Multiple Analyses as an Alternative Method to Avoid Ecological Problem, Using Students and Teacher Data of Timss 2011

Ecological problem in the analysis of aggregate data is a serious methodological issue in most studies. In the present paper, the problems of Aggregate analysis and disaggregate analysis has been analyzed, and multilevel modeling procedures, specifically Hierarchical Linear Modeling (HLM) as an alternative method is introduced. This research is correlational and has analyzed the relationships of variables among 6029 eighth-grade students (2816 girls and 3213 boys) who had participated in TIMSS 2011 and completed the TIMSS questionnaire. The results showed that there is a significant relationship between Homework time and Homework opportunity with mathematics achievement at students-level; however, it was not significant at teachers-level. The utility of multiple analyses at nested data set were also confirmed.