The number of topic-attributed features affects speakers' metaphor production.

This study examined whether or not the number of topic-attributed features affects the speakers' use of metaphor production rather than literal expressions. Across two experiments, participants were asked to produce an expression that best paraphrased a given sentence. The number of features attributed to each topic was manipulated: one feature ("Her sarcasm hurts people"), two features ("Her sarcasm hurts people and is sharp"), and three features ("Her sarcasm hurts people, is sharp, and is piercing to the heart"). Participants' responses were classified into nominal metaphor/simile, literal, other metaphor/simile, and others. In both Experiment 1 and Experiment 2, participants' nominal metaphor responses (e.g., "Her sarcasm is a knife") increased with the number of topic-vehicles that shared significant features in a given sentence. These results suggest that the number of topic-attributed features affects participants' preference for the use of metaphorical expressions. We discussed the results based on the compactness hypothesis (Ortony, Educational Theory, 25: 45-53, 1975) of metaphor production.

[Development of free will and determinism scale in Japanese].

We developed a free will and determinism scale in Japanese (FAD-J) to assess lay beliefs in free will, scientific determinism, fatalistic determinism, and unpredictability. In Study 1, we translated a free will and determinism scale (FAD-Plus) into Japanese and verified its reliability and validity. In Study 2, we examined the relationship between the FAD-J and eight other scales. Results suggested that lay beliefs in free will and determinism were related to self-regulation, critical thinking, other-oriented empathy, self-esteem, and regret and maximization in decision makings. We discuss the usefulness of the FAD-J for studying the psychological functions of lay beliefs in free will and determinism.

Performance evaluation of automated scoring for the descriptive similarity response task.

We examined whether a machine-learning-based automated scoring system can mimic the human similarity task performance. We trained a bidirectional encoder representations from transformer-model based on the semantic similarity test (SST), which presented participants with a word pair and asked them to write about how the two concepts were similar. In Experiment 1, based on the fivefold cross validation, we showed the model trained on the combination of the responses (N = 1600) and classification criteria (which is the rubric of the SST; N = 616) scored the correct labels with 83% accuracy. In Experiment 2, using the test data obtained from different participants in different timing from Experiment 1, we showed the models trained on the responses alone and the combination of responses and classification criteria scored the correct labels in 80% accuracy. In addition, human-model scoring showed inter-rater reliability of 0.63, which was almost the same as that of human-human scoring (0.67 to 0.72). These results suggest that the machine learning model can reach human-level performance in scoring the Japanese version of the SST.