Cognitive style and academic performance of Advanced Biology students

Abstract

Literature about Individual Differences in Learning proposes various constructs possibly determining performance in different domains of knowledge. Besides intellectual abilities and motivation, cognitive style has been proposed as one important factor contributing to scholastic performance. This study investigates the relationship between cognitive style and academic performance in Maltese students taking Biology at advanced level. A sample of 581, 212 males (36.5%) and 369 females (63.5%), at the Gian Frangisk Abela Junior College, University of Malta, were administered the Cognitive Style Analysis (CSA), and thus grouped according to their position on two fundamental cognitive styles dimensions (Wholist-Analytic and Verbal-Imagery). Their academic performance in five different subjects, two taken at A-level (Biology being one of them) and three taken at Intermediate level, was measured through a computed mark. This mark consisted of five semester assessments, given over the two-year course, added to the mark obtained at the end-of-year test in the first year. Pearson’s Correlation showed that the Wholist-Analytic dimension, the Verbal-Imagery dimension and gender were not correlated. Regression Analysis using the computed subject mark as dependent variable and Gender; WA and VI as the Independent variable showed no correlation between cognitive style and performance in any of the subjects considered. A statistically significant correlation was recorded for gender. Though no gender effect was shown on student performance in Biology, a significant influence of gender on performance in Chemistry, Physics, languages, and social studies was obtained, with standardized beta coefficients showing that females on average perform 0.13% better than males. Besides the effect of separate style dimensions on performance, the sample was split into nine categories, according to the different style dimension combinations given by the CSA. Since some of the resulting category frequencies were smaller than 30, a Chisquare test was used to compare means. This showed no significant differences for the various categories on the basis of gender. A style combination matrix was created. Regression showed that none of the matrix variables (style dimension combinations) proved to have significant effect on performance in any of the subjects. But gender proved to be a stronger determinant in performance. These results were interpreted in the light of previous research and from a Cognitive Neuroscience perspective. Implications for instruction and assessment were discussed.