Understanding math and reading achievement through motivational and contextual variables in the Early Childhood Longitudinal Survey (ECLS-K)

The primary focus of this study was to isolate the motivational and contextual variables (i.e., instructional practices, classroom relationships, student motivation, socioeconomic status, parental expectations and support, and their respective sub-variables) that contributed the most variance to math and reading achievement scores. A sample of (n = 17,212) grade one students from the Early Childhood Longitudinal Survey (ECLS-K) was the source of the data used in this study. Multiple regression and canonical analyses were employed to analyze spring 2000 data of the ECLS-K.;Multiple regression (R square = .35) revealed that cognitive skills and strategy use (17.8%), parents' beliefs in their child's performance in math (7.2%), child tutored regularly (3.2%), highest degree parents expect of their child (2.0%), and student anxiety (2.0%) contributed the most variance to math achievement scores. Similarly, cognitive skills and strategy use in reading (26.1%), parents' beliefs in their child's performance in math (12.2%), child tutored regularly (4.1%), and the highest degree parents expect of their child (2.5%) contributed the most variance to reading achievement scores (R square = .57).;Canonical analyses created two pairs of canonical variates that explained 47% of the variance in math and reading achievement scores. The first canonical variable of the motivational and contextual variables (W1) explained 37.5% of the variance in math and reading achievement scores. W1 was comprised of cognitive skills and strategy use in reading, socioeconomic status, students' attitudes toward learning, and child tutored regularly. The second canonical variable of the motivational and contextual variables (W2) only explained 1.5% of the variance in math and reading achievement scores. Square multiple correlations revealed that the canonical variate of the motivational variables had significant predictive power for math and reading achievement. However, the canonical variate of math and reading achievement scores had little or no predictive power for the motivational and contextual variables.;Overall, the findings indicated that cognitive skills and strategy use in math and reading, socioeconomic status, parental beliefs, and regular tutoring were the major contributors to the variance in math and reading achievement scores. Implications for parents and educators are discussed.