A causal model of variables predicting mathematical achievement in first-grade children

Michele Kasson Loebl, Fordham University


A causal model of mathematical achievement of first-grade children was evaluated. This model consisted of sex of the student as the exogenous variable. Verbal ability; spatial ability; attitude toward school, learning, teacher, and mathematics; and mathematical problem solving were endogenous variables.^ Sex of the student was hypothesized to directly affect verbal ability, spatial ability, and attitude but not mathematical problem-solving ability or mathematical achievement. Indirect effects of sex on mathematical ability were hypothesized through the endogenous variables. The endogenous variables were hypothesized to directly affect mathematical achievement, and to produce indirect effects on mathematical achievement through their causal relationships with each other.^ The sample included 73 boys and 83 girls from first-grade classes in New York City. Verbal ability and mathematical achievement were assessed by the Comprehensive Test of Basic Skills, vocabulary, reading comprehension, mathematical concepts, and computation subtests, level C. Spatial ability was assessed through the Arlin-Hills Attitude Surveys. Mathematical problem solving was assessed by addition and subtraction tasks, and linear measurement tasks which have previously been used in research with young children.^ An analysis of the causal model was provided with the use of LISREL. A measure of goodness of fit of the model was provided with the chi-square statistic. Results indicated that variables predicting mathematical achievement affect mathematical achievement directly, and indirectly through endogenous variables. While the paths between sex and verbal ability and sex and attitude were not significant, these paths approach significance. Direct and indirect effects for verbal ability and spatial ability were found on mathematical achievement. Similarly, mathematical problem solving also had a significant effect on mathematical achievement. The attitude variable did not show a significant effect on mathematical achievement.^ Concurrent validity was established for verbal, spatial, and problem-solving tasks, but not for the attitude survey. Based upon the results, three alternative, but more restricted, models were assessed.^ A better fit of the model was provided when the attitude variables were constrained, but not evidenced when an alternative spatial construct was hypothesized. ^

Subject Area

Educational psychology|Developmental psychology|Psychology

Recommended Citation

Loebl, Michele Kasson, "A causal model of variables predicting mathematical achievement in first-grade children" (1990). ETD Collection for Fordham University. AAI9304530.