Abstract: Question asking has been a critical tool for teaching and learning since the time of Socrates and is important in the creative problem-solving process. Yet, its role in creativity has insofar not been thoroughly explored. We empirically investigate the role of question asking in the creative process with the alternative questions task (AQT). In the AQT, participants are asked to generate creative and unusual questions for common objects. Responses are rated for their question level using the Bloom’s taxonomy, a widely accepted guideline in designing examination questions of differing levels of complexity, as well as their subjective and objective creativity. First, I will present a series of studies where we find a significant positive relation between AQT question level and objective and subjective creativity scores: Higher, more complex questions were more creative, with the inverse effect true for lower-level questions. We interpret these findings as supporting the hypothesis that higher question complexity is related and predictive of creative ability. Next, I will present a large language model we trained to automatically rate the Bloom level (complexity) of questions asked in the AQT, showing strong correspondence with human ratings (r = .76) paving the way for automatic scoring of question complexity. Finally, I will present an online game we developed, Spot the Spy – where we study question asking in natural settings and relate it to cognitive capacities such as creativity, intelligence, and curiosity. Overall, these lines of work highlight the role of question asking, and especially question complexity, in creativity.

Bio: Dr. Yoed Kenett is an Assistant Professor at the Faculty of Data and Decision Sciences at the Technion – Israel Institute of Technology in Haifa, Israel. His research computationally and empirically investigates the complexity of high-level cognition in typical and atypical populations, focusing on creativity, associative thought, knowledge, and memory search. To investigate these issues, he applies computational tools from network science, natural language processing, and machine learning, coupled with empirical cognitive and neural research.