| The purposes of this study were (a) to determine the internal factor structure of the six latent variables investigated, including home environment, peer influences, school environment, educational aspirations, attitudes toward mathematics and study habits, (b) to examine the effects of the above six variables on mathematics achievement as measured by TIMSS (Third International Mathematics and Science Study), (c) to investigate whether the same internal factor structure and pattern of influences existed for students from three selected countries. A learning model was developed on the basis of Walberg's educational productivity model and was tested by the LISREL multi-sample approach. The TIMSS mathematics achievement test and the student background questionnaire for eighth grade students were used to achieve the above purposes. This study involved a total of 14,651 eighth grade students, including 2,920 in the Republic of South Korea, 4,644 in Singapore, and 7,087 in the United States.This study found different factor structures and different influences on mathematics achievement across the three selected countries. Nevertheless, the factor loading estimates of observed indicators on their respective latent variables ranged from moderate to high. These results supported the effectiveness of the twenty observed indicators in measuring the six latent variables investigated. In addition, Home environment, attitudes toward mathematics, and educational aspiration emerged as the more important and consistent predictors of mathematics achievement for the three countries. The other three variables had mixed or inconsistent effects on mathematics achievement. For Korean students, the effects of the six manipulable variables on mathematics achievement were most discrepant with the proposed model, whereas the effects for Singaporean students were least discrepant. Nevertheless, the multiple squared correlations for the structural model of mathematics achievement were very similar across countries. Furthermore, the Singaporean learning model supported the value of Walberg's educational productivity model. |
