Available online at www.sciencedirect.

com

Procedia - Social and Behavioral Sciences 69 (2012) 422 – 427

International Conference on Education and Educational Psychology (ICEEPSY 2012)

The Cross-Cultural Examination of 3×2 Achievement Goal Model

in Taiwan
Chun-Chin Wu*
Institute of Education, National Cheng Kung University, Tainan, 70101, Taiwan

Abstract

There were many problems in dichotomy, trichotomy and 2×2 achievement goals questionnaires. In addition,
the 2×2 achievement goal models compound two foci- absolute/intrapersonal in mastery goal, resulted in failing
to assess the achievement goals. Therefore, researchers proposed 3×2 achievement goal model to better assess
students' achievement goals. Until now, there is a few empirical and cross-cultural researches to verify about
the 3×2 achievement goal model. The purpose of present research is to investigate the construct and structural
validity between junior high and elementary school students in Taiwan. If any gender differences exist on
students' achievement goals also discussed. The questionnaire is shown to have good reliability of internal
consistency on both school levels. For junior high school students, the 3×2 achievement goal model fit very
well, but the elementary school students' data does not. This result may imply the elementary school students
cannot clearly discriminate between 3×2 achievement goals as well as the junior high school students.
Generally, the 3×2 achievement goal model fits better than the other three competing models (dichotomy,
trichotomy and 2×2 achievement goals) in both groups. Results from the chi-square tests indicate that there are
no gender differences in the pattern of achievement goals. Finally, the implication of future researches are
discussed.
© 2012 The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
Selection and peer-review
© 2012 Published by Elsevierunder responsibility
Ltd. Selection of Dr. Zafer
and/or peer-review Bekirogullari
under of of
responsibility Cognitive
Dr. Zafer–Bekirogullari
Counselling, of Research
Cognitive & Conference
– Counselling,
Services C-crcs.
Research & Conference Services C-crcs.
Keywords: Achievement goals; goal orientation; cross-cultural examination

1. Introduction

Most of people concern about students' academic achievement, especially focus on their mathematics grades in
Taiwan. People always overemphasize on external criteria led to ignore the most complicated psychological process.
Students who get the same grades are regarded as similar performance individuals. But, students may pursue the
same grades for varied reasons, and lead to different cognitive, affective and behavioral consequences (Schunk,

* Corresponding author. Tel.: +002886426878989; fax: +002886426878988.

E-mail address: [email protected]; [email protected].

1877-0428 © 2012 The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
Selection and peer-review under responsibility of Dr. Zafer Bekirogullari of Cognitive – Counselling, Research & Conference
Services C-crcs.
doi:10.1016/j.sbspro.2012.11.429
 Chun-Chin Wu / Procedia - Social and Behavioral Sciences 69 (2012) 422 – 427 423

Pintrich, & Meece, 2008). The main purpose which achievement goal theorists striving to answer is why students
possessing different reasons when they are engaging in the same task or activity (Maehr & Zusho, 2009).
The earlier taxonomy of achievement goal is dichotomy and named differently, but they represented similar
constructs: mastery goal and performance goal. Mastery goals means students engaged in learning task in order to
develop their competence. In contrast, performance goals means students engaged in learning task in order to
demonstrate their competence (Senko, Durik, & Harackiewicz, 2008).
Elliot and Harackiewicz (1996) argued that performance goals in dichotomous achievement goals did not take
two foci approach and avoidance into account. The focus of approach leads students to move toward positive end-
states, conversely the focus of avoidance leads students to move away from negative end-state (Higgins, 1997).
Therefore, it is best to separate them from performance goals, and form the trichotomous achievement goals,
including mastery, performance-approach and performance-avoidance. Further, Elliot and McGregor (2001)
grounded on their former theoretical assumption and practical observation, they discriminated achievement goals
along two dimensions: definition and valence of competence. They added mastery-avoidance into trichotomous
achievement goals, and proposed the 2×2 achievement goals which contained mastery-approach, mastery-avoidance,
performance-approach and performance-avoidance. Empirical researches also indicated students possessed mastery-
avoidance are prevalent (van Yperen & Renkema, 2008), and this kind of goal can undermine their performance and
emotion when they engaged in achievement task (Huang, 2011; Jang & Liu, 2012). Moreover, the 2×2 achievement
goals model also had good discriminative, criterion-related validity (Huang, 2012) and convergent validity (Cherng,
2003).
Unexpectedly, results of dichotomous, trichotomous and 2×2 achievement goals may be biased, because many
problems existed in the measurement of achievement goal questionnaires. The problems including as following
(Elliot & Murayama, 2008): (1) Putting one goal against another and more than one goal focus in a question. (2)
Failing to focus on achievement goals per se. (3) Use the same content for different goals. (4) Different comparative
norms between performance-related goals. (5) Differently hypothetical and affective component in questions. (6)
Intense components of content. (7) Mixed social goals or motivation with individual achievement goals. In addition,
the 2×2 achievement goal model was confounded two foci- absolute/intrapersonal in mastery-based goals, resulting
in failing to assess the achievement goals. In order to clarify the constructs of achievement goals, Elliot and
Murayama (2008) revised those problems, and Elliot, Murayama, and Pekrun (2011) differentiated between
mastery-based goals, then proposed 3×2 achievement goal model. Elliot et al. (2011) based on the definition and
valence of competence and divided the definition of competence into three types of goals: task, self, and others. The
valence of competence is as well as the 2×2 achievement goal model: approach and avoidance. Consequently, there
are six dimensions in the 3×2 achievement goals: approach-task, avoidance-task, approach-self, avoidance-self,
approach-others, and avoidance-others. Elliot et al. (2011) argued that cross-cultural examination is needed to
confirm the validity and invariance of the 3×2 achievement goal model. Until now, there is a few empirical and
cross-cultural researches to verify about it under Eastern culture. In order to provide some empirical and cross-
cultural evidences, the main purpose of this study is to examine if the 3×2 achievement goal model can be use to
interpret the data draw from different school levels in Taiwan. In addition, we also interested in testing if this model
is better than dichotomy, trichotomy, and 2 ×2 achievement goal models. If gender differences exist on achievement
goal pattern of mathematics also been discussed.

2. Method

There were 373 eighth grade junior high (90 joined in pilot test, 283 joined in formal study) and 146 fifth
elementary school students (40 joined in pilot test, 106 joined in formal study) in Taiwan involved in this study. In
order to attain the purpose of this study, the 3×2 achievement goals questionnaire developed by Elliot et al. (2011)
was used to measure students' achievement goals, after been approved by the author. There are six dimensions in the
3×2 achievement goals questionnaire, each dimension has three items with Likert's 7-point scale ranging from 1 (not
true of me) to 7 (extremely true of me). Due to its property of domain-specific, the Mandarin version of 3×2
achievement goals questionnaire focus on mathematics. After translation and back-translation, the pilot tests were
implemented to assure that Mandarin version of 3×2 achievement goal questionnaire is accurate in wording.
After the reliability of the Mandarin version of 3×2 achievement goal questionnaire have been confirmed. The
confirmatory factor analysis is used to investigate if the 3×2 achievement goal model fit the data draw from different
424 Chun-Chin Wu / Procedia - Social and Behavioral Sciences 69 (2012) 422 – 427

school levels. In order to choose the best model, the 3×2 achievement goals model has been compared to the other
three competing models (dichotomy, trichotomy, 2×2 achievement goal models). Finally, we categorized the
responses of students into five patterns: half of approach and avoidance, strong approach and weak avoidance,
strong avoidance and weak approach, uni-approach, and uni-avoidance. The chi-square test was used to examine if
gender differences exist in on the pattern of achievement goals.

3. Results

There are 275 junior high and 102 elementary school students including in final data analysis because of missing
data. Responses to the 3×2 achievement goals questionnaires from both different school levels are shown to have
good reliability of internal consistencies. The Cronbach's Į coefficient ranges from .90 to.95 calculating from junior
high school students' data, and .75 to.83 calculating from elementary school students' data (see Table 1). Chiu (2010)
suggests that the middle value of 7-point rating scale should be roughly about 4, and standard deviation have to
larger than 1.17. The middle value of the 3×2 achievement goal questionnaire calculating from junior high school
students' data ranges from 4.34 to 4.61, and standard deviation ranges from 1.62 to 1.75. The middle value of the
3×2 achievement goal questionnaire calculating from elementary school students' data ranges from 4.72 to 5.32, and
standard deviation ranges from 1.39 to 1.65 (also see Table 1). In General, these results show good quality of our
Mandarin version of 3×2 achievement goal questionnaire.

Table 1 Descriptive statistics and internal consistencies of different school levels

Junior high school students Elementary school students

Variable
M SD Cronbach's Į M SD Cronbach's Į
Task-approach goal 4.42 1.68 .912 5.07 1.39 .832
Task-avoidance goal 4.50 1.63 .896 5.24 1.46 .750
Self-approach goal 4.37 1.82 .952 4.88 1.65 .860
Self-avoidance goal 4.50 1.62 .896 5.32 1.65 .803
Other-approach goal 4.34 1.75 .949 4.72 1.44 .768
Other-avoidance goal 4.61 1.65 .919 5.31 1.40 .812

In order to examine the construct validity of the 3×2 achievement goal model, different indices are used to
evaluate the fit of models. As Wang, Shih, and Li (2007) recommended, these indices including the goodness of fit
index (GFI), the root-mean-square error of approximation (RMSEA), the normed fit index (NFI), the comparative fit
index (CFI) and the Tucker–Lewis index (TLI). In addition, we used diverse criteria suggested by Kaplan (2009) to
select the best model between dichotomous, trichotomous, 2×2, and 3×2 achievement goal models, these criteria
including Akaike information criterion (AIC) , the Bayesian information criterion (BIC) and the expected cross-
validation index (ECVI). The following cutoff criteria were used to evaluate the model fit: GFI • .90, RMSEA ” .06
(Hu & Bentler, 1999), NFI • .90, CFI • .90, TLI • .90, and the model with smallest values of AIC, BIC and ECVI is
regarded as the best fit model (Kaplan, 2009). Finally, ǻȤ2 are the values of one model subtract the baseline model
(the 3×2 achievement goals model). If ǻȤ2 is significant, it suggests that the 3×2 achievement goals model is the best
fit model.
As it is indicated in Table2, these indices of model fit calculating from junior high school students' data shown
good fit of baseline model (GFI, NFI, CFI and TLI • .90, and RMSEA ” .06). The AIC, BIC and ECVI of baseline
model are entirely smaller than the other three competing models (2×2, trichotomy and dichotomy model). The ǻȤ2
also show the significant differences between baseline model and the other three competing models.
 Chun-Chin Wu / Procedia - Social and Behavioral Sciences 69 (2012) 422 – 427 425

Table 2 Indices of hypothesized and competing models: Junior high school students

model Ȥ2 df ǻȤ2 GFI RMSEA NFI CFI TLI AIC BIC ECVI
baseline 236.17** 120 .91 .059 .96 .98 .98 338.17 522.63 1.23
2×2 354.43** 129 118.26** .85 .080 .95 .96 .96 438.43 590.33 1.60
Trichotomy 377.35** 132 128.79** .85 .082 .94 .96 .96 355.35 596.41 1.66
Dichotomy 468.03** 134 219.48** .81 .095 .93 .95 .94 542.03 675.85 1.98
Note. Baseline = 3×2 achievement goal model; 2×2 = 2×2 achievement goal model; Trichotomy = trichotomous
achievement goal model; Dichotomy = dichotomous achievement goal model; N = 275.
**
p .01
In Table 3, as we can see, GFI ” .9, RMSEA .6, there are only NFI, CFI, TLI .9 . These indices indicate
that the baseline model not fit very well to the data draw from elementary school students. Moreover, this model
reveals one modification index. If we modify this relationship, independent assumption of observation variables and
error will be violated (Kaplan, 2009). Generally, as we can see, the values of AIC, BIC, and ECVI from baseline
model which show this model is more suitable for interpreting the elementary school students' data than the other
competing models. The ǻȤ2 also show the significant differences between baseline model and the other three
competing models.

Table 3 Indices of hypothesized and competing models: Elementary school students

model Ȥ2 df ǻȤ2 GFI RMSEA NFI CFI TLI AIC BIC ECVI
baseline 248.55** 120 .81 .070 .94 .97 .96 350.55 444.52 3.08
2×2 354.43** 129 105.88** .84 .080 .95 .96 .96 438.43 461.76 3.48
Trichotomy 377.35** 132 128.79** .74 .082 .94 .96 .96 355.35 471.39 3.65
Dichotomy 468.03** 134 219.48** .66 .095 .93 .95 .94 542.03 559.05 4.57
Note. Baseline = 3×2 achievement goal model; 2×2 = 2×2 achievement goal model; Trichotomy = trichotomy
achievement goal model; Dichotomy = dichotomy achievement goal model; N = 102.
**
p .01

Finally, according to the data, we classify the 3×2 achievement goals into five patterns: half of approach and
avoidance, strong approach and weak avoidance, strong avoidance and weak approach, uni-approach, and uni-
avoidance. The chi-square values range from .016 to 2.63, which are all smaller than the cutoff value 6.635 with
degree of freedom 1 and significant level .01 ( Ȥ2.99(1) = 6.635 ). Substantially, there is no gender difference in each
3×2 achievement goals patterns.

4. Conclusion and Discussion

At first, the Mandarin version of have good reliability, construct validity, and moderate difficulty for junior high
students. For elementary school students, the questionnaire also shown good reliability, but less good in the
construct validity, and it is slightly difficult (the mean value are higher).
Second, the intercorrelations of 3×2 achievement goals are higher in the elementary school students' data.
Specifically, comparing to junior high school students, elementary school students may not well discriminate
between the 3×2 achievement goals. This result may imply the elementary school students have not well develop in
their cognitive ability, and cannot clearly discriminate between 3×2 achievement goals as well as junior high school
students. It means the developmental differences between two groups (Pintrich, 2000). A few similar words are used
in the 3×2 achievement goal questionnaire may be also another reason which can be used to interpret the high
intercorrelations.
Third, as we noted earlier, the 3×2 achievement goal model fit the junior high school students' data, but
elementary school students' data not fit as well. Contrasting the 3×2 achievement goal model to the other three
competing models, the 3×2 achievement goal model is the best fit, which can be used for interpreting junior and
426 Chun-Chin Wu / Procedia - Social and Behavioral Sciences 69 (2012) 422 – 427

elementary school students' achievement goals. Specifically, the 3×2 achievement goal model is more accurate in its
competence-related definition (task, self, and other) than the other three competing models (only contained mastery
and performance). Therefore, the Mandarin version of 3×2 achievement goal questionnaire can be used for better
measuring the junior high and elementary school students in Taiwan. But the questionnaire may have to revise some
wording for better measuring the achievement goals of elementary school students.
Finally, the results from chi-square tests are shown no gender differences in the patterns of 3×2 achievement goal.
This result is unexpected for two reasons. The first reason is that we find girls are more dislike mathematics than
boys whatever their school levels in Taiwan. The second reason is the stereotype of boys may be more competitive
than girl (Schunk et al., 2008). As a result, the amount of boys' other-approach goal pattern should be higher than
girls.
For explaining this phenomenon, we propose it may due to the parents and teachers usually remind girls that
mathematics is very important for their future. Therefore, getting good grades for parents, and conform to teachers'
expectation or requirements may become their social goals or social motivation (Wentzel, 1999, 2000), then affect
their achievement goal toward mathematics in turn. If girls do so, it may imply that girls abandon their individual
goals to pursue the social goals which are transmitted from parents or teachers. This may be the reason why there is
no gender difference between boys' and girls' achievement goals toward mathematics in junior high school and
elementary school students in Taiwan.

5. Limitations and Future directions

Although, the 3×2 achievement goal questionnaire may not entirely suit for measuring and interpreting the
achievement goals of elementary school students. But this model is the best fit than the other three competing
models. For future researches on measuring the achievement goals of elementary school students, we suggest
researchers can modify the wording which is used in the present Mandarin version of 3×2 achievement goal
questionnaire. In addition to the wording, more researches are needed to investigate if there are other goals different
from achievement goals which elementary students may pursue. Specifically, elementary school students may be
more likely to be influenced by parents, teachers, and peers. Therefore, the achievement goal may partially interpret
students' achievement goals. For elementary school students, they may pursue achievement goals for social reason
or from social motivation. Because the sense of relatedness related to social reason or social motivation is very
important in Chinese culture. Therefore, future research can incorporate social motivation into achievement goals
questionnaire to examine if it can be used to better interpret students' achievement goals.

References

Cherng, B.-L. (2003). Study of the model of 4 dimensions goal orientation. Journal of Taiwn Normal University:
Education, 48(1), 15-40.
Chiu, H.-C. (2010). Quantitative research and statistics analysis: Examples of SPSS(PASW) data analysis. Taipei:
Wu-Nan Culture Enterprise.
Elliot, A. J., & Harackiewicz, J. M. (1996). Approach and avoidance achievement goals and intrinsic motivation: A
mediational analysis. Journal of Personality and Social Psychology, 70, 461-475.
Elliot, A. J., & McGregor, H. A. (2001). A 2 x 2 achievement goal framework. Journal of Personality and Social
Psychology, 80(3), 501-519. doi: 10.1O37//OO22-3514.80.3.501
Elliot, A. J., & Murayama, K. (2008). On the measurement of achievement goals: Critique, illustration, and
application. Journal of Educational Psychology, 100(3), 613-628.
Elliot, A. J., Murayama, K., & Pekrun, R. (2011). A 3 × 2 achievement goal model. Journal of Educational
Psychology, 103(3), 632-648. doi: 10.1037/a0023952
Higgins, E. T. (1997). Beyond pleasure and pain. American Psychologist, 52(12), 1280-1300.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria
versus new alternatives. Structural Equation Modeling, 6, 1-55.
Huang, C. (2011). Achievement goals and achievement emotions: A meta-analysis. Educational Psychology Review,
23(3), 359-388. doi: 10.1007/s10648-011-9155-x
 Chun-Chin Wu / Procedia - Social and Behavioral Sciences 69 (2012) 422 – 427 427

Huang, C. (2012). Discriminant and criterion-related validity of achievement goals in predicting academic
achievement: A meta-analysis. Journal of Educational Psychology, 104(1), 48-73. doi: 10.1037/a0026223
Jang, L. Y., & Liu, W. C. (2012). 2 x 2 achievement goals and achievement emotions: A cluster analysis of students'
motivation. European Journal of Psychology of Education, 27(1), 59-76. doi: 10.1007/s10212-011-0066-5
Kaplan, D. (2009). Structural equation modeling: Foundations and extensions (2nd ed.). Los Angeles: SAGE.
Maehr, M. L., & Zusho, A. (2009). Achievement goal theory: The past, present, and the future. In K. R. Wentzel &
A. Wigfield (Eds.), Handbook of motivation at school (pp. 77-104). New York: Routledge.
Pintrich, P. R. (2000). An achievement goal theory perspective on issues in motivation terminology, theory, and
research. Contemporary Educational Psychology, 25, 92-104.
Schunk, D. H., Pintrich, P. R., & Meece, J. L. (2008). Motivation in education : Theory, research, and applications.
Upper Saddle River, N.J.: Pearson/Merrill Prentice Hall.
Senko, C., Durik, A. M., & Harackiewicz, J. M. (2008). Historical perspectives and new directions in achievement
goal theory: Understanding the effects of mastery and performance-approach goals. In J. Y. Shah & W. L.
Gardner (Eds.), Handbook of motivation science (pp. 100-113). New York: Guilford Press.
van Yperen, N. W., & Renkema, L. J. (2008). Performing great and the purpose of performing better than others: On
the recursiveness of the achievement goal adoption process. European Journal of Social Psychology, 38(2),
260-271. doi: 10.1002/ejsp.425
Wang, H.-F., Shih, C.-T., & Li, M.-N. (2007). Factorial validity test of theoy constructs: First order CFA model In
M.-N. Li (Ed.), An introduction to AMOS and its uses in scale development: Graphics & basic (pp. 157-
202). Taipei: Psychology.
Wentzel, K. R. (1999). Social-motivational processes and interpersonal relationships: Implications for understanding
motivation at school. Journal of Educational Psychology, 91(1), 76-97.
Wentzel, K. R. (2000). What is it that I'm trying to achieve? Classroom goals from a content perspective.
Contemporary Educational Psychology, 25, 105-115.