Examining component processes of performance on the Tower of London

Abstract

Three experiments examined the component processes underlying performance on a widely used executive function task, the Tower of London (TOL). All of the experiments examined the relationship of TOL performance to measures of working memory and inhibition. Experiment I examined the contributions of working memory, inhibition, and fluid intelligence to performance on the TOL. It was found that although working memory and inhibition did not significantly contribute to TOL performance, fluid intelligence did account for a significant amount of variance. In contrast, Experiment 2 found that, overall, the working memory and inhibition variables accounted for 34% of the variance in TOL performance. Experiment 2 also showed that the underlying problem structure affected accuracy on the task, but the hypothesis that there would be a stronger relationship between external measures of inhibition and problems with conflict moves and suboptimal solution paths was not supported. Experiment 3 found that as the secondary task became more difficult, performance was less accurate and slower in both tasks. These results support a working memory maintenance component to Tower of London performance. There was not an interaction between move number and the number of letters in Tower of London accuracy indicating that complexity and working memory maintenance may not have a strong relationship. Based on these findings, it is concluded that the commonly held view that executive function as measured by the Tower of London can be conceptualized in terms of working memory and inhibition needs to be reexamined. The finding that problem structure affects complexity in the Tower of London problems is important both in theoretical and practical terms.