Table 1 ChatGPT performance for multiple-choice questions with prompt 4 (relevant paragraph)
Cognitive taxonomy level (Number of test items) | Correct answers to Prompt 4 (relevant paragraph) (%) | Overall accuracy over four lead-in prompts (%) |
|---|---|---|
Recall (44) | 81.8 | 86.4 |
Interpretation (20) | 75.0 | 77.5 |
Problem-solving (14) | 35.7 | 41.1 |
Total test items (78) | 71.8 | 76.0 |