post-test and four-week memory recall task of listening and
speaking. Students were asked to summarize the major characters, time, location, and plots of the stories in the reading and listening tests. The Unicorn
in the Garden includes five characters, one time, two locations, and six major
plots. The Giving Tree contains two characters, one time, one location, and
eight major plots. Scores for the short-term and long-term retention of information were calculated based on their correct descriptions of characters,
time, location, and story plots. Cronbach's in this sample was .79.
The video The Unicorn in the Garden was downloaded from http://www.youtube.com/
watch?v=1teJjX-smdE by Erose7en (2008, May 8; Classic UPA Cartoon). The video The Giving
Tree was downloaded from http://www.youtube.com/watch?v=9PaDSZfZFZc by K. Hamil &
T. Newman (2008, September 16).

16-PMS_Lee_140120.indd 582

15/04/15 6:36 PM

AUDIO-VISUAL AIDS, EFL ANXIETY

583

Procedure
At the beginning of the fall semester in 2010, the researchers went to
six classes to administer the survey at prearranged times and to explain to
the participants the nature and the purpose of this study. After the explanation, the researchers recruited volunteers, who signed a written informed
consent, and provided instructions on how to answer the questionnaires.
The questionnaires used to measure general anxiety and general foreign
language anxiety were the STAIT, the FLCAS, the FLRAS, and the FLLAS.
The participants were given as much time as required to complete the selfreport sections of the questionnaires. When all were finished, the volunteers were told which computer labs were to be used for the comprehension tests two weeks later.
In the reading comprehension test, each participant of the control
group (n = 94) read the printed text (The Unicorn in the Garden) for 30 min.,
while the experimental group (n = 83) read the printed text for 20 min. and
then was allowed to watch the video of the same story, The Unicorn in the
Garden, which took 10 min. After reading, all participants were asked to
answer 15 multiple-choice content questions within 10 min.; the printed
story text was available to both groups. After the reading test, the reading
state anxiety (STAIS for reading) was assessed.
In the listening comprehension test, the control group read the printed story text while they listened to the second story (The Giving Tree)
via audio (the audio track from the video watched by the experimental
group) for 10 min. The experimental group watched the video for 10 min.,
but without captions or the printed story text. After listening to the audio
(with text, control) or watching the video (without text, experimental), the
participants were asked to complete the 15 multiple-choice content questions within 10 min. The listening state anxiety (STAIS for listening) was
assessed immediately after the listening comprehension test.
The tests of short-term retention for information in the reading and listening tests were followed by the reading and listening state anxiety questionnaires. Tests of long-term retention for information in the reading and
listening tests were administered four weeks after exposure to the experimental material. The participants were required to recall the two stories.
Their descriptions should include the plot, main characters, time, and place.
Statistical Analysis
Sex ratio, year in school, age, and general English proficiency of the
control and experimental groups did not dier significantly. The participants were randomly assigned to the control and experimental groups.
Dierences in general anxiety (STAIT), general foreign language anxiety
(FLCAS, FLRAS, and FLLAS), state anxiety (STAIS) in the reading and listening comprehension tests, test anxiety scales in the reading and listening

16-PMS_Lee_140120.indd 583

15/04/15 6:36 PM

584

S-P. LEE, ET AL.

comprehension tests (STAIS minus STAIT), test scores in reading and listening comprehension, and short-term and long-term retention of information in reading and listening (all dependent variables) were analyzed by
Students t test to assess the eects of audio-visual aids on the experimental
group vs. the control group. In all cases, a dierence at p < .05 was considered statistically significant.
RESULTS
Table 1 shows comparisons of the two groups on general anxiety
and general foreign language anxiety assessments using the STAIT, the
FLCAS, the FLRAS, and the FLLAS. General anxiety mean scores (STAIT,
FLCAS, FLRAS, FLLAS) were similar in the two groups.
Mean reading test anxiety in the experimental group was significantly
lower than in the control group. Although reading comprehension scores
were similar, short-term and long-term retention scores in the reading test
in the experimental group were significantly higher than those of the control group. The audio-visual aids had a significant eect on reducing students' reading test anxiety, as well as short-term and long-term retention
of information, but were not beneficial in improving students' reading
comprehension.
In the listening test, listening test anxiety was significantly lower in
the experimental group. Also, the mean listening comprehension score in
the experimental group was significantly higher than that of the control
group. However, mean short-term and long-term retention scores for listening were not significantly dierent between the groups. Audio-visual
aids were found to be helpful in enhancing students' listening comprehension and reducing students' listening test anxiety, but were not beneficial
to short-term and long-term retention of information through listening.
DISCUSSION
Audio-visual aids helped to diminish reading test anxiety and improve short-term and long-term retention of information through reading,
but did not influence reading comprehension test scores assessed through
multiple-choice questions. Audio-visual aids greatly reduced students'
listening test anxiety and improved students' listening comprehension
test scores, but had no influence on retention of information through listening tasks.
On reading comprehension scores, there was no significant dierence
with audio-visual aids, consistent with Secules, et al. (1992), who found
no significant dierences in test scores between groups using video and
groups not using video in reading comprehension tasks. Both groups
were provided with the printed story text (The Unicorn in the Garden) during the reading comprehension test, so when students did not know the

16-PMS_Lee_140120.indd 584

15/04/15 6:36 PM

585

AUDIO-VISUAL AIDS, EFL ANXIETY

TABLE 1
CONTROL AND EXPERIMENTAL GROUPS GENERAL ANXIETY AND GENERAL FOREIGN
LANGUAGE ANXIETY; READING SCORES, READING TEST ANXIETY, SHORT-TERM RETENTION, AND
LONG-TERM RETENTION THROUGH READING; LISTENING SCORES, LISTENING TEST ANXIETY,
SHORT-TERM RETENTION, AND LONG-TERM RETENTION THROUGH LISTENING
Scale

Control
Group (n = 94)
M

SD

Experimental
Group (n = 83)
M

SD

Cohens
d

STAIT

43.98

8.50

45.11

6.86

0.96

.34

0.15

FLCAS

101.86

10.98

100.48

13.34

0.76

.45

0.11

FLRAS

59.59

8.17

59.86

8.55

0.22

.83

0.03

FLLAS

56.32

7.31

58.59

8.18

1.95

.05

0.30

47.84

10.29

45.23

8.90

1.79

.07

0.27

3.86

8.07

0.12

7.69

3.15

.002

0.48

Reading test
STAIS for reading
Reading test anxiety
Reading scores

80.11

17.69

81.85

12.64

0.75

.46

0.11

Short-term reading
retention scores

7.17

2.82

8.54

2.56

3.37

.001

0.51

Long-term reading
retention scores

6.21

3.15

7.94

2.39

4.07

< .001

0.62

47.93

9.81

45.13

9.42

1.93

.06

0.29

Listening test
STAIS for listening
Listening test anxiety

3.96

5.88

0.02

7.99

3.76

< .001

0.57

83.05

13.71

88.35

9.59

2.94

.004

0.45

Short-term listening
retention scores

6.59

2.54

6.80

2.44

0.56

.58

0.09

Long-term listening
retention scores

5.00

2.48

5.42

2.33

1.16

.25

0.18

Listening scores

Note.STAIT = Trait Anxiety Scale; FLCAS = Foreign Language Classroom Anxiety Scale;
FLRAS = Foreign Language Reading Anxiety Scale; FLLAS = Foreign Language Listening Anxiety Scale; STAIS = State Anxiety Scale; Reading test anxiety = STAIS minus STAIT for reading; Listening test anxiety = STAI-S minus STAIT in listening.

meaning of a word in the reading comprehension test they could go back


to look for the target word in context and obtain contextual cues. Audiovisual aids helped to diminish reading test anxiety, in contrast to previous
studies that examined the eects of audio-visual aids on listening anxiety
(Snyder & Golon, 1988; Arthur, 1999).
In the listening test, audio-visual aids resulted in higher listening comprehension scores and lower listening test anxiety, consistent with Secules,
et al. (1992), who found that exposure to audio-visual stimuli increased
learners' listening comprehension scores. Elkhafaifi (2005a) found a negative correlation (r = .65) between listening anxiety and listening comprehension; listening is highly anxiety-provoking if the discourse is incom-

16-PMS_Lee_140120.indd 585

15/04/15 6:36 PM

586

S-P. LEE, ET AL.

prehensible (Krashen, cited in Young, 1992, p. 163). When anxiety occurs


in the input stage, external stimuli such as audio-visual aids might reduce
students' anxiety (MacIntyre & Gardner, 1994a) and improve comprehension. In the current study, decreased listening test anxiety and higher listening scores with audio-visual aids suggest that audio-visual aids provide contextual support and could be applied in literary story teaching in
EFL courses to ease feelings of anxiety. Stories can be better understood
through visual cues instead of auditory components (Canning-Wilson,
2000), which helps overcome learners listening barriers.
In the reading test, when students had an audio-visual aid with a
printed text they achieved higher short-term (Cohen's d = 0.51) and longterm reading retention scores (Cohen's d = 0.62), but their reading comprehension scores (Cohen's d = 0.11) were unchanged. In contrast, in the listening test, when students had an audio-visual aid without a printed text
they obtained higher listening comprehension scores (Cohen's d = 0.45),
but their short-term (Cohen's d = 0.09) and long-term retention scores (Cohen's d = 0.18) did not increase. Thus, in the listening test audio-visual aids
facilitated listening comprehension better than audio used alone, as demonstrated in Rubin's study (1990; Cohen's d = 0.4), but they did not help
students remember the content. In other words, audio-visual aids might
help with information input, as suggested by Mayer (2003), but they did
not help the retention of listened-to information. The findings partially
supported the results of previous research, which has shown that visual retention is superior to auditory retention (Cohen, Horowitz, & Wolfe,
2009).
The visual aid increased students' listening comprehension but did
not improve short-term and long-term memory retention. The audio-visual aid helped students to understand, but not to remember, what they had
listened to. Foreign language instructors could link listening and reading
skills to activate learners auditory and visual memory in the input stage
to help students retain new information long enough to build new knowledge. Then the new knowledge could be used in subsequent productive
language tasks, such as active construction of a speaking or writing task.
Audio-visual aids could be useful tools both to reduce the test anxiety
caused by the listening and reading of foreign language stories and to improve listening comprehension, because the visual cues help deliver the
story line and compensate for listening barriers. Audio-visual aids do not
facilitate acquisition of language and literary story lines that require interpretation. It is suggested that a well-designed audio-visual aid combined
with reading skills tasks could be embedded into the learning cycle to integrate both language and content in literary story teaching for text appreciation and interpretation.

16-PMS_Lee_140120.indd 586

15/04/15 6:36 PM

AUDIO-VISUAL AIDS, EFL ANXIETY

587

Limitations and Future Research


Although this study provides some useful insights and empirical support for using audio-visual aids in EFL classrooms, it nonetheless has several limitations. The first limitation is that most of the participants were
women. A previous study has indicated that sex is an important variable in foreign language learning anxiety, with women being more anxious, but no significant dierence was found in listening anxiety (Elkhafaifi, 2005a). Second, this study used short stories as stimuli to test the
eect of audio-visual aids in reading and listening. The length or syntactic complexity of the selected texts may influence the results. Third, Swaffar, Arens, and Byrnes (1991) emphasized that reading comprehension is
a process of meaning-making. Multiple choice questions as assessment do
not require meaning-making and hence do not test students understanding of the story contexts.
Future studies could use a random sampling of participants, including more male participants or non-English majors, to observe the eects of
sex on measuring anxiety scales. Furthermore, it remains to be seen if applying longer texts or texts with higher contextual or syntactic complexities may result in dierent influences on test anxiety, test scores, and retention. A comparative study using short texts and long texts for measuring
test anxiety, performance, and retention may also provide interesting data.
It is of interest how much repetitions of the listening activity could improve memory retention. A method should be developed for training students to be meaning-oriented literary story readers with audio-visual aids
for text appreciation.
REFERENCES

AHMAD, I. S., AL-SHBOUL, M. M., NORDIN, M. S., RAHMAN, Z. A., BURHAN, M., & MADARSHA,
K. B. (2013) The potential sources of foreign language reading anxiety in a Jordanian EFL context: a theoretical framework. English Language Teaching, 6(11), 89.
ARNOLD, J. (2000) Seeing through listening comprehension exam anxiety. TESOL Quarterly, 34(4), 777-786.
ARONEN, E. T., VUONTELA, V., STEENAR, M. R., SALMI, J., & CARLSON, S. (2005) Working
memory, psychiatric symptoms, and academic performance at school. Neurobiology of Learning and Memory, 83(1), 33-42.
ARTHUR, P. (1999) Why use video? A teacher's perspective. VSELT, 2(4), 4.
BALTOVA, I. (1994) Impact of video on the comprehension skills of core French students.
Canadian Modern Language Review, 50(3), 506-531.
BEKTA-ETINKAYA, Y. (2011) Foreign language reading anxiety: a Turkish case. Journal of
Language Teaching and Learning, 1(2), 44-56.
AKIR, . (2006) The use of video as an audio-visual material in foreign language teaching classroom. Turkish Online Journal of Educational Technology, 5(4), 67-72.
CANNING-WILSON, C. (2000) Practical aspects of using video in the foreign language
classroom. Internet TESL Journal, 6(11), 1-6.

16-PMS_Lee_140120.indd 587

15/04/15 6:36 PM

588

S-P. LEE, ET AL.

CAROL, A. D. (1994) Comprehensible input through the practical application of videotext in second language acquisition. Italica, 71(3), 465-483.
CHANG, A. C. S., & READ, J. (2008) Reducing listening test anxiety through various forms
of listening support. TESL-EJ, 12(1), 1-25.
CHANG, C. S. (2008) Sources of listening anxiety in learning English as a foreign language. Perceptual & Motor Skills, 106(1), 21-34.
CHASTAIN, K. (1975) Aective and ability factors in second language learning. Language
Learning, 25(1), 153-161.
CHUNG, S. K., & LONG, C. F. (1984) A study of the revised State-Trait Inventory. Psychological Testing, 1, 27-36.
COHEN, M. A., HOROWITZ, T. S., & WOLFE, J. M. (2009) Auditory recognition memory is
inferior to visual recognition memory. PNAS, 106(14), 6008-6010. Retrieved from
http://www.pnas.org/content/106/14/6008.full.
DEVELLIS, R. F. (1991) Scale development theory and applications. London, UK: Sage.
ELKHAFAIFI, H. (2005a) Listening comprehension and anxiety in the Arabic language
classroom. The Modern Language Journal, 89(2), 206-220.
ELKHAFAIFI, H. (2005b) The eect of prelistening activities on listening comprehension in
Arabic learners. Foreign Language Annals, 38(4), 505-513.
ELLIS, R. (1981) The role of input in language acquisition: some implications for second
language teaching. Applied Linguistics, 11(1), 70-82.
EROSE7EN. (2008, May 8) Classic UPA cartoon: the unicorn in the garden [Video file].
Retrieved from http://www.youtube.com/watch?v=1teJjX-smdE.
GARRETT, N. (1989) Computers in foreign language teaching and research: a new humanism. In William H. Graves (Ed.), Computing across the curriculum: academic perspectives. McKinney, TX: Academic Computing Publications. Pp. 101-139.
HAMIL, K., & NEWMAN, T. (2008, September 16) The Giving Tree [Video file]. Retrieved
from http://www.youtube.com/watch?v=9PaDSZfZFZc.
HERRON, C., & HANLEY, J. (1992) Using video to introduce children to a foreign culture.
Foreign Language Annals, 25(5), 419-426.
HERRON, C., HANLEY, J., & COLE, S. (1995) A comparison study of two advance organizers
for introducing beginning foreign language students to video. Modern Language
Journal, 79(3), 387-394.
HERRON, C., MORRIS, M., SECULES, T., & CURTIS, L. (1995) A comparison study of the eects
of video-based versus text-based instruction in the foreign language classroom.
The French Review, 68(5), 775-795.
HERRON, C., YORK, H., COLE, S. P., & LINDEN, P. A. (1998) A comparison study of student
retention of foreign language video: declarative versus interrogative advance
organizer. Modern Language Journal, 82(2), 237-247.
HORWITZ, E. K., HORWITZ, M. B., & COPE, J. (1986) Foreign language classroom anxiety.
Modern Language Journal, 70(2), 125-132.
KLEINMANN, H. H. (1977) Avoidance behavior in adult second language acquisition. Language Learning, 27(1), 93-107.
KRASHEN, S. D. (1985) The input hypothesis: issues and implications. London, UK: Longman.
LEE, S. P., SU, H. K., & LEE, S. D. (2012) Eects of computer-based immediate feedback on
foreign language listening comprehension and test-associated anxiety. Perceptual
& Motor Skills, 114(3), 995-1006.

16-PMS_Lee_140120.indd 588

15/04/15 6:36 PM

AUDIO-VISUAL AIDS, EFL ANXIETY

589

MACINTYRE, P. D., & GARDNER, R. C. (1989) Anxiety and second language learning:
toward a theoretical clarification. Language Learning, 39(2), 251-275.
MACINTYRE, P. D., & GARDNER, R. C. (1991) Language anxiety: its relationship to other
anxieties and to processing in native and second languages. Language Learning,
41(4), 513-534.
MACINTYRE, P. D., & GARDNER, R. C. (1994a) The eects of induced anxiety on three stages
of cognitive processing in computerized vocabulary learning. Studies in Second
Language Acquisition, 16(1), 1-17.
MACINTYRE, P. D., & GARDNER, R. C. (1994b) The subtle eects of language anxiety on
cognitive processing in the second language. Language Learning, 44(2), 283-305.
MAYER, R. E. (2003) Nine ways to reduce cognitive load in multimedia learning. Educational Psychologist, 38(1), 43-52.
MILLS, N., PAJARES, F., & HERRON, C. (2006) A reevaluation of the role of anxiety: selfecacy, anxiety, and their relation to reading and listening proficiency. Foreign
Language Annals, 39(2), 276-295.
MUELLER, G. (1980) Visual contextual cues and listening comprehension: an experiment.
Modern Language Journal, 64, 335-340.
OMAGGIO, C. (1979) Picture and second language comprehension: do they help? Foreign
Language Annals, 12(2), 107-116.
PHILLIPS, E. M. (1992) The eects of language on students oral test performance and
attitudes. The Modern Language Journal, 76(1), 1-13.
RUBIN, J. (1990) Improving foreign language listening comprehension. In J. E. Alatis
(Ed.), Georgetown University round table. Washington, DC: Georgetown Univer.
Press. Pp. 309-316.
RUBIN, J. (1994) A review of second language listening comprehension research. Modern
Language Journal, 78(2), 199-221.
SAITO, Y., HORWITZ, E. K., & GARZA, T. J. (1999) Foreign language reading anxiety. The
Modern Language Journal, 83(2), 202-218.
SECULES, T., HERRON, C., & TOMASELLO, M. (1992) The eect of video context on foreign
language learning. The Modern Language Journal, 76(4), 480-490.
SELLERS, V. (2000) Anxiety and reading comprehension in Spanish as a foreign language.
Foreign Language Annals, 33(5), 512-521.
SHARWOOD, S. M. (1993) Input enhancement in instructed SLA. Studies in Second Language Acquisition, 15(2), 165-179.
SILVERSTEIN, S. (1964) The giving tree. New York: Harper & Row.
SNYDER, H., & GOLON, I. (1988) Foreign language acquisition and audio-visual aids. Foreign Language Annals, 21(4), 343-348.
SPIELBERGER, C. D. (1983) Manual for the State-Trait Anxiety Inventory (Form Y). Palo Alto,
CA: Consulting Psychologists Press.
SPIELBERGER, C. D., GORSUCH, R. L., & LUSHENE, R. E. (1970) STAT manual: for the State-Trait
Anxiety Inventory (Self-Evaluation questionnaire). Palo Alto, CA: Consulting
Psychologists Press.
SWAFFAR, J. K., ARENS, K. M., & BYRNES, H. (1991) Reading for meaning: an integrated
approach to language learning. Englewood Clis, NJ: Prentice Hall.
THURBER, J. (1940) The unicorn in the garden. In J. Thurber (Ed.), Fables for our time and
famous poems illustrated. New York: Harper & Brothers.

16-PMS_Lee_140120.indd 589

15/04/15 6:36 PM

590

S-P. LEE, ET AL.

TSAI, Y. C., & LI, Y. C. (2012) Test anxiety and foreign language reading anxiety in a
reading-proficiency test. Journal of Social Sciences, 8(1), 95-103.
WAGNER, E. (2008) Video listening tests: what are they measuring? Language Assessment
Quarterly, 5(3), 218-243.
YOUNG, D. J. (1986) The relationship between anxiety and foreign language oral proficiency ratings. Foreign Language Annals, 23(5), 539-553.
YOUNG, D. J. (1992) Language anxiety from the foreign language specialist's perspective:
interviews with Karasen, Omaggio Hadley, Terrel, and Rardin. Foreign Language
Annals, 25(2), 157-172.
ZHANG, X. (2013) Foreign language listening anxiety and listening performance: conceptualizations and causal relationships. System, 41(1), 164-177.
Accepted March 23, 2015.

16-PMS_Lee_140120.indd 590

15/04/15 6:36 PM

Perceptual & Motor Skills: Perception


2015, 120, 2, 535-555. Perceptual & Motor Skills 2015

THE ROLE OF CHUNKING IN DRAWING REY COMPLEX FIGURE1, 2


U. H. OBAIDELLAH
Faculty of Computer Science and IT, University of Malaya
and
School of Engineering and Informatics, University of Sussex
P. C-H. CHENG
School of Engineering and Informatics, University of Sussex
Summary.The study investigated the eects of chunking and perceptual patterns that guide the drawings of Rey complex figure. Ten adult participants (M
age = 22.2 yr., SD = 4.1) reproduced a single stimulus in four drawing modes including delayed recall, tracing, copying, and immediate recall across 10 sessions producing a total of 400 trials. It was hypothesized that the eect of chunking is most
obvious in the free recall tasks than in the tracing or copying tasks. Measures such
as pauses, patterns of drawings, and transitions among patterns of drawings suggested that participants used chunking to aid rapid learning of the diagram. The
analysis of the participants' sequence of chunk production further revealed that
they used a spatial schema to organize the chunks. Findings from this study provide additional evidence to support prior studies that claim graphical information
is hierarchically organized.

The process of drawing is increasingly of interest among educators to


be adopted as a method to facilitate the learning experience. This is often
practiced by teachers teaching early childhood education, when young
children are introduced to doodling or drawing to communicate freely, particularly children with low literacy skills (Young & Barrett, 2001).
There are many modes of drawing. For example, the tracing mode commonly requires the drawer to outline the figure, i.e., shown in dotted or
faded lines. The copying mode requires the drawer to refer to a target figure while reproducing the reference figure as accurately as possible. Another mode of drawing can be referred to as recall from memory, where
reproductions of the target figures are drawn based on recall of the previously seen figure, without any reference to the target figure at the time of
drawing. Recall from memory drawing tasks are useful to test a learner's
understanding of a particular concept that a figure expresses.
Address correspondence to Unaizah Obaidellah, Department of Artificial Intelligence, Faculty of Computer Science and IT, University of Malaya, 50603 Kuala Lumpur, Malaysia or
e-mail (unaizah@um.edu.my).
2
This research was in part supported financially by the University of Malaya Research Grant
(UMRG-RG110-12ICT). We thank the reviewers and associate editor for their comments
which improved this manuscript.
1

DOI 10.2466/24.PMS.120v17x6

13-PMS_Obaidellah_150013.indd 535

ISSN 0031-5125

15/04/15 6:33 PM

536

U. H. OBAIDELLAH & P. C-H. CHENG

An interesting question arising from drawing by using these various


modes relates to the kinds of strategies people use to reproduce figures. A
better understanding on how these dierent modes of drawings aect the
strategy of reproduction will be able to explain the eectiveness of using
these modes of drawing in education to benefit the learning process. Furthermore, if a figure is repeatedly drawn using these modes over a specified period, little is presently known about the kinds of eects they produce
on how well the graphically learned materials are recalled, such as whether tracing (as compared to copying) is a better drawing mode that can be
used to improve the rate of recall of a particular figure. The mental structures representing the knowledge acquired from graphical material remain
unclear. Previous studies (Bower, 1970; Palmer, 1977; McNamara, Hardy, &
Hirtle, 1989) on mental representation associated with semantic and spatial
memory have collectively proposed that information of these types are hierarchically organized. In this type of mental representation, information
such as items to be remembered' is categorized based on shared similarities such as functions or properties (i.e., apples and oranges are considered
fruits, while roses and camellias are flowers). This mental representation
comprises many levels, the most general information being represented at
the highest level and the most specific unit of information at the lowest level. These units of information are referred to as chunks (Miller, 1956), defined
as a group of related information that shares related characteristics and normally appears in sets or small groups (Miller, 1956; Tulving, 1962; Chase &
Simon, 1973; Buschke, 1976). For example, studies on semantic recall typically employ word association or categorization of word lists to assess the
eects of semantic structure on retrieval. Thus, the highest level of the hierarchy consists of category names that correspond to larger categories (e.g.,
animal), and the lowest level consists of smaller category names (e.g., dog,
bird) (Collins & Quillian, 1969, 1970). Mental representation of meaningful
knowledge in this manner is referred to as a schema (Piaget, 1926; Bartlett,
1932); schemas organize the knowledge a person possesses about the world,
events, people, and action (Mandler, 1984; Brewer, 1999).
It would be interesting to know whether drawing actions are guided
by a perceptual pattern or the existing chunks and schemas established by
and cultivated from prior experience, such as those developed in childhood, have a greater influence on drawing. One approach proposed by
Gestalt theory's perceptual principles (Wertheimer, 1923) posits that units
of elements are retrieved based on the principles of grouping together
items according to shared characteristics such as proximity (i.e., items are
spatially close together), similarity (i.e., items are similar in shape, color, or
size), common fate (i.e., items appear to move in the same direction), good
continuation (i.e., points for smooth curving lines are seen as belonging

13-PMS_Obaidellah_150013.indd 536

15/04/15 6:33 PM

CHUNKING IN DRAWINGS

537

together following the smoothest path), closure (i.e., items are perceived
as complete figures even when parts of the information are missing) and
symmetry (i.e., items are perceived as symmetrical that can be divided into
an even number of symmetrical parts) (Wertheimer, 1923). Bartlett (1932),
Egan and Schwartz (1979), Karmilo-Smith (1990), and Palmer (1977) offered another approach. They reported that when people draw, they retrieve graphical components in groups of elements sharing similar functions akin to chunks (Bartlett, 1932; Palmer, 1977; Egan & Schwartz, 1979;
Karmilo-Smith, 1990). These functions serving similar purposes are not
necessarily consistent in terms of the characteristics described by Gestalt's
perceptual principles. This means the elements of a chunk do not have
to belong to a strictly Gestalt principle of groupings, such as whether all
members of elements for a chunk share a symmetry pattern. For example,
Egan and Schwartz (1979), who sought to provide evidence of expertise
between skilled and unskilled participants on symbolic diagram reading,
reported that experts recalled symbols based on functional units sharing
related operations, rather than if the appearance of the symbolic representation look similar. The study reported in this paper describes an experiment that supports the second approach. Elements retrieved in chunks of
figures obtained from graphical material may be derived from an organized schema such as that described in a hierarchical structure (Palmer,
1977; Koedinger & Anderson, 1990).
Van Sommers (1984) investigated the strategies used by adults when
copying simple patterns. He developed a cognitive model of drawing positing that the process of drawing consists of an interaction between two
hierarchical systems: the visual perception and the graphical production
systems. Gurin, Ska, and Belleville (1999) accepted van Sommers' (1984)
model as a comprehensive cognitive model of graphical production that
incorporated principles governing drawings from a mechanical (i.e., action) and cognitive (i.e., perceptual) point of view. Van Sommers (1984)
proposed that drawing processes are influenced by graphic rules. One of
the rules van Sommers (1984) described suggests that participants adopt
common principles such as a preference for drawing in a counterclockwise
direction when drawing a circle at a starting point above a virtual axis (e.g.,
the top half of the circle such as going from 11 o'clock to 5 o'clock), whereas
they use a clockwise movement when a starting point is on the lower half
of the circle. This approach indicates that drawers commonly employ strategies that improve motor eciency (i.e., the movement of the hand and
fingers that influence drawing behavior), because the frequency of hand
movements decreases with temporal and spatial accuracy. In response to
this notion, Burke and Roodenrys (2000) and Vinter and Perruchet (2002)
further proposed that in relation to learning from drawings, the factors that

13-PMS_Obaidellah_150013.indd 537

15/04/15 6:33 PM

538

U. H. OBAIDELLAH & P. C-H. CHENG

influence the eectiveness of a drawing task are the level of observation,


the directionality of drawing movements, the size of the figure, and the average speed of hand movement. This means that the movement of the hand
increases as the size of the figure to be drawn becomes larger. On this assumption, it is necessary for the drawer to apply selective attention, emphasizing drawing groups of preferred components chosen from the respective whole figure, when strategizing an ecient execution (Farran, Jarrold,
& Gathercole, 2003). The selection of these parts, relative to chunks, may
potentially rely on a hierarchical organization and schemas (Palmer, 1977;
Koedinger & Anderson, 1990). Alternatively, the elements may be selected
line-by-line irrespective of any grouping relations between the lines during
the planning process, prior to the execution of any drawing actions. The
present study will assess whether graphical elements may be retrieved in
a hierarchical manner, predicted to consist of chunks, corresponding to the
notion of schemas as suggested by Palmer (1977).
Palmer (1977) presented a framework for perceptual representation in
the assessment of drawing tasks, such as verifying parts within a figure and
synthesizing the figure from separate parts. Participants were asked to draw
and rate simple straight-line figures in various configurations. He proposed
that graphical elements (i.e., simple, straight-line figures consisting of 16 segments) are processed in units that integrate hierarchically, forming a complete
network of chunks (units of elements) which he defined as a schema. According to Palmer, the use of a schema is advantageous during information selection for further processing, including the analysis of the selected data and eye
fixation on specific data of interest. Thus, a unified and coherent integration
of visual-spatial properties is made up of a hierarchy of progressively more
detailed and complex visual components from individual segmented patterns to complex figures. This hierarchy of visual-spatial properties provides
a way to envision the organization of graphic information.
The use of chunks in the drawings of simple geometrical shapes has
been reported by Cheng, McFadzean, and Copeland (2001) and Obaidellah and Cheng (2009). In a dierent experimental setting, Karmilo-Smith
(1990) found that children used chunks and schemas in their graphical
productions when they were asked to draw usual and unusual (i.e., items
the children usually see are defined by Karmilo-Smith as existent and
unusual items derived from their imagination are defined as non-existent)
sets of pictures (i.e., a house, an animal, or a man). For example, when instructed to draw a non-existent object older children show a tendency to
use chunks and schemas more noticeably, as they are able to interrupt patterns of chunks (e.g., older children drew parts of a house while they were
drawing a man) more easily than younger children (e.g., younger children
drew a complete drawing of the house before adding the features of the

13-PMS_Obaidellah_150013.indd 538

15/04/15 6:33 PM

CHUNKING IN DRAWINGS

539

man). This eect shows that younger children often perform whole parts
of their drawings before they continue drawing a dierent part of a figure.
Hence, with respect to drawings, the study reported by Karmilo-Smith
(1990) suggests that drawing in chunks of elements sharing similar characteristics could have been guided by the individual's knowledge of the
world, based on personal experience and principled understanding of a
particular concept or event (i.e., a schema). Vinter and Picard (1996) reported similar findings. According to Vinter and Picard (1996), the eects
of chunks were apparent when children (both younger and older) decomposed the objects into basic components (e.g., a house consists of a body,
a roof, windows, and a door). Each of the components is formed by an integration of several elements. The errors in the drawings were: change of
size, deletion, replication, changes of position or orientation of elements,
modification of whole shape, assimilation to another object, and the ability to adopt a cross-category change (e.g., drawing wings belonging to
the animal domain to the side walls of the house domain) from dierent
domains in the middle of their drawing activity (i.e., inter-representation
change). Younger children were more commonly observed to demonstrate
intra-representational element-based changes, a condition where drawings occur within categories and within chunks (i.e., complete their drawings of
all elements associated with the animal domain before they begin drawing elements associated with the house domain). In contrast, older children tended to show inter-representational whole-based changes, which indicate that they integrate and access components from dierent categories.
Therefore, older children could manipulate the drawings by integrating
dierent components from distinct categories more flexibly than only elements within a specific component from a single category.
Although Palmer (1977) suggested that while a simple graphical diagram consists of three levels, the potential number of hierarchical levels in
a complex diagram is unclear. Furthermore, no studies (Palmer, 1977; van
Sommers, 1984; Karmilo-Smith, 1990) have investigated the selection and
order of execution of parts from the whole figure in drawing strategies for
dierent types of drawing tasks, such as tracing, copying, and recall from
memory. It is important to discover how these drawing tasks facilitate a
drawer's learning from diagrams. The drawing mode, whether tracing or
copying, which facilitates learning the most is still unknown (Gonzalez,
Anderson, Culmer, Burke, Mon-Williams, & Wilkie, 2011). Improved understanding will enable proper complexity of graphical materials used for
learning, particularly in educational subjects such as physics and chemistry. In addition, at present little is known about how graphical elements are
stored mentally and retrieved with respect to drawing mode. Although van
Sommers (1984) successfully demonstrated that common drawing practic-

13-PMS_Obaidellah_150013.indd 539

15/04/15 6:33 PM

540

U. H. OBAIDELLAH & P. C-H. CHENG

es (i.e., a sequence of graphical productions such as the anti-clockwise approach when writing the letter O' among right-hand writers) are generally
applied among individuals regardless of age, sex, and cultural background,
it is presently uncertain if this generalizes to relatively complex diagrams.
The main purpose of this study is to determine if drawings based on
dierent modes (i.e., tracing, copying, immediate recall from memory, and
delayed recall from memory) are more likely driven by strict perceptual
patterns or by an influence of chunks and schemas. The previously discussed research (Cheng, et al., 2001) used simple and straight-line figures
as stimuli. In this study, a more complex figure was employed as a stimulus to compare drawing strategies to those used for simpler diagrams
(Palmer, 1977; van Sommers, 1984). The goal was to investigate the types
of strategies (i.e., order of elements drawn) used to reproduce the complex
figure in dierent drawing modes. Data from 10 adult participants who
reproduced the Rey figure in four modes of drawings across ten sessions
producing a total of 400 trials were analysed. As previously discussed,
the participants may adopt any or a combination of the following potential strategies: (1) drawing patterns in an arbitrary order based on Gestalt
principles, e.g., selection of drawn elements based on symmetry before
those elements in close proximity; (2) drawing lines that maximize motor eciency; and (3) using some kind of hierarchical structure. Finally, if
a hierarchical structure is used, it was of interest how the elements of the
figure would be represented in the hierarchy. Hence, the motivation is further developed by investigating which of these strategies show the strongest influence in the drawing protocols.
Hypothesis 1. Drawers will recall and reproduce graphical elements of a figure in chunks that are derived from an organized
schema in the form of a hierarchical structure.
Hypothesis 2. If chunking strategies are adopted, the production of
the drawings will be more consistent, demonstrated by regular
patterns of chunks (i.e., each chunk consists of the same collection of elements), and will reduce the number of transitions between elements as compared to higher transition counts when
no chunking strategies are used (i.e., elements were drawn
randomly) or if irregular chunk patterns are drawn.
METHOD
Participants
Ten adult (6 men, 4 women) undergraduate and postgraduate university students participated in the experiment. Their ages were between 21
and 30 years (M = 22.2 yr., SD = 4.1). Eight of them were right-handed and

13-PMS_Obaidellah_150013.indd 540

15/04/15 6:33 PM

CHUNKING IN DRAWINGS

541

two were left-handed. Eight participants had normal vision, while two of
them had corrected-to-normal vision. All had sucient fundamental experience with drawing scientific figures, as all of them had science education backgrounds. There were no specific criteria for the selection of their
participation. However, the pool of participants was focused on trained
drawers (i.e., university students) since the authors were interested to investigate their research aims among normal and healthy adults. Thus, the
university community serves as an ideal choice for this. None of them
had memory, drawing, or motor impairments as demonstrated by successful drawing of simple shapes prior to the experiment. All participants
received a small monetary reward for their participation.
Material
All drawings were performed on a Wacom Intuos3 graphics tablet
with a special inking pen. An A4 paper was taped on the graphics tablet in
a landscape orientation and participants were required to draw a large figure on the blank sheet. A specially written program, TRACE (Cheng & Rojas-Anaya, 2004), was used to capture all drawing actions, extract pen positions, and compute pause durations between the drawn elements. The
single stimulus adopted from the original Rey figure as shown in Fig. 1(a)
was printed in black ink on a white sheet for all tasks in this experiment.
Modifications such as blank spaces introduced at the ends of all lines were
used to enable recording of the pauses that occur during the transitions
between drawing strokes.
Measures
Four types of tasks were chosen. A Tracing task was administered to
assess the eects and extent of chunking employed in drawings when little eort was necessary during retrieval. A possible approach taken in the
Tracing task is to draw consecutive elements based on the nearest-neighbor strategy as an economical drawing method. To test whether parts of
a figure are drawn in groups of related elements based on Gestalt principles, a Copying task was used to evaluate whether the process of copying
parts of a figure would change over time or the retrieval of the elements
would exhibit a recurring and consistent pattern between and within individuals. An immediate recall task was used to study the eects and accuracy of retrieval from memory after recent exposure to the stimulus. A delayed recall task was used to study how coherently and precisely a figure
could be reproduced following a delay of as long as 24 hours.
As shown in Fig. 1(a), the Rey-Osterrieth Complex Figure (hereafter
called the Rey Figure) was adopted with a few modifications (i.e., gaps in
between lines, break longer lines into shorter ones) for the technique called
Graphical Protocol Analysis (Cheng & Rojas-Anaya, 2008). The method

13-PMS_Obaidellah_150013.indd 541

15/04/15 6:33 PM

542

U. H. OBAIDELLAH & P. C-H. CHENG

FIG. 1. (a) Modified Rey Figure, (b) Default pattern, (c) L1 and L2 pauses

measures the pause durations between the drawn elements, i.e., the time
between the end point of the last drawn line and the starting point of the
successive line. Graphical Protocol Analysis was employed, as this method
uses pauses as an indication of chunk boundaries, similar to other methods
such as the standard reaction time, qualitative scores of drawings, and verbal protocol analysis (Egan & Schwartz, 1979; Karmilo-Smith, 1990; van
Mier & Hulstijn, 1993). The Rey Figure was chosen because of its extensive use in established empirical psychological studies related to memory
tests. This relatively complex figure comprising 56 lines in various putative
groups of elements is a suitable stimulus to assess the hypotheses.
Graphical Protocol Analysis is an improved method of analyzing
drawing behavior compared to the standard reaction time technique, as
data recording is accurate and the analysis is simpler: drawings are done
on a graphics tablet to capture natural drawing actions, with accurate
measurement of drawing and pause timing. A number of writing studies employing Graphical Protocol Analysis have demonstrated a coherent temporal pattern indicating the structure of chunks in memory during
tasks such as writing number sequences, word phrases, and mathematical formulae (Cheng & Rojas-Anaya, 2005, 2006, 2007). The chunk patterns
are indicated by statistically significant dierences between the duration
of the pauses at dierent hierarchical levels, and were observed in all of
the above tasks (Cheng & Rojas-Anaya, 2008). Specifically, in drawings
this method is adequate for the investigation of the internal mental representations when drawing simple geometric patterns (Cheng, et al., 2001).
The present study applies the Graphical Protocol Analysis method for all
drawing activities obtained from more complex graphical stimuli.
As shown in Fig. 1(c), the two types of pauses examined in this experiment were coded as (a) L1-element within patterns, defined as the

13-PMS_Obaidellah_150013.indd 542

15/04/15 6:33 PM

CHUNKING IN DRAWINGS

543

pauses measured between successive elements of the same pattern; and


(b) L2-element between patterns, defined as the pause between the last
line of a pattern and the first line of a dierent pattern. To facilitate and
standardize the process of scoring the participants' drawings, as shown
in Fig. 1(b), the Rey Figure was decomposed into 13 patterns according to
specific grouping criteria where each was named (e.g., gill, fin, eye, brow)
following the terms used to describe the anatomy of a fish. These criteria,
otherwise known as coding schemes, largely similar to the categorization
criteria of Osterrieth (1944), Lezak (1983), and Corwin and Bylsma (1993),
were developed to test the consistency of the sequence of the elements
drawn by the participants (i.e., consistency of drawing patterns for both
between and within participants) and to provide rules to resolve ambiguities in scoring confusing drawings such as whether an element or pattern
is drawn by mistake and thus should not be considered (e.g., if a participant draws a flag at the tip of the fin pattern) or should be accepted as
part of the closely related pre-defined pattern (e.g., if a participant draws
a slanting line instead of a straight vertical line to represent the element of
the brow pattern).
Procedure
Experiments were conducted in independent sessions with each participant. They were initially given instructions about the tasks they were
required to perform. Each participant completed 10 sessions with four
tasks (i.e., Tracing, Copying, Immediate recall, and Delayed recall from
memory), producing a total of 400 trials. The description of the tasks were:
(1) Copying, in which the figure was reproduced on a blank piece of paper with the original figure placed next to the blank sheet; (2) Tracing, requiring replication of the figure printed in faded grey lines; (3) Delayed
recall, a reproduction of the figure based on recall from memory. The target figure was presented prior to both of the copying and tracing tasks;
(4) Immediate recall, in which the goal was to reproduce the figure based
on recall from memory, after the copying and tracing tasks. The participants initiated the experiment with a practice task to familiarize themselves with drawing on a graphics tablet.
In the first session, all participants began with the Tracing task followed by the Copying task and the Immediate recall task. In the evennumbered sessions, the participants drew the figure in this order: Delayed
recall, Copying, Tracing, and Immediate recall. In the odd-numbered sessions, the participants followed the order: Delayed recall, Tracing, Copying, and Immediate recall. Only the Copying and Tracing tasks were counterbalanced to reduce the possibility of an order eect, so that the results
are less likely aected by the order of the stimulus presentation since the

13-PMS_Obaidellah_150013.indd 543

15/04/15 6:33 PM

544

U. H. OBAIDELLAH & P. C-H. CHENG

participants may develop a regular strategy applied in all the drawing activities if the same stimulus ordering were given in all 10 sessions. The Delayed and Immediate recall tasks were not counterbalanced, as the authors
were particularly interested to find if recall performance was statistically
dierent prior and after executing both the Tracing and Copying tasks.
In all drawings, participants began by writing a hash (#) to enable recognition of the time at which the figure was begun and to ensure drawing
and thinking processes were underway. There was no time restriction for
the duration of the drawings. The participants drew until they had nothing
else to draw. They were given an option to end the session for the particular drawing task after being asked if they had anything else to draw. Most
participants did not recall more elements. There was a delay of at least one
day, but no more than two days, between the Immediate recall task and the
Delayed recall task in the subsequent session. The inclusion of 10 sessions
was meant to provide an opportunity for the participants to complete their
learning process, so as to be able to perform chunking eectively.
RESULTS
Pause Analysis
During drawings, a few of the participants made mistakes by drawing
broken lines. The frequency of this kind of mistake was recorded as errors
at the time of data pre-processing. The errors were divided into three categories according to the level and type of errors committed. For example, a
structure error was defined as a type of error that was made when a participant produced lines that were too short or too long, or misplaced elements
that could occur at the level of the entire pattern (i.e., drawing the eye pattern in the wrong location) or a single element of a pattern (i.e., drawing a
single short line for the box pattern). However, the frequencies of these errors
were very few (e.g., less than five in each session when aggregated across the
participants) to qualify for a statistics test. Thus, given that these data were
removed from the other data for analysis, the pauses used for analysis were
due to changes between elements and not the broken lines of a particular element. This coding is based on the experimenter's observation as the participants did the drawings. For example, the participants correctly traced over
the original faded lines, as can be seen in the hardcopies of the tasks. Thus,
the pauses the participants established were due to changes from one element to another and not during the drawing of that particular element.
As shown in Fig. 2, across all modes of drawing for all participants and
sessions, the L1-within pattern pauses were shorter (within the range of 387
649 msec.) than the L2-between patterns pauses (6951833 msec.). The L2
pauses were more variable. A test of equal variance on the pauses showed
that the variances were unequal for the L1 and L2 pauses (F1, 718 = 78.1, p < .001).

13-PMS_Obaidellah_150013.indd 544

15/04/15 6:33 PM

Copyright of Perceptual & Motor Skills is the property of Ammons Scientific, Ltd. and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.