Anda di halaman 1dari 7

ATMI TO MEASURE THE MATHEMATICS ATTITUDE

IN ELEMENTARY STUDENTS

Siska Nur Rahmawati1,


1,2
Yogyakarta State University, Jln. Colombo No.1, Sleman, Daerah Istimewa Yogyakarta, Indonesia
email: siskagamaya@gmail.com

Abstract
Character education could not be separated from the affective sphere or student attitudes.
Students 'attitudes have been known to be influenced the students' activeness and achievement
toward mathematics learning. Some instruments have long been developed to measure students'
mathematical attitudes such as Attitudes Toward Mathematics Inventory (ATMI) that developed
by Tapia & Marsh. The purpose of this research is to analyze the construct validity of ATMI.
Instruments were distributed to 150 students at 6th grade elementary school. Confirmatory
Factor Analysis were used to see the good model for dimensions in mathematics attitude. This
validation tried to reveal the suitability of the instruments to measure the elementary students'
mathematics attitude in the Yogyakarta context.

Keywords: Mathematics attitude, ATMI, Construct Validity, CFA

1
1 INTRODUCTIO cognitive more stable than theoretical
N understanding, belief and/ or framework
Many while affective value. (McLeod, 1992;
researchers still factors play a . Based on Zan & Di
interest to assess crucial role in the the statement, Martino, 2003)
the attitudes of process of somehow the that needs to be
students, parents, learning primary students developed further.
teachers, or even mathematics. The are suitable to The
administrators study conducted explore the Attitudes Toward
(Adelson and by McLeod attitudes. At the Mathematics
McCoach: 2010). around the 90's, in growing phase of Inventory (ATMI)
Allport in Pickens the newest, was them, the attitudes (Tapia & Marsh,
(2005) defined an arranged about of primary 2004) is one of
attitude as a the affective students are a the latest
readiness, domain in dominant instruments, but it
organized through mathematics subdomain that has not enjoyed
experience, education. recognizably to significant
exerting a McLeod (1992) measure. The application in
directive or Together with his development of research
dynamic colleagues study about (Chamberlin,
influence on the divided affective student attitudes 2010). It was
individual’s spheres into three at long period of chosen for this
response to all sub-domains: time conclude study because it
objects and emotions, that it has provides a sharp
situations to attitudes, and powerful impacts and distinct focus
which it is beliefs. DeBellis on their effective and identifies four
related. So that and Goldin engagement, dimensions along
the mathematics (2000) add participation and which attitudes
is field that an subdomain ie achievement in toward
attitude could value. Recent mathematics mathematics
impact the mental research, (Majeed, could be
process of "attitude" is used Darmawan, measured. The
individuals even to sum it all up. Lynch, 2013). improved ATMI
students (Majeed, Affective math is Many researchers comprised 40
Darmawan, directed to chart claim that despite items that
Lynch, 2013; Di the basics of the fact that measured four
Martino & Zan, anxiety and research on domains:
2011; Goldin, mathematical attitude, as enjoyment, value,
2000, 2002; procedures compared to other motivation, and
Grootenboer & (Evans, 2006). subdomain of self-confidence
Hemmings, 2007; Leder and affects, has the (Majeed,
Malmifuori, 2001, Grootenboer longest history, Darmawan,
2006, 2007; (2005) express the term attitude Lynch, 2013).
Schloglmann, statements with remains an Scoring was done
2003). the category “ambiguous with a five-point
There is an inside. The construct” (e.g. Likert Scale, with
increased attitude is more Hart, 1989; response options
understanding of stable than Hannula, 2002) from “strongly
the factors that emotions and with an disagree” to
influence feelings, but not ambiguous “strongly agree”.
2
Siska Nur Rahmawati

As a of the instrument, Indonesia keeping RMSEA (Root


development only 32 were used in mind that this Mean Square
process of ATMI, and omitting eight study conducted Error of
based on that items to make the to explore the Approximation),
definition above, instrument more ATMI in RMR (Root
(1) is the ATMI a manageable Yogyakarta Mean-square
reliable and a length and context. After Residual), GFI
valid instrument prevent the translating and (Goodness-of-Fit
to measure negative impact proofreading it, Index), NFI
elementary on completion by 40 items of ATMI (Normed Fit
students’ attitudes the respondent then reduced to Index), Non-
toward without 32 items and Normed Fit Index
mathematics in compromising its omitting eight (NNFI), and CFI
the Yogyakarta validity. items without (Comparative Fit
context? So, the 2.1 Research compromising its Index). Loading
purpose of this Settings validity to prevent factor and t-Value
research is to find The study the respondent are criteria that
the construct was conduct at dropouts during were used to
validity of the September 2017 data collecting. analyze the
ATMI in in 7 public The final measurement
Yogyakarta elementary school instrument then model fit. The
context especially in Sleman distributed to the standard of
in elementary Regency, Special respondent. significance value
students. Region of The data for the validity
Yogyakarta. then analysed based on Hair,
2 METHOD 2.2 Research using Lisrel 8.80 Black, Babin,
This study Subject to measure the Ander-son, &
conducted as Using construct validity. Tatham (2010)
quantitative cluster random Confirmator were “factor
descriptive to sampling out of y factor analysis loadings ± .3 to .4
explore the 14 public was used to are minimally
construct validity elementary analyse the acceptable”.
of ATMI for schools in Sleman construct validity
elementary Regency as of the data. 3 RESULTS AND
student in population, 7 Firstly, the overall DISCUSSION
Yogyakarta elementary model fit was Before
context. A total of schools were conduct to examine the
184 elementary chosen, and examine the fit of construct validity
students reaching 184 the model based of the instrument,
participated in elementary on the goodness reliability was
this study. The students at 6 fit indices then assessed to
respondents were grades as measure the confirm the
elementary respondent. measurement of internal
students at 6 2.3 Procedure and fit. The criteria consistency of
grades in 7 public Data Analysis that were used to instruments items.
elementary The ATMI evaluate the The Cronbach’s
schools in Sleman instrument was goodness of fit alpha statistics for
regency. adapted by are: normed chi- the instrument
Reducing translating it from square ( ), was .741. These
the total of items English to Bahasa results confirmed
3
Siska Nur Rahmawati

the adequacy of necessary to the suggested score contruxt validity


the internal overall model of (RMSEA< .05), of the instrument
consistency of the fit first. If overall so the level of fit that the
instrument. model fit match is in poor fit level is marginal
To examine with the criteria, (Browne & and good fit.
the overall and then the Cudeck, 1993). (Why the
measurement measurement The result of marginal can be
model fit, model fit could be RMR and CFI is . use to measure,
confirmatory conduct to fulfil 089 and .92 because the
factor analysis the construct respectively, with marginal fit is
was used. The validity analysis N≤250 the model closely fit)
explanation of condition. will in good fit Marginal fit = fit
each criteria in (menentukan fit level if diambang batas.)
overall model fit model) Standardized 3.2 Measurement
that mentioned The table RMR ≤ .09 and Model Fit
before i.e. below shows the CFI> .92 (Hair, After the
Normed Chi- overall model fit Black, Babin, overall model is
Square is ratio value after the Anderson, & fit, the
between Chi- data was analysed Tatham, 2006) so measurement
Square and using Lisrel 8.80. that those criteria model fit was
degree of are in marginal conduct. The
freedom. RMSEA GOF Resul Level of Fitfit. The criteria of result of
is most
Criteria t GFI range from 0 measurement
informative Normed χ² 3.11 Poor fit (poor fit) to 1 model fit for
indicator for
RMSEA .107 Poor fit (perfect fit). The ATMI were
model fit. RMR RMR .089 GFI scores .67 so
Marginal fit described here.
represent the
GFI .67 Good fit that the fit level is Measurement
residual mean NFI by .88 in good criteria.
Marginal fit model fit conduct
matching the
NNFI .91 The criteria of
Marginal fit using first order
covariance matrix CFI .92 GFI is same for
Marginal fit confirmatory
of the data. GFI is NFI and NNFI. factor analysis or
a scale of Table 1 Considering the 1st CFA. The
precision of the shows the normed condition above, criteria for
model that χ2, that is the it could say that construct validity
resulting ratio between the the overall model are loading factor
covariance χ2 and degree of is fit to measure and t-Value. If
matrix. NFI has a freedom. Good fit the construct loading factor is
tendency to level suggests that validity of the greater than .3
lowering the fit in the score must be instrument. and t-Value more
small sample size. range from 1.00 (based in than 1.96 then the
NNFI was used to to 2.00. because table above we item is
fix the problem the score is 3.11 can use GOF categorized as
that caused by the so that the Criteria RMR, valid item. Those
complexity of the normed χ2 is in GFI, NFI, NNFI, criteria refer to
model. the poor level. CFI to assumed assumption that
3.1 Overall Model RMSEA scored . that the overall “factor loadings ±
Fit 055. Because the model is fit, so we .3 to .4 are
To analyse score is greater can go to the next minimally
the construct than the step to acceptable.”
validity, it measurement the (Hair, Black,
4
Siska Nur Rahmawati

Babin, Anderson, shows that the whose sample REFERENCES


& Tatham, 2006). items of the involved 699 Adelson, J.L., &
instrument have Osuth Australian McCoach, D.B.
First Order CFA loading factor > . students in 7 and (2010).
Factor Item 3 and t-Value > 8 grades. Measuring the
LF t-Val
SlfCon SF1 .80 10.28 1.96 it means the By reducing mathematical
SF2 .79 10.91 instrument was the 40 items of attitudes of
valid ATMI, the 24 elementary
SF3 .81 11.20
constructively and items he students: The
SF4 .76 11.82
could measure the reliability and effects of a 4-
SF5 .83 11.57 gap of attitude in. validity estimates point or 5-point
SF6 -.46 -5.33 Item SF 6, SDF 7, for ATMI are Likert-type
SF7 -.57 -9.18 SF 8, SF 9, SF 11, stable over many scale.
SF8 -.50 -6.85 SF 12, Enj 2, Mot years after its Educational and
SF9 -.49 -7.22 2 is not valid initial Psychological
SF10 .61 7.87 constructively to administration in Measurement
SF11 -.49 -6.63 measure the 1996 and beyond 70(5) 796-807.
SF12 -.51 -7.98 affective attitude the initial Browne, M.W., &
Val Val1 .52 9.82 from elementary samples. These Cudeck, R.
Val2 .30 5.84 student of considerations (1993).
mathematics. provide Alternative
Val3 .39 7.04
compelling ways of
Val4 .68 10.53 4 CONCLUSION rationale for its assessing model
Val5 .66 10.83 Based on use in future fit. In K. A.
Val6 .68 10.67 analysis result, it research about Bollen & J.S
Val7 .59 8.70 can be concluded attitudes toward Long (Eds),
Enj Enj1 .69 10.40 that conclusion 24 mathematics Testing
Enj2 -.73 -8.02 items have factor (Majeed, Structural
Enj3 .58 7.74 loading > .3 and Darmawan, & Equatin Models.
Enj4 .60 7.74 t-value > 1.96 so Lynch, 2013). Newbury Park,
Enj5 .71 9.04 that out of 32 The ATMI is CA: Sage, 136-
Enj6 .47 6.09 items only 24 particularly 162 (Electronic
items are valid. 2) useful, both for Version).
Enj7 .74 10.97
The 24 items of teachers, who Chamberlin, S.A.
Enj8 .44 6.77
adapted ATMI are want to monitor (2010). A review
Enj8 .57 8.43 constructively students attitude of instruments
Mot Mot1 .62 9.37 valid to measure toward created to assess
Mot2 -.52 -5.83 the mathematics mathematics, and affect in
Mot3 .65 8.99 attitude from for researchers, mathematics.
Mot4 .55 8.43 elementary who often use Journal f
student in different Mathematics
Not Yogyakarta instruments in Education, 3(1),
valid=because context. The their studies. For 167-182.
didn’t range the results are the further study, Di Martino, P., &
minimum criteria. consistent with larger sample and Zan, R. (2010).
Table 2 the factor better translating Me and maths:
shows the structure reported may give more Towars a
summary of by Majeed, credible result. definition of
construct validity Darmawan, & attitude
using 1st CFA. It Lynch (2013), grounded on
5
Siska Nur Rahmawati

students mathematical R.I. (2006). affect,


narratives. problem solving. Multivarite cognition, and
Journal of Mathematical Data Analysis, social
Mathematics Thingking and 6th edition. New environment in
Teacher Learning, 2, Jersey: Person the regulation of
Education, 209-219. Prentice Hall. personal
13(1), 27-48. Goldin, G.A. (2002). Hair, J.F., Anderson, learning
Di Martino, P., & Affect, meta- R.E., Tatha,, processes: the
Zan, R. (2011). affect, and R.L. & Black, case of
Attitude towards mathematical W.C. (2010). mathematics.
mathematics: a belief structures, Multivariate Department of
bridge between in G. Leder, E. data analysis Teacher
beliefs and Pehkonen & G. (7th ed). New Education,
emotions. ZDM: Torner (Eds.), Jersey: Prentice- Research Report
The Beliefs:A hidden Hall. 172. Helsinki:
International variable in Hart, L.E. (1989). Helsinski
Journal on mathematics Describing te University
Mathematics education (pp. affective Press.
Education, 43, 59-72). domain: saying Malmivuori, M.L.
471-482. Dordrecht: what we mean. (2006). Affect
Evans, J. (2006). Kluwer. In D.B. McLed and self-
Affect and Grootenboeer, P. & & V.M. Adms regulation.
Emoticon in Hemmings, B. (Eds). Affect Educational
Mathematical (2007). and Studies in
Thingking and Mathematics mathematical Mathematics,
Learning. The performance and problem 63(2), 149-164.
Turn to the the role played solving; A new Malmivuori, M.L.
Social: by affective and perspective (2007).
Sociocultural background (pp.37-45). New Understanding
Approaches factors. York, NY: student affect in
Introduction: Mathematics Springer-Verlag. learning
Recent Education Majeed, A.A., mathematics. In
Developments Research Darmawan, C.L. Petroselli
in Research On Journal, 19(3), I.G.N., & (Ed), Science
Affect. In J. 3-20. Lynch, P. education:
Maasz, W. Hannula, M.S. (2013). A Issues and
Schloeglmann (2002). Attitude confirmatory developments
(Eds), New towards factor analysis (pp. 125-149).
mathematics mathematics: of atttudes New York, NY:
education Emotions, toward Nova Science
research and epectations and athematics Publishers.
practice (p.203- values. inventory McLeod, D.B.
208). Educational (ATMI). The (1992).
Rotterdam: Studies in Mathematics Research on
Sence Mathematics, Educator: Vol affect in
Publishers. 49(1), 25-46. 15, No. 1, 121- mathematics
Goldin, G.A. (2000). Hair, Jr.J.F., Black, 135. education: A
Affective W.C., Babin, Malmivuori, M.L reconceptualizat
pathways and B.J., Anderson, (2001). The ion. In D.A.
representation in R.E. & Tatham, dynamics of Grouws (Ed).
6
Siska Nur Rahmawati

Handbook of Educational
research on Research
mathematics Association,
learning and Chattanooga,
teaching TN. (ERIC
(pp.575-596). Document
New York, NY: Reproduction
Macmillan. Service No. ED
Leder, G.C. & 471 301).
Grootenboer, P.
(2005).
Editorial: affect
and mathematics
eduation.
Mathematics
Educaton
Research
Journal, 17(2),
1-8.
Pickens, J. (2005).
Attitudes and
perceptions.
Organizational
behaviour in
health care, 43-
76.
Schloglmann, W.
(2003). Affect
and cognition:
Two poles of a
learning
process.
Retrieved from
http://www.educ
ation.monash.ed
u.au/project/vam
p/schloglmann2
001.pdf.
Tapia, M., & Marsh,
G. E., II. (2002).
Confirmatory
factor analysis
of the Attitudes
Toward
Mathematics
Inventory. Paper
presented at the
Annual Meeting
of the Mid-
South
7

Anda mungkin juga menyukai