Effects of generating versus evaluating distant analogies on subsequent relational integration: mediating and moderating mechanisms.

BACKGROUND: Relational integration is a core component of higher-order cognition; however, little is known about whether it can be temporarily influenced by prior analogy processing or whether such effects vary across individuals. In the present study, we examined whether generating, rather than evaluating, distant analogies was associated with better subsequent relational integration; whether this association was mediated by cognitive load; and whether it was moderated by individual differences in cognitive abilities and cognitive styles. METHODS: One hundred twenty Chinese college students were randomly assigned to either a distant analogy-generation condition or a distant analogy-evaluation condition. After the priming task, participants completed a multidimensional relational reasoning task (MRRT), a cognitive load measure, and several measures of cognitive ability and cognitive style. RESULTS: Participants in the generation condition performed better on the MRRT than those in the evaluation condition. This difference was partially mediated by lower cognitive load during the subsequent relational integration task. Working memory capacity substantially moderated the direct association between priming task type and relational integration, whereas fluid intelligence and cognitive style did not moderate this effect to a significant degree. CONCLUSION: Overall, the findings suggest that generating distant analogies may temporarily support subsequent relational integration, especially among individuals with relatively lower working memory.