Better answers, broader thinking: What students gain from ChatGPT and critical-thinking training
A Bocconi University study with over 1,000 students found ChatGPT improved work quality while causal reasoning training boosted originality, with combined benefits when used together.
A randomized experiment involving 1,000 first-year Bocconi University students examined the effects of ChatGPT access and causal reasoning training on academic work. Students were divided into four groups: those with ChatGPT access, those given critical-thinking training, those with both, and a control group. The study aimed to assess how each intervention influenced the quality and originality of student submissions on a real-world business case.
Students with ChatGPT access produced work that scored nearly a full point higher on a five-point rubric, with answers that were more coherent, idea-rich, and aligned with expert recommendations. However, their submissions did not necessarily demonstrate greater originality. The AI tool required students to refine prompts, evaluate responses, and curate final submissions, ensuring active engagement rather than passive use.
Critical-thinking training, which focused on causal reasoning through exercises unrelated to AI, did not improve rubric scores but led to a wider variety of more distinct ideas among students. Text analysis revealed these students questioned assumptions and explored multiple approaches, highlighting a gap in traditional grading methods that prioritize conventional answers over innovation.
The study concludes that ChatGPT enhances the quality and polish of student work, while critical-thinking training fosters originality and broader thinking. Students who received both interventions showed the strongest overall improvements, suggesting that AI and human-centered skills training can complement each other in education. The findings underscore the need for educational assessments to evolve beyond traditional rubrics to better measure creativity and reasoning.