Spatial Abilities and Chemistry Achievement: Contributions and Limitations of Correlational Studies

• In Event: Spatial Thinking in Chemistry

Abstract

An obvious first step in studying spatial thinking in chemistry is to examine correlations between spatial ability measures and performance in chemistry. Given the importance of spatial concepts and representations in chemistry, it is not surprising that there have been several demonstrations that spatial ability significantly predicts chemistry performance (e.g., Bodner & McMillan, 1986; Carter, LaRussa & Bodner, 1987; Keig & Rubba, 1993; Pribyl & Bodner, 1987). However, correlations between spatial ability measures and chemistry achievement are not always high. I will present results of a new study in which we tested the spatial abilities of a group of 150 students in an introductory college chemistry class and then followed them through their first 4 quarters in college, examining how spatial abilities predicted their grades in chemistry classes. Only two of the four spatial ability tests, paper folding (r = .23) and a cross section test (r = .27) were significantly correlated with performance in their first chemistry class. By the second semester, only the cross section test (r = .18) was significantly correlated. By the fourth semester, in which they began studying organic chemistry, none of the spatial ability tests were predictive of performance. Although organic chemistry is often thought to be demanding of imagistic processes such as mental rotation and perspective taking, tests of these abilities were not significantly predictive of performance in any of the chemistry classes.

While correlational studies are a first step in providing evidence for the importance of spatial skills in chemistry, it is important to interpret them critically. First, many of the studies conducted to date are based on small samples. Second, most of the observed correlations of spatial ability with chemistry achievement measures, while statistically significant, are relatively small (.3 or lower). Third, some studies do not control for the possibility that the correlation between spatial ability and science performance reflects common variance shared with general intelligence. Fourth, the tasks that are most highly correlated with spatial ability measures are not necessarily those that have the highest face validity as spatial tasks. For example, in one study, the strongest correlations were with stoichiometry problems, which are mathematical rather than spatial in nature (Bodne & McMillan, 1986).

A final limitation of correlational studies is that they do not inform us about the specific cognitive processes that are used to solve chemistry problems. As a result, they do not provide good insights into how to alleviate student’s difficulties. Rather than focusing on correlations with existing measures of spatial ability, a better approach is to examine performance in specific spatial tasks in chemistry to identify the cognitive processes.

Author

• Mary Hegarty, University of California - Santa Barbara