Cluster Analysis Findings Over 20 Years of TIMSS

Abstract

Student motivation clusters were identified by analyzing IEA’s Trends in International Mathematics and Science Study (TIMSS) mathematics data from 1995, 2007, and 2015 at both grade four and grade eight in 12 jurisdictions. Deeper examination of the distributions of student motivation variables within each cluster revealed that high confidence scores, followed by high enjoyment scores, were generally associated with better student achievement in mathematics. Students’ value for the subject was dissociated from the other two motivation variables in at least one cluster in nearly all samples. Statistical tests indicated that mean achievement, gender composition, and mean home resources scores differed across clusters in systematic ways at both grades and across all three TIMSS administrations. Twenty years of TIMSS data reveal interesting patterns in the number of clusters, the relationships among the three motivational variables and their association with achievement, and concerning the gender composition of distinct motivational clusters. Clusters with similar mixes of motivational variables seem to produce consistent and almost universal effects across jurisdictions.