Abstract
"This study was a nonexperimental design that incorporated multiple correlation
and regression: (a) to investigate the relationship of cognitive development and disembedding ability to student performance in a college-level astronomy course, and (b) to assess which variable is the best predictor of performance. Cross-validation was used to validate the findings.
The sample consisted of 184 students (105 females, 79 males) enrolled in General Astronomy at a state university located in Southeastern Pennsylvania during the 1993 Spring semester. An Inventory of Piaget’s Developmental Tasks (IPDT) was used to access cognitive development. The Group Embedded Figures Test (GEFT) was used to assess disembedding ability. The mean score for the IPDT was 59.41, standard deviation = 7.45. The mean score for the GEFT for males was 12.65, standard deviation = 4.61; the mean score for females was 11.45, standard deviation = 4.43. The mean score for the final grade in astronomy was 2.98, standard deviation = .88.
The correlation between IPDT scores and final grades in astronomy was .23
(p < .01) indicating a low (5.3% of the variance), but significant, correlation between cognitive development and student performance. The correlation between GEFT scores and final grades in astronomy was .17 (p < .05) indicating a low (2.9% of the variance), but significant, correlation between disembedding ability and student performance.
The multiple regression analysis for the sample (N = 184) yielded a multiple R value of .23 (p < .01) between the final grade in astronomy and the two predictor variables. Cross-validation produced multiple R values of similar magnitude which suggested that the combination of cognitive development and disembedding ability provide a composite measure that accounts for a statistically significant amount of the variation in grade performance of students enrolled in a college-level astronomy course.
Stepwise multiple regression analysis for the sample (N = 184) yielded a multiple R value of .23 (p < .01) and the following regression equation: Y = .03 X] + 1.38. Cross-validation produced multiple R values of similar magnitude and the following regression equation: Y = .03 x, + 1.28. In both equations, x, represented the IPDT scores. This result indicates that cognitive development best predicts student performance in a college-level astronomy course.
Discussion, conclusions, and further recommendations for research are presented."
