Predictors of grade seven mathematics achievement

The study investigated the predictive relationship between Grade Seven mathematics achievement and five intellectual variables: upper‚ÄövÑvÆprimary school mathematics achievement, mathematics aptitude, non‚ÄövÑvÆverbal I.Q., verbal I.Q. and reading comprehension. Subjects for the study were Grade Seven students from a secondary school. Data was provided by student measures on five test instruments which are widely used in primary and secondary schools as measures of academic ability. Basic descriptive statistics, zero‚ÄövÑvÆorder correlations and multiple linear regression techniques allowed for the exploration of several broad questions and issues that emerged from the review of related literature. The study investigated the following issues concerned with prediction of Grade Seven mathematics achievement: 1. that measures of prior mathematics performance are more efficient predictors than global performance measures; 2. that mathematics aptitude is a more efficient predictor than prior mathematics achievement; 3. that non‚ÄövÑvÆverbal I.Q. is a more efficient predictor than verbal I.Q.; 4. that where verbal I.Q. has already been included as a predictor then reading comprehension provides no additional predictive information. Answers to these issues were generally consistent with previous research and were in agreement with theory. The following interpretation was made. The two measures of prior mathematics performance were significantly more efficient predictors than the two measures of global performance. The measures of prior mathematics performance were not significantly different in predictive efficiency, nor were the global performance measures, but mathematics aptitude achieved the highest correlation, accounting for 56% of variance. Verbal I.Q. was a more efficient predictor than reading comprehension. The most efficient prediction was obtained by utilising both measures of prior mathematics performance and both measures of global performance. The four significant predictors together accounted for 66.067% of variance. Using one‚ÄövÑvÆyear cross‚ÄövÑvÆvalidation, a prediction equation was determined. The economic equation used both global performance measures and mathematics aptitude as predictors and achieved a cross‚ÄövÑvÆvalidated correlation coefficient of 0.78.