Article Type:

Report

Subject:

College students
(Psychological aspects)

Motivation in education (Analysis)

Learning strategies (Analysis)

Teaching (Equipment and supplies)

Teaching (Usage)

Motivation in education (Analysis)

Learning strategies (Analysis)

Teaching (Equipment and supplies)

Teaching (Usage)

Authors:

Alkharusi, Hussain

Neisler, Otherine

Barwani, Thuwayba Al-

Clayton, David

Sulaimani, Humaira Al-

Khan, Mohammad

Yahmadi, Hamad Al-

Kalbani, Muna Al-

Neisler, Otherine

Barwani, Thuwayba Al-

Clayton, David

Sulaimani, Humaira Al-

Khan, Mohammad

Yahmadi, Hamad Al-

Kalbani, Muna Al-

Pub Date:

09/01/2012

Publication:

Name: College Student Journal Publisher: Project Innovation (Alabama) Audience: Academic Format: Magazine/Journal Subject: Education Copyright: COPYRIGHT 2012 Project Innovation (Alabama) ISSN: 0146-3934

Issue:

Date: Sept, 2012 Source Volume: 46 Source Issue: 3

Topic:

Event Code: 310 Science & research

Product:

Product Code: E197500 Students, College

Organization:

Organization: Sultan Qaboos University

Geographic:

Geographic Scope: Oman Geographic Code: 7OMAN Oman

Accession Number:

302464020

Full Text:

Research on university student learning has suggested that
students' motivational orientations and learning strategies are
critical to their academic achievement. In an attempt to assess the
levels of motivation and the usage of learning resources and strategies
of college students, Pintrich, Smith, Garica and McKeachie (1993)
developed an 81-item instrument entitled the Motivated Strategies for
Learning Questionnaire (MSLQ) based on the social cognitive view of
motivation and learning. This paper examines the psychomteric properties
of the MSLQ for Sultan Qaboos University. A sample of 952 students (506
females and 446 males) who were admitted at Sultan Qaboos University in
2010 completed the MSLQ. Confirmatory factor analyses indicated that
first-order factor structures of a reduced form (71 items) of the MSLQ
provided a good fit to the observed data compared to the second-order
factor structures. The scores from the reduced form of the MSLQ
exhibited acceptable levels of internal consistency compared to the long
form of 81-items. Overall, the results indicate that MSLQ provides
reliable and valid interpretations of the motivational orientations and
usage of learning strategies and resources of students entering SQU.

Keywords: validity, reliability, confirmatory factor analysis, MSLQ, university students

Introduction

Academic success of university students has been a concern for many educators in Oman. Several examinations of the General Education System in Oman have revealed that students were not academically prepared for the high expectations of university education (Ministry of Education, 1994, 1995, 2007). In addition, many studies have been conducted at Sultan Qaboos University (SQU), which is the only government university in Oman, to identify the factors affecting student success in the university. For example, Al-Shahwarzi, Al-Ghufaili, and Al-Jabri (1991) found no significant correlation between secondary school scores and university grade point averages (GPA). Likewise, Ibrahim, Yahya, and Al-Barwani (1992) found that secondary school scores alone were not a good predictor of university GPA. These results suggest that doing well in secondary school did not mean that students would succeed in university studies. Other personal factors might affect the academic success in the university. In a survey of the university faculty about the basic cognitive skills and attitudes needed for academic success, Al-Barwani, Yahya, and Ibrahim (1997) concluded that first year students tended to have a deficiency in 88% of the skills and attitudes needed for academic success. Along similar lines, Yahya (1997) found that academic motivation might play a role in the academic achievement of the university students. Taken together these SQU studies explained the need for a reliable and valid measure of the motivation- and learning-related factors that might affect the academic success of the university students.

Research on university student learning has suggested that students' motivational orientations and learning strategies are critical to their academic achievement (Artino, 2005; Pintrich & DeGroot, 1990). In an attempt to assess the motivational levels and the usage level of learning resources and strategies of the college students, Pintrich, Smith, Garica and McKeachie (1993) developed an 81-item instrument entitled the Motivated Strategies for Learning Questionnaire (MSLQ). The design of the MSLQ was based on the social cognitive view of motivation and learning. The motivation section included the components of value, expectations, and affect. The learning section included the components of cognitive-metacognitive strategies and resource management use. Many studies have established relationships between academic achievement and a combination of self-regulated learning skills, motivation, and metacognitive strategies as measured by MSLQ (e.g., Davenport, 1999; Garcia & Pintirch, 1993; Howey, 1999; Pintrich, 2000).

Since its development, there has been an increased interest in utilizing the MSLQ to understand students' motivational orientations and learning strategies in various countries of the world such as China (e.g., Rao, Moely, & Sachs, 2000), Germany (e.g., Neber & Heler, 2002), Greek (e.g., Andreo & Metallidou, 2004), and South Africa (e.g., Watson, McSorley, Foxcroft & Watson, 2004). Despite the extensive amount of research utilizing the MSLQ in various parts of the world, little is known about the validity and reliability of its scores when applied with university students in Oman. In addition, there is a paucity of research examining the higher-order latent structure underlying the responses to the MSLQ. Hence, the primary objective of the present study is to provide a rigorous psychometric assessment in terms of validity, reliability, and norms of the MSLQ's scores for SQU students.

Purposes of the Study

Given the growing interest in the academic success of university students, research is needed to support the assertion that score interpretations from the available instruments such as the MSLQ are accurate representations of the university students' motivational orientations and learning resources and strategies. Therefore, this study is part of a larger strategic project funded by His Majesty Grant to investigate college readiness of students entering SQU. The study sought evidence of the MSLQ score validity and reliability by:

1. Testing the first- and second-order latent factor structures of the MSLQ's scores for SQU students.

2. Providing evidence regarding the internal consistency of the MSLQ's scores for SQU students.

3. Extracting the norms for SQU students' raw scores on the MSLQ.

Methods

Sample

The sample for this study included 952 Omani students selected conveniently from the population of students entering SQU in September 2010. There were 506 females and 446 males. The participants ranged in age from 17 to 25 with a mean of 18.55 and a standard deviation of .64.

Instrumentation

The primary instrument in this study was The Motivated Strategies for Learning Questionnaire (MSLQ) developed by Pintrich et al. (1993) to measure motivational orientations and learning resources and strategies of college students. It consisted of 81 items divided into two sections: a motivation section and a learning resources and strategies section. The motivation section consisted of 31 items assessing students' intrinsic goal orientation (4 items, [alpha] = .74), extrinsic goal orientation (3 items, [alpha] = .62), task value (6 items, [alpha] = .90), control beliefs about learning (4 items, [alpha] = .68), self-efficacy (8 items, [alpha] = .93), and test anxiety (5 items, [alpha] = .80). The learning resources and strategies section consisted of 50 items assessing students' use of rehearsal (4 items, [alpha] = .69), elaboration (6 items, [alpha] = .75), organization (4 items, [alpha] = .64), critical thinking (5 items, [alpha] = .80), metacognitive self-regulation (12 items, [alpha] = .79), time and study environment (8 items, [alpha] = .76), effort regulation (4 items, [alpha] = .69), peer learning (3 items, [alpha] = .76), and help seeking (4 items, [alpha] = .52). Responses were obtained on a 7-point Likert scale ranging from 1 (not at all true of me) to 7 (very true of me). Scoring of the negative items was reversed so that higher scores reflected a more positive motivational orientation and use of learning strategies. The items were translated from English to Arabic by the research team.

Procedures

Permission to collect data from the students during the orientation week was obtained from the University Administration. The research team informed the students that a study was being conducted on the motivational orientations and learning resources and strategies adopted by university students. Emphasis was placed on the fact that information to be gathered would not influence their admission in the university in any way and that the study would hopefully lead to improved learning in the university. Students who wished to participate in the study were given the MSLQ. Total time for completing the instrument averaged approximately twenty minutes.

Statistical Analyses

The first phase of the analysis was to test the first-order factor structures of the MSLQ. Two confirmatory factor analyses (CFA) with maximum likelihood estimation in EQS 6.1 were conducted on the MSLQ item data: one for the set of motivation items and another for the set of learning resources and strategies items. The analysis was conducted using the covariance matrix. Each item was constrained to load only on its hypothesized factor described by Pintrich et al. (1993). One item on each factor was constrained to equal one in order to set a metric for the factors. Factor covariances were left free to be estimated, but the measurement errors were not allowed to covary. During this phase, items that resulted in large standardized residuals were identified for possible deletion. Items were deleted one at a time in order to study changes in the factor structure solution. The purpose of this item deletion was to develop the best fitted and the most parsimonious model to the data, while maintaining the original factor structure.

The second phase of the analysis involved developing two second-order factor structures. As implied by Pintrich et al. (1993), one second-order factor model tested the existence of the motivation structure, whereas the other tested the existence of the learning resources and strategies structure. All covariances among the first-order factor residuals were constrained to equal zero. A matrix of the loadings of the first-order factors on the second-order factors was set up in which all subscales of the MSLQ were allowed to load freely on their hypothesized second-order factors. The variances of the second-order factors were constrained to equal 1.

In evaluating the fit of the models in all CFAs, recommendations by Schermelleh-Engel, Moosbrugger, and Muller (2003) were followed. These recommendations state that for an acceptable model fit, the ratio [X.sup.2] / df should be less than or equal to 3, the Root Mean Square Error of Approximation (RMSEA) should be less than or equal to .08, the Nonnormed Fit Index (NNFI) that is also called the Tuker-Lewis Index (TLI) should be greater than or equal to .95, and the Comparative Fit Index (CFI) should be greater than or equal to .95 (Schermelleh-Engel et al., 2003). The RMSEA, NNFI, and CFI were chosen because they were found to be less affected by the size of the sample when compared to the Normative Fit Index (NFI), the Goodness-of-Fit Index (GFI), and the Adjusted Goodness-of-Fit Index (AGFI) (Schermelleh-Engel et al., 2003).

The third phase of the analysis involved estimating the internal consistency reliability of the MSLQ's scores resulting from the CFAs obtained in the previous phases. Finally, percentile ranks were calculated as norms for interpreting the raw scores of the sample on MSLQ.

Results

First-Order Factor Structures

Motivation model. A CFA was first used to evaluate the dimensionality of the motivation model by constraining the 31 items to fall onto the six motivational latent factors: intrinsic goal orientation (4 items), extrinsic goal orientation (3 items), task value (6 items), control beliefs about learning (4 items), self-efficacy (8 items), and test anxiety (5 items). Results yielded an inferential test of = 1170.05 (p = .00, df = 419) with the following descriptive fit indices (RMSEA = .04 with 90% CI = [.04-.05], NNFI = .88, and CFI = .89). Although the fit indices compare favorably to those reported in Davenport (1999) and Pintrich et al. (1993) and all of the item parameters were statistically significant, the fit indicators pointed out that the 31-item motivation model was not fit at the expected level.

Thus, an attempt was made to modify the first-order 31-item motivation model to fit the data. Items with large standardized residuals were identified for deletion. Items were deleted one at a time until an acceptably fitting model was obtained. In the deletion of items, it was of interest to maintain the original six factor structure of motivation with a minimum of three items per factor to avoid identification problems. This process resulted in a 28-item model with an inferential test of = 953.22 (p = .00, df = 335) and with satisfactory descriptive fit indices (RMSEA = .04 with 90% CI = [.04-.05], NNFI = .90, and CFI = .91).

Table 1 presents standardized factor loadings for this first-order 28-item model of motivation along with factor intercor-relations. The first factor explained between 9 and 38%, the second factor accounted for between 29 and 42%, the third factor explained between 23 and 48%, the fourth factor accounted for between 9 and 40%, the fifth factor explained between 25 and 53%, and the sixth factor accounted for between 13 and 40% of the variance in the item data. All factor loadings were statistically significant at p < .01. As expected, the positive motivational factors (intrinsic goal orientation, task value, control beliefs of learning, and self-efficacy) are all correlated positively with one another with correlations ranging from .42 to .68. Test anxiety was negatively correlated with the positive motivational beliefs with correlations ranging from .09 to .23, but not significantly correlated with extrinsic goal orientation, which is a less desirable motivational characteristic (Pintrich et al., 1993).

Learning model. A CFA was first used to evaluate the dimensionality of the learning model by constraining the 50 items to fall onto the nine learning latent factors: students' use of rehearsal (4 items), elaboration (6 items), organization (4 items), critical thinking (5 items), self-regulation (12 items), time and study environment (8 items), effort regulation (4 items), peer learning (3 items), and help seeking (4 items). Results yielded an inferential test of = 3385.69 (p = .00, df = 1139) with the following descriptive fit indices (RMSEA = .05 with 90% CI = [.04-.05], NNFI = .82, and CFI = .84). Although the fit indices compare favorably to those reported in Davenport (1999) and Pintrich et al. (1993) and all of the item parameters were statistically significant, the fit indicators pointed out that the 50-item learning model was not fit at the expected level.

Thus, an attempt was made to modify the first-order 50-item learning model to fit the data. Items with large standardized residuals were identified for deletion. Items were deleted one at a time until an acceptably fitting model was obtained. In the deletion of items, it was of interest to maintain the original nine factor structure of learning with a minimum of three items per factor to avoid identification problems. This process resulted in a 43-item model with an inferential test of = 2152.13 (p = .00, df = 824) and with satisfactory descriptive fit indices (RMSEA = .04 with 90% CI = [.04-.05], NNFI = .90, and CFI = .90).

Table 2 presents standardized factor loadings for this first-order 43-item model of learning along with factor intercor-relations. The first factor explained between 25 and 44%, the second factor accounted for between 28 and 46%, the third factor explained between 34 and 45%, the fourth factor accounted for between 20 and 36%, the fifth factor explained between 28 and 40%, the sixth factor accounted for between 10 and 45%, the seventh factor explained between 9 and 46%, the eighth factor accounted for between 18 and 46%, and the ninth factor explained between 27 and 45% of the variance in the item data. All factor loadings were statistically significant at p < .01. As expected, all cognitive strategies, metacognitive strategies, and resource management subscales were positively related to one another with correlations ranging from. 17 to .77.

Second-Order Factor Structures

Motivation model. As shown in Table 1, the first-order factors of the motivation section of the MSLQ are correlated with one another. It may be hypothesized that the correlations among the factors can be accounted for by second-order factors. As such, based on the first-order factor model of motivational orientations, an alternative second-order factor model was developed to test the existence of the motivation structure implied by Davenport (1999) and Pintrich et al. (1993). This structure included two second-order factors, which were Expectancy and Value. Intrinsic goal orientation, extrinsic goal orientation, and task value were loaded on the second-order Value factor. Self-efficacy and control of learning beliefs were loaded on the second-order Expectancy factor. As stated by Davenport (1999), test anxiety was hypothesized to be a function of both expectations and value of success, and as such it was cross-loaded on both second-order factors. Results yielded an inferential test of = 983.94 (p = .00, df = 334) with the following descriptive fit indices (RMSEA = .05 with 90% CI = [.04-.05], NNFI = .89, and CFI = .90). It seems that the first-order factor model of motivation represents the sample data much better than the third-order factor model.

In an attempt to evaluate the fit of the second-order factor model in relation to the first-order model, Marsh and Hocevar (1985) designed a goodness-of-fit index called the target coefficient (T), which is the ratio of the chi-square value from the first-order model to that of the second-order model. A value of 1.0 indicates that the covariation among the first-order factors is completely accounted for by the second-order model. Thus, based on the chi-square values obtained for the first-and second-order factor models of motivation, the value of T was .97, suggesting that 97% of the covariation among the six first-order factors of the motivational orientations can be accounted for by the two second-order factors of Expectancy and Value.

Learning model. As shown in Table 2, the first-order factors of the learning strategies section of the MSLQ are correlated with one another. It may be hypothesized that the correlations among the factors can be accounted for by second-order factors. As such, based on the first-order factor model of learning strategies, an alternative second-order model was developed to test the existence of the learning strategies structure implied by Davenport (1999) and Pintrich et al. (1993). This structure included two second-order factors, which were Learning Strategies and Resource Use. Rehearsal, elaboration, organization, critical thinking, and self-regulation were loaded on the second-order Learning Strategies factor. Time and study environment, effort regulation, peer learning, and help seeking were loaded on the second-order Resource Use factor. Results yielded an inferential test of = 2410.14 (p = .00, df = 839) with the following descriptive fit indices (RMSEA = .04 with 90% CI = [.04-.05], NNFI = .87, and CFI = .88). It seems that the first-order factor model of learning represents the sample data much better than the second-order factor model. Using Marsh and Hocevar's (1985) goodness-of-fit index, 89% of the covariation among the nine first-order factors of the learning strategies can be accounted for by the two second-order factors of Learning Strategies and Resource Use.

Internal Consistency Reliability

Based upon the first-order factor analysis, it appears that the reduced 71-item MSLQ models provide a reasonable fit to the data. As such, for comparative purposes, Table 3 presents internal consistency reliabilities as measured by Cronbach's alpha for the sample's scores on the short and long versions of the MSLQ. As shown in Table 3, the reliabilities of the short version's scores seem comparable to those of the long version of the MSLQ. Although the magnitude of the score reliability is relatively lower than that reported by Pintrich et al. (1993), no precision was lost in the short version obtained in this study compared to the long version. This suggests that the motivational orientations and learning resources and strategies of students entering SQU can be effectively represented by a smaller number of items while keeping the same degree of measurement precision as the long version 81-item MSLQ.

Norms

Percentile ranks were extracted as norms for the raw scores of the participants to indicate the relative position of an individual student in the sample. Table 4 shows means and standard deviations along with the percentile ranks for the scores on the MSLQ. As seen in Table 4, the means of the participants on the MSLQ subscales were above the scales' midpoints, suggesting that, on average, the participants tended to adopt desirable motivational orientations and learning resources and strategies. The norms of the present study's sample were slightly higher than the norms reported in Pintrich, Smith, Garcia, and McKeachie's (1991) study.

Conclusion

Students' positive motivational orientations and desirable use of learning resources and strategies have been recognized as the most influential factors of the academic success of the university students (Pintrich et al., 1993). To adequately assess these motivational and learning factors, Pintrich et al. (1993) developed an 81-item instrument entitled the Motivated Strategies for Learning Questionnaire (MSLQ). There has been an increased interest in utilizing the MSLQ to understand students' motivational characteristics and use of learning resources and strategies around the world (Duncan & Mckeachie, 2005). However, the measurement properties of this instrument have not been explored statistically when applied on university students in Oman. The present study reports on the validity and reliability of the MSLQ's scores for SQU students in Oman. The importance of this study stemmed from Messick's (1995) assertion that establishing evidence of score meaning is a continual and evolving process. Additionally, the significance of the study comes from the concern over the inadequate level of motivation and usage of learning strategies possessed by students entering SQU (Al-Barwani et al., 1997; Yahya, 1997).

The current findings indicated that a reduced version of MSLQ consisting of 71 items displayed conceptually sound first-order factor structures of motivational orientations and use of learning resources and strategies as well as reasonable levels of internal consistency. From a practical point of view, the development of a shortened MSLQ would be of benefit to researchers who are conducting studies in which MSLQ's constructs are one of several constructs being measured. Overall, the results indicate that MSLQ provides reliable and valid interpretations of the motivational orientations and usage of learning resources and strategies of students entering SQU. This makes MSLQ a valuable assessment tool for university educators to help academically unprepared students entering the university. Specifically, the MSLQ can be administered during the orientation week to identify academically unprepared students who got low scores on the MSLQ. Appropriate developmental instructional programs can then be designed to enhance their motivational orientations and equip them with the learning skills necessary for a successful beginning to their university education. Early identification of academically at-risk students may improve students' academic success and retention rates (Adelman, 1986; Jacobi, Astin, & Ayala, 1987).

Finally, the generalizability of the study findings was limited by the non-random sample used in the study. Also, a larger sample size will enable future researchers to test whether the factorial structure of the MSLQ is invariant across males and females. Future research might consider examining the changes in students' motivational orientations and usage of learning resources and strategies across time during their university education.

References

Adelman, C. (1986). Assessment in American higher education. Washington, DC: Department of Education.

Al-Barwani, T. H., Yahya, A., & Ibrahim, A. (1997). Basic skills and attitudes for university study as perceived by faculty members at Sultan Qaboos University. Series of Educational and Psychological Studies, 2, 39-63.

Al-Shahwarzi, J. F., AlGhufaili, H., & Al-Jabri A.M. (1991). Factors that influence the success of students in Sultan Qaboos University. College of Science, Sultan Qaboos University.

Andreou, E., & Metallidou, P. (2004). The relationship of academic and social cognition to behaviour in bullying situations among Greek primary school children. Educational Psychology, 24, 27-41.

Artino, A. R., Jr. (2005). Review of the Motivated Strategies for Learning Questionnaire. Manuscript submitted for publication.

Davenport, M. A. (1999). Modeling motivation and learning strategy use in the classroom: An assessment of the factorial, structural, and predictive validity of the Motivated Strategies for Learning Questionnaire. Unpublished doctoral dissertation, Auburn University, Alabama.

Duncan, T. G., & McKeachie, W. J. (2005). The making of the Motivated Strategies for Learning Questionnaire. Educational Psychologist, 40, 117-128.

Garcia, T., & Pintrich, P. R. (1993). Self-schemas, motivational strategies and self-regulated learning. Paper presented at the annual meeting of the American Educational Research, Atlanta (ERIC Document No.ED359234).

Howey, S. C. (1999). The relationship between motivation and academic success at community college freshmen orientation students. Unpublished doctoral dissertation, Kansas State University, Kansas.

Ibrahim, A., Yahya, A., & Al-Barwani, T. (1992). A Study of the Omani secondary school certificate examination as a predictor of academic performance at Sultan Qaboos University. Series of educational and Psychological Studies, 2.

Jacobi, M., Astin, A., & Ayala, E, Jr. (1987). College student outcomes assessment: A talent development perspective. ASHE-ERIC Higher Education Report No. 7. Association for the Study of Higher Education: Washington, D.C.

Marsh, H. W., & Hocevar, D. (1985). Application of confirmatory factor analysis to the study of self-concept: First and higher-order factor models and their invariance across groups. Psychological Bulletin, 97, 562-582.

Messick, S. (1995). Validity of psychological assessment: Validation of inferences from persons' responses and performances as scientific inquiry into score meaning. American Psychologist, 50, 741-749.

Ministry of Education. (1994). The educational system in Oman: Achievements and challenges. Sultanate of Oman: Author.

Ministry of Education. (1995). Reform and development of general education in the Sultanate of Oman. Sultanate of Oman: Author.

Ministry of Education. (2007). Basic education: Aims, implementation and evaluation. Sultanate of Oman: Author.

Neber, H., & Heller, K. A. (2002). Evaluation of a summer-school program for highly gifted secondary-school students: The German Pupils Academy. European Journal of Psychological Assessment, 18, 214-228.

Pintrich, P. R. (2000). The role of goal orientation in self-regulated learning. In M. Boekaerts & P. R. Pintrich(Eds.), Handbook of self-regulation (pp.13-39). San Diego, CA: Academic Press.

Pintrich, P. R., & DeGroot, E. V. (1990). Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology, 82, 33-40.

Pintrich, P. R., Smith, D. A. E, Garcia, T., & McKeachie, W. J. (1991). A manual for the use of the Motivated Strategies for Learning Questionnaire (MSLQ). National Center for Research to Improve Postsecondary Teaching and Learning, Ann Arbor, MI. (ERIC Document Reproduction Service No. ED338122)

Pintrich, P. R., Smith, D. A. E, Garcia, T., & McKeachie, W. J. (1993). Reliability and predictive validity of the Motivated Strategies for Learning Questionnaire (MSLQ). Educational and Psychological Measurement, 53, 801-814.

Rao, N., Moely, B. E., & Sachs, J. (2000). Motivational beliefs, study strategies, and mathematics attainment in high- and low-achieving Chinese secondary school students. Contemporary. Educational Psychology, 25, 287-316.

Schermelleh-Engel, K., Moosbrugger, H., & Muller, H. (2003). Evaluating the fit of structural equation models: Tests of significance and descriptive goodness-of-fit measures. Methods of Psychological Research Online, 8, 23-74.

Watson, M., McSorley, M., Foxcroft, C. & Watson, A. (2004). Exploring the Motivation Orientation and Learning Strategies of First Year University Learners. Tertiary Education and Management, 10, 193-207.

Yahya, A. (1997). Achievement motivation as a function of sex, age, and academic achievement of Sultan Qaboos University Students. Series of educational and Psychological Studies, 2, 1-15.

HUSSAIN ALKHARUSI

OTHERINE NEISLER

THUWAYBA AL-BARWANI

DAVID CLAYTON

HUMAIRA AL-SULAIMANI

MOHAMMAD KHAN

Sultan Qaboos University, Oman

HAMAD AL-YAHMADI

Diwan Of Royal Court, Oman

MUNA AL-KALBANI

Ministry of Education, Oman

Keywords: validity, reliability, confirmatory factor analysis, MSLQ, university students

Introduction

Academic success of university students has been a concern for many educators in Oman. Several examinations of the General Education System in Oman have revealed that students were not academically prepared for the high expectations of university education (Ministry of Education, 1994, 1995, 2007). In addition, many studies have been conducted at Sultan Qaboos University (SQU), which is the only government university in Oman, to identify the factors affecting student success in the university. For example, Al-Shahwarzi, Al-Ghufaili, and Al-Jabri (1991) found no significant correlation between secondary school scores and university grade point averages (GPA). Likewise, Ibrahim, Yahya, and Al-Barwani (1992) found that secondary school scores alone were not a good predictor of university GPA. These results suggest that doing well in secondary school did not mean that students would succeed in university studies. Other personal factors might affect the academic success in the university. In a survey of the university faculty about the basic cognitive skills and attitudes needed for academic success, Al-Barwani, Yahya, and Ibrahim (1997) concluded that first year students tended to have a deficiency in 88% of the skills and attitudes needed for academic success. Along similar lines, Yahya (1997) found that academic motivation might play a role in the academic achievement of the university students. Taken together these SQU studies explained the need for a reliable and valid measure of the motivation- and learning-related factors that might affect the academic success of the university students.

Research on university student learning has suggested that students' motivational orientations and learning strategies are critical to their academic achievement (Artino, 2005; Pintrich & DeGroot, 1990). In an attempt to assess the motivational levels and the usage level of learning resources and strategies of the college students, Pintrich, Smith, Garica and McKeachie (1993) developed an 81-item instrument entitled the Motivated Strategies for Learning Questionnaire (MSLQ). The design of the MSLQ was based on the social cognitive view of motivation and learning. The motivation section included the components of value, expectations, and affect. The learning section included the components of cognitive-metacognitive strategies and resource management use. Many studies have established relationships between academic achievement and a combination of self-regulated learning skills, motivation, and metacognitive strategies as measured by MSLQ (e.g., Davenport, 1999; Garcia & Pintirch, 1993; Howey, 1999; Pintrich, 2000).

Since its development, there has been an increased interest in utilizing the MSLQ to understand students' motivational orientations and learning strategies in various countries of the world such as China (e.g., Rao, Moely, & Sachs, 2000), Germany (e.g., Neber & Heler, 2002), Greek (e.g., Andreo & Metallidou, 2004), and South Africa (e.g., Watson, McSorley, Foxcroft & Watson, 2004). Despite the extensive amount of research utilizing the MSLQ in various parts of the world, little is known about the validity and reliability of its scores when applied with university students in Oman. In addition, there is a paucity of research examining the higher-order latent structure underlying the responses to the MSLQ. Hence, the primary objective of the present study is to provide a rigorous psychometric assessment in terms of validity, reliability, and norms of the MSLQ's scores for SQU students.

Purposes of the Study

Given the growing interest in the academic success of university students, research is needed to support the assertion that score interpretations from the available instruments such as the MSLQ are accurate representations of the university students' motivational orientations and learning resources and strategies. Therefore, this study is part of a larger strategic project funded by His Majesty Grant to investigate college readiness of students entering SQU. The study sought evidence of the MSLQ score validity and reliability by:

1. Testing the first- and second-order latent factor structures of the MSLQ's scores for SQU students.

2. Providing evidence regarding the internal consistency of the MSLQ's scores for SQU students.

3. Extracting the norms for SQU students' raw scores on the MSLQ.

Methods

Sample

The sample for this study included 952 Omani students selected conveniently from the population of students entering SQU in September 2010. There were 506 females and 446 males. The participants ranged in age from 17 to 25 with a mean of 18.55 and a standard deviation of .64.

Instrumentation

The primary instrument in this study was The Motivated Strategies for Learning Questionnaire (MSLQ) developed by Pintrich et al. (1993) to measure motivational orientations and learning resources and strategies of college students. It consisted of 81 items divided into two sections: a motivation section and a learning resources and strategies section. The motivation section consisted of 31 items assessing students' intrinsic goal orientation (4 items, [alpha] = .74), extrinsic goal orientation (3 items, [alpha] = .62), task value (6 items, [alpha] = .90), control beliefs about learning (4 items, [alpha] = .68), self-efficacy (8 items, [alpha] = .93), and test anxiety (5 items, [alpha] = .80). The learning resources and strategies section consisted of 50 items assessing students' use of rehearsal (4 items, [alpha] = .69), elaboration (6 items, [alpha] = .75), organization (4 items, [alpha] = .64), critical thinking (5 items, [alpha] = .80), metacognitive self-regulation (12 items, [alpha] = .79), time and study environment (8 items, [alpha] = .76), effort regulation (4 items, [alpha] = .69), peer learning (3 items, [alpha] = .76), and help seeking (4 items, [alpha] = .52). Responses were obtained on a 7-point Likert scale ranging from 1 (not at all true of me) to 7 (very true of me). Scoring of the negative items was reversed so that higher scores reflected a more positive motivational orientation and use of learning strategies. The items were translated from English to Arabic by the research team.

Procedures

Permission to collect data from the students during the orientation week was obtained from the University Administration. The research team informed the students that a study was being conducted on the motivational orientations and learning resources and strategies adopted by university students. Emphasis was placed on the fact that information to be gathered would not influence their admission in the university in any way and that the study would hopefully lead to improved learning in the university. Students who wished to participate in the study were given the MSLQ. Total time for completing the instrument averaged approximately twenty minutes.

Statistical Analyses

The first phase of the analysis was to test the first-order factor structures of the MSLQ. Two confirmatory factor analyses (CFA) with maximum likelihood estimation in EQS 6.1 were conducted on the MSLQ item data: one for the set of motivation items and another for the set of learning resources and strategies items. The analysis was conducted using the covariance matrix. Each item was constrained to load only on its hypothesized factor described by Pintrich et al. (1993). One item on each factor was constrained to equal one in order to set a metric for the factors. Factor covariances were left free to be estimated, but the measurement errors were not allowed to covary. During this phase, items that resulted in large standardized residuals were identified for possible deletion. Items were deleted one at a time in order to study changes in the factor structure solution. The purpose of this item deletion was to develop the best fitted and the most parsimonious model to the data, while maintaining the original factor structure.

The second phase of the analysis involved developing two second-order factor structures. As implied by Pintrich et al. (1993), one second-order factor model tested the existence of the motivation structure, whereas the other tested the existence of the learning resources and strategies structure. All covariances among the first-order factor residuals were constrained to equal zero. A matrix of the loadings of the first-order factors on the second-order factors was set up in which all subscales of the MSLQ were allowed to load freely on their hypothesized second-order factors. The variances of the second-order factors were constrained to equal 1.

In evaluating the fit of the models in all CFAs, recommendations by Schermelleh-Engel, Moosbrugger, and Muller (2003) were followed. These recommendations state that for an acceptable model fit, the ratio [X.sup.2] / df should be less than or equal to 3, the Root Mean Square Error of Approximation (RMSEA) should be less than or equal to .08, the Nonnormed Fit Index (NNFI) that is also called the Tuker-Lewis Index (TLI) should be greater than or equal to .95, and the Comparative Fit Index (CFI) should be greater than or equal to .95 (Schermelleh-Engel et al., 2003). The RMSEA, NNFI, and CFI were chosen because they were found to be less affected by the size of the sample when compared to the Normative Fit Index (NFI), the Goodness-of-Fit Index (GFI), and the Adjusted Goodness-of-Fit Index (AGFI) (Schermelleh-Engel et al., 2003).

The third phase of the analysis involved estimating the internal consistency reliability of the MSLQ's scores resulting from the CFAs obtained in the previous phases. Finally, percentile ranks were calculated as norms for interpreting the raw scores of the sample on MSLQ.

Results

First-Order Factor Structures

Motivation model. A CFA was first used to evaluate the dimensionality of the motivation model by constraining the 31 items to fall onto the six motivational latent factors: intrinsic goal orientation (4 items), extrinsic goal orientation (3 items), task value (6 items), control beliefs about learning (4 items), self-efficacy (8 items), and test anxiety (5 items). Results yielded an inferential test of = 1170.05 (p = .00, df = 419) with the following descriptive fit indices (RMSEA = .04 with 90% CI = [.04-.05], NNFI = .88, and CFI = .89). Although the fit indices compare favorably to those reported in Davenport (1999) and Pintrich et al. (1993) and all of the item parameters were statistically significant, the fit indicators pointed out that the 31-item motivation model was not fit at the expected level.

Thus, an attempt was made to modify the first-order 31-item motivation model to fit the data. Items with large standardized residuals were identified for deletion. Items were deleted one at a time until an acceptably fitting model was obtained. In the deletion of items, it was of interest to maintain the original six factor structure of motivation with a minimum of three items per factor to avoid identification problems. This process resulted in a 28-item model with an inferential test of = 953.22 (p = .00, df = 335) and with satisfactory descriptive fit indices (RMSEA = .04 with 90% CI = [.04-.05], NNFI = .90, and CFI = .91).

Table 1 presents standardized factor loadings for this first-order 28-item model of motivation along with factor intercor-relations. The first factor explained between 9 and 38%, the second factor accounted for between 29 and 42%, the third factor explained between 23 and 48%, the fourth factor accounted for between 9 and 40%, the fifth factor explained between 25 and 53%, and the sixth factor accounted for between 13 and 40% of the variance in the item data. All factor loadings were statistically significant at p < .01. As expected, the positive motivational factors (intrinsic goal orientation, task value, control beliefs of learning, and self-efficacy) are all correlated positively with one another with correlations ranging from .42 to .68. Test anxiety was negatively correlated with the positive motivational beliefs with correlations ranging from .09 to .23, but not significantly correlated with extrinsic goal orientation, which is a less desirable motivational characteristic (Pintrich et al., 1993).

Learning model. A CFA was first used to evaluate the dimensionality of the learning model by constraining the 50 items to fall onto the nine learning latent factors: students' use of rehearsal (4 items), elaboration (6 items), organization (4 items), critical thinking (5 items), self-regulation (12 items), time and study environment (8 items), effort regulation (4 items), peer learning (3 items), and help seeking (4 items). Results yielded an inferential test of = 3385.69 (p = .00, df = 1139) with the following descriptive fit indices (RMSEA = .05 with 90% CI = [.04-.05], NNFI = .82, and CFI = .84). Although the fit indices compare favorably to those reported in Davenport (1999) and Pintrich et al. (1993) and all of the item parameters were statistically significant, the fit indicators pointed out that the 50-item learning model was not fit at the expected level.

Thus, an attempt was made to modify the first-order 50-item learning model to fit the data. Items with large standardized residuals were identified for deletion. Items were deleted one at a time until an acceptably fitting model was obtained. In the deletion of items, it was of interest to maintain the original nine factor structure of learning with a minimum of three items per factor to avoid identification problems. This process resulted in a 43-item model with an inferential test of = 2152.13 (p = .00, df = 824) and with satisfactory descriptive fit indices (RMSEA = .04 with 90% CI = [.04-.05], NNFI = .90, and CFI = .90).

Table 2 presents standardized factor loadings for this first-order 43-item model of learning along with factor intercor-relations. The first factor explained between 25 and 44%, the second factor accounted for between 28 and 46%, the third factor explained between 34 and 45%, the fourth factor accounted for between 20 and 36%, the fifth factor explained between 28 and 40%, the sixth factor accounted for between 10 and 45%, the seventh factor explained between 9 and 46%, the eighth factor accounted for between 18 and 46%, and the ninth factor explained between 27 and 45% of the variance in the item data. All factor loadings were statistically significant at p < .01. As expected, all cognitive strategies, metacognitive strategies, and resource management subscales were positively related to one another with correlations ranging from. 17 to .77.

Second-Order Factor Structures

Motivation model. As shown in Table 1, the first-order factors of the motivation section of the MSLQ are correlated with one another. It may be hypothesized that the correlations among the factors can be accounted for by second-order factors. As such, based on the first-order factor model of motivational orientations, an alternative second-order factor model was developed to test the existence of the motivation structure implied by Davenport (1999) and Pintrich et al. (1993). This structure included two second-order factors, which were Expectancy and Value. Intrinsic goal orientation, extrinsic goal orientation, and task value were loaded on the second-order Value factor. Self-efficacy and control of learning beliefs were loaded on the second-order Expectancy factor. As stated by Davenport (1999), test anxiety was hypothesized to be a function of both expectations and value of success, and as such it was cross-loaded on both second-order factors. Results yielded an inferential test of = 983.94 (p = .00, df = 334) with the following descriptive fit indices (RMSEA = .05 with 90% CI = [.04-.05], NNFI = .89, and CFI = .90). It seems that the first-order factor model of motivation represents the sample data much better than the third-order factor model.

In an attempt to evaluate the fit of the second-order factor model in relation to the first-order model, Marsh and Hocevar (1985) designed a goodness-of-fit index called the target coefficient (T), which is the ratio of the chi-square value from the first-order model to that of the second-order model. A value of 1.0 indicates that the covariation among the first-order factors is completely accounted for by the second-order model. Thus, based on the chi-square values obtained for the first-and second-order factor models of motivation, the value of T was .97, suggesting that 97% of the covariation among the six first-order factors of the motivational orientations can be accounted for by the two second-order factors of Expectancy and Value.

Learning model. As shown in Table 2, the first-order factors of the learning strategies section of the MSLQ are correlated with one another. It may be hypothesized that the correlations among the factors can be accounted for by second-order factors. As such, based on the first-order factor model of learning strategies, an alternative second-order model was developed to test the existence of the learning strategies structure implied by Davenport (1999) and Pintrich et al. (1993). This structure included two second-order factors, which were Learning Strategies and Resource Use. Rehearsal, elaboration, organization, critical thinking, and self-regulation were loaded on the second-order Learning Strategies factor. Time and study environment, effort regulation, peer learning, and help seeking were loaded on the second-order Resource Use factor. Results yielded an inferential test of = 2410.14 (p = .00, df = 839) with the following descriptive fit indices (RMSEA = .04 with 90% CI = [.04-.05], NNFI = .87, and CFI = .88). It seems that the first-order factor model of learning represents the sample data much better than the second-order factor model. Using Marsh and Hocevar's (1985) goodness-of-fit index, 89% of the covariation among the nine first-order factors of the learning strategies can be accounted for by the two second-order factors of Learning Strategies and Resource Use.

Internal Consistency Reliability

Based upon the first-order factor analysis, it appears that the reduced 71-item MSLQ models provide a reasonable fit to the data. As such, for comparative purposes, Table 3 presents internal consistency reliabilities as measured by Cronbach's alpha for the sample's scores on the short and long versions of the MSLQ. As shown in Table 3, the reliabilities of the short version's scores seem comparable to those of the long version of the MSLQ. Although the magnitude of the score reliability is relatively lower than that reported by Pintrich et al. (1993), no precision was lost in the short version obtained in this study compared to the long version. This suggests that the motivational orientations and learning resources and strategies of students entering SQU can be effectively represented by a smaller number of items while keeping the same degree of measurement precision as the long version 81-item MSLQ.

Norms

Percentile ranks were extracted as norms for the raw scores of the participants to indicate the relative position of an individual student in the sample. Table 4 shows means and standard deviations along with the percentile ranks for the scores on the MSLQ. As seen in Table 4, the means of the participants on the MSLQ subscales were above the scales' midpoints, suggesting that, on average, the participants tended to adopt desirable motivational orientations and learning resources and strategies. The norms of the present study's sample were slightly higher than the norms reported in Pintrich, Smith, Garcia, and McKeachie's (1991) study.

Conclusion

Students' positive motivational orientations and desirable use of learning resources and strategies have been recognized as the most influential factors of the academic success of the university students (Pintrich et al., 1993). To adequately assess these motivational and learning factors, Pintrich et al. (1993) developed an 81-item instrument entitled the Motivated Strategies for Learning Questionnaire (MSLQ). There has been an increased interest in utilizing the MSLQ to understand students' motivational characteristics and use of learning resources and strategies around the world (Duncan & Mckeachie, 2005). However, the measurement properties of this instrument have not been explored statistically when applied on university students in Oman. The present study reports on the validity and reliability of the MSLQ's scores for SQU students in Oman. The importance of this study stemmed from Messick's (1995) assertion that establishing evidence of score meaning is a continual and evolving process. Additionally, the significance of the study comes from the concern over the inadequate level of motivation and usage of learning strategies possessed by students entering SQU (Al-Barwani et al., 1997; Yahya, 1997).

The current findings indicated that a reduced version of MSLQ consisting of 71 items displayed conceptually sound first-order factor structures of motivational orientations and use of learning resources and strategies as well as reasonable levels of internal consistency. From a practical point of view, the development of a shortened MSLQ would be of benefit to researchers who are conducting studies in which MSLQ's constructs are one of several constructs being measured. Overall, the results indicate that MSLQ provides reliable and valid interpretations of the motivational orientations and usage of learning resources and strategies of students entering SQU. This makes MSLQ a valuable assessment tool for university educators to help academically unprepared students entering the university. Specifically, the MSLQ can be administered during the orientation week to identify academically unprepared students who got low scores on the MSLQ. Appropriate developmental instructional programs can then be designed to enhance their motivational orientations and equip them with the learning skills necessary for a successful beginning to their university education. Early identification of academically at-risk students may improve students' academic success and retention rates (Adelman, 1986; Jacobi, Astin, & Ayala, 1987).

Finally, the generalizability of the study findings was limited by the non-random sample used in the study. Also, a larger sample size will enable future researchers to test whether the factorial structure of the MSLQ is invariant across males and females. Future research might consider examining the changes in students' motivational orientations and usage of learning resources and strategies across time during their university education.

References

Adelman, C. (1986). Assessment in American higher education. Washington, DC: Department of Education.

Al-Barwani, T. H., Yahya, A., & Ibrahim, A. (1997). Basic skills and attitudes for university study as perceived by faculty members at Sultan Qaboos University. Series of Educational and Psychological Studies, 2, 39-63.

Al-Shahwarzi, J. F., AlGhufaili, H., & Al-Jabri A.M. (1991). Factors that influence the success of students in Sultan Qaboos University. College of Science, Sultan Qaboos University.

Andreou, E., & Metallidou, P. (2004). The relationship of academic and social cognition to behaviour in bullying situations among Greek primary school children. Educational Psychology, 24, 27-41.

Artino, A. R., Jr. (2005). Review of the Motivated Strategies for Learning Questionnaire. Manuscript submitted for publication.

Davenport, M. A. (1999). Modeling motivation and learning strategy use in the classroom: An assessment of the factorial, structural, and predictive validity of the Motivated Strategies for Learning Questionnaire. Unpublished doctoral dissertation, Auburn University, Alabama.

Duncan, T. G., & McKeachie, W. J. (2005). The making of the Motivated Strategies for Learning Questionnaire. Educational Psychologist, 40, 117-128.

Garcia, T., & Pintrich, P. R. (1993). Self-schemas, motivational strategies and self-regulated learning. Paper presented at the annual meeting of the American Educational Research, Atlanta (ERIC Document No.ED359234).

Howey, S. C. (1999). The relationship between motivation and academic success at community college freshmen orientation students. Unpublished doctoral dissertation, Kansas State University, Kansas.

Ibrahim, A., Yahya, A., & Al-Barwani, T. (1992). A Study of the Omani secondary school certificate examination as a predictor of academic performance at Sultan Qaboos University. Series of educational and Psychological Studies, 2.

Jacobi, M., Astin, A., & Ayala, E, Jr. (1987). College student outcomes assessment: A talent development perspective. ASHE-ERIC Higher Education Report No. 7. Association for the Study of Higher Education: Washington, D.C.

Marsh, H. W., & Hocevar, D. (1985). Application of confirmatory factor analysis to the study of self-concept: First and higher-order factor models and their invariance across groups. Psychological Bulletin, 97, 562-582.

Messick, S. (1995). Validity of psychological assessment: Validation of inferences from persons' responses and performances as scientific inquiry into score meaning. American Psychologist, 50, 741-749.

Ministry of Education. (1994). The educational system in Oman: Achievements and challenges. Sultanate of Oman: Author.

Ministry of Education. (1995). Reform and development of general education in the Sultanate of Oman. Sultanate of Oman: Author.

Ministry of Education. (2007). Basic education: Aims, implementation and evaluation. Sultanate of Oman: Author.

Neber, H., & Heller, K. A. (2002). Evaluation of a summer-school program for highly gifted secondary-school students: The German Pupils Academy. European Journal of Psychological Assessment, 18, 214-228.

Pintrich, P. R. (2000). The role of goal orientation in self-regulated learning. In M. Boekaerts & P. R. Pintrich(Eds.), Handbook of self-regulation (pp.13-39). San Diego, CA: Academic Press.

Pintrich, P. R., & DeGroot, E. V. (1990). Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology, 82, 33-40.

Pintrich, P. R., Smith, D. A. E, Garcia, T., & McKeachie, W. J. (1991). A manual for the use of the Motivated Strategies for Learning Questionnaire (MSLQ). National Center for Research to Improve Postsecondary Teaching and Learning, Ann Arbor, MI. (ERIC Document Reproduction Service No. ED338122)

Pintrich, P. R., Smith, D. A. E, Garcia, T., & McKeachie, W. J. (1993). Reliability and predictive validity of the Motivated Strategies for Learning Questionnaire (MSLQ). Educational and Psychological Measurement, 53, 801-814.

Rao, N., Moely, B. E., & Sachs, J. (2000). Motivational beliefs, study strategies, and mathematics attainment in high- and low-achieving Chinese secondary school students. Contemporary. Educational Psychology, 25, 287-316.

Schermelleh-Engel, K., Moosbrugger, H., & Muller, H. (2003). Evaluating the fit of structural equation models: Tests of significance and descriptive goodness-of-fit measures. Methods of Psychological Research Online, 8, 23-74.

Watson, M., McSorley, M., Foxcroft, C. & Watson, A. (2004). Exploring the Motivation Orientation and Learning Strategies of First Year University Learners. Tertiary Education and Management, 10, 193-207.

Yahya, A. (1997). Achievement motivation as a function of sex, age, and academic achievement of Sultan Qaboos University Students. Series of educational and Psychological Studies, 2, 1-15.

HUSSAIN ALKHARUSI

OTHERINE NEISLER

THUWAYBA AL-BARWANI

DAVID CLAYTON

HUMAIRA AL-SULAIMANI

MOHAMMAD KHAN

Sultan Qaboos University, Oman

HAMAD AL-YAHMADI

Diwan Of Royal Court, Oman

MUNA AL-KALBANI

Ministry of Education, Oman

Table 1 Standardized Factor Loadings for the First-Order 28-Item Model of the Motivation Section of the MSLQ alonj with the Factor Intercorrelations Item Factor loadings FI F2 F3 F4 F5 F6 1 0.35 2 0.62 3 0.30 4 0.54 5 0.60 6 0.65 7 0.48 8 0.69 9 0.60 10 0.63 11 0.53 12 0.58 13 0.41 14 0.63 15 0.30 16 0.60 17 0.52 18 0.72 19 0.50 20 0.73 21 0.58 22 0.66 23 0.53 24 0.36 25 0.45 26 0.47 27 0.63 28 0.48 Factor intercorrelations F1 -- F2 .34 *** -- F3 .54 *** .55 *** -- F4 .42 *** .45 *** .57 *** -- F5 .51 *** .54 ** .68 *** .52 *** -- F6 -.13 *** -0.03 -.16 *** -.09 ** -.23 *** -- Note. F 1 = intrinsic goal orientation. F2 = extrinsic goal orientation. F3 = task value F4 = control beliefs about learning. F5 = self-efficacy. F6 = test anxiety. ** p<.01. *** p<.001. Table 2 Standardized Factor Loadings for the First-Order 43-Item Model of the Learning Section of the MSLf alon'S with the Factor Intercorrelations Item Factor loadings F1 F2 F3 F4 F5 F6 F7 1 0.50 2 0.61 3 0.66 4 0.62 5 0.53 6 0.60 7 0.65 8 0.62 9 0.68 10 0.52 11 0.59 12 0.58 13 0.61 14 0.67 15 0.45 16 0.52 17 0.60 18 0.50 19 0.50 20 0.55 21 0.58 22 0.53 23 0.54 24 0.61 25 0.54 26 0.56 27 0.63 28 0.58 29 0.47 30 0.63 31 0.49 32 0.32 33 0.67 34 0.64 35 0.56 36 0.30 37 0.68 38 39 40 41 42 43 Factor intercorrelations F1 -- F2 .66 ** -- F3 .62 ** .68 ** -- F4 .56 ** .65 ** .55 ** -- F5 .70 ** .77 ** .68 ** .65 ** -- F6 .62 ** .62 ** .57 ** .49 ** .71 ** -- F7 .47 ** .49 ** .44 ** .41 ** .58 ** .56 ** -- F8 .37 ** .37 ** .39 ** .40 ** .38 ** .24 ** .17 ** F9 .52 ** .55 ** .45 ** .40 ** .56 ** .55 ** .43 ** Item Factor loadings F8 F9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 0.46 39 0.43 40 0.68 41 0.52 42 0.56 43 0.67 Factor intercorrelations F1 F2 F3 F4 F5 F6 F7 F8 -- F9 .37 ** -- Note. F 1 = rehearsal. F2 = elaboration. F3 = organization. F4 = critical thinking. F5 = self-regulation. F6 = time and study environment. F7 = effort regulation. F8 = peer learning. F9 = help seeking. ** P <.01. Table 3 Internal Consistency Reliabilities by Cronbach's Alpha of the Short (71-item) and Long (81-item) Versions of the MSLQ Subscale Present sample Pintrich et Short version Long version al. (1993) Intrinsic goal 0.40 0.41 0.74 orientation Extrinsic goal 0.62 0.49 0.62 orientation Task value 0.76 0.74 0.90 Control of learning 0.40 0.48 0.68 beliefs Self-efficacy 0.82 0.82 0.93 Test anxiety 0.59 0.56 0.80 Rehearsal 0.69 0.67 0.69 Elaboration 0.77 0.75 0.76 Organization 0.71 0.69 0.64 Critical thinking 0.65 0.64 0.80 Self-regulation 0.82 0.78 0.79 Time and study 0.67 0.52 0.76 environment Effort regulation 0.54 0.32 0.69 Peer learning 0.52 0.49 0.76 Help seeking 0.60 0.39 0.52 Table 4 Means, Standard Deviations, and Percentile Ranks for the Sample Scores on the MSLQ Sabscales Subscale M SD Percentile ranks 25 50 75 Intrinsic goal orientation 5.27 1.06 4.75 5.50 6.00 Extrinsic goal orientation 5.82 1.12 5.25 6.00 6.75 Task value 5.75 1.04 5.33 6.00 6.50 Control of learning beliefs 5.53 1.04 5.00 5.75 6.25 Self-efficacy 5.74 0.96 5.25 5.88 6.38 Test anxiety 3.59 1.33 2.60 3.60 4.60 Rehearsal 5.45 1.18 4.75 5.75 6.25 Elaboration 5.33 1.10 4.67 5.50 6.17 Organization 5.39 1.24 4.50 5.50 5.80 Critical thinking 5.02 1.05 4.40 5.00 5.80 Self-regulation 5.28 0.93 4.75 5.33 6.00 Time and study environment 4.98 0.83 4.50 5.00 5.63 Effort regulation 4.96 1.06 4.25 5.00 5.75 Peer learning 4.44 1.33 3.42 4.67 5.33 Help seeking 4.99 1.06 4.5 5.25 5.75

Gale Copyright:

Copyright 2012 Gale, Cengage Learning. All rights
reserved.