Computational analysis of individual differences in metaphor comprehension and cognitive abilities

In this study, we investigated how individual differences in cognitive abilities (i.e., fluid intelligence, broad retrieval ability, and similarity reasoning ability) relate to the human metaphor comprehension process using the prediction model with two processes (spreading activation of the vehicle, m, and constraint in the topic, k) to compute metaphor interpretations. The participants (N = 85) completed cognitive ability tasks (Japanese version of the Semantic Similarities Test, Raven Progressive Matrices Short, and Ideation fluency task) and a metaphorical interpretation task, in which they listed multiple features of the presented metaphors. Each participant’s interpretation of the metaphors was modeled using the predication model, and the parameters (k and m) were determined to maximize the evaluation metrics (Mean Average Precision and Mean Reciprocal Rank). We found that the similarity reasoning ability affected the spreading activation of the vehicle concept and the degree of constraints on the topic. Compared to previous studies that found that the outcome of metaphor comprehension relates to similarity reasoning ability, our results showed that similarity reasoning ability is also related to the processes of metaphor comprehension.