Synthetic Artificial Intelligence in Verbal Creative Thinking Assessment: A Case Study on the Torrance Verbal Creativity Test

This study investigates the use of synthetic artificial intelligence (AI) in assessing verbal creativity through the Torrance Verbal Creativity Test (TVCT). The research was conducted with 300 participants who completed the TVCT. Their responses were manually scored by two experts, assessing fluency, flexibility, originality, and overall creativity. The consistency between the manual scores was examined. Additionally, the study involved training two AI models, ChatGPT-4 and DeepSeek, using identical guidelines to score the same responses. The reliability of AI assessments was compared with manual scoring, highlighting the effectiveness, limitations, and alignment of AI in creative thinking assessments.