Table 1.

This Table Demonstrates the Number of Correct Answers Selected by OpenAI's ChatGPT 3.5 and ChatGPT4 Based on Question Difficulty (Ranked by Board Vitals as “easy,” “moderate,” or “hard”)

Difficulty level	Correct answers	Total questions	Percentage, %
ChatGPT 3.5
Easy	23	39	58.97
Moderate	14	31	45.16
Hard	5	23	21.74
ChatGPT4
Easy	38	39	97.44
Moderate	28	31	90.32
Hard	11	23	47.83