Performance of ChatGPT on the India Undergraduate Community Medicine Examination: Cross-Sectional Study
Bibliographic record
Abstract
BACKGROUND: Medical students may increasingly use large language models (LLMs) in their learning. ChatGPT is an LLM at the forefront of this new development in medical education with the capacity to respond to multidisciplinary questions. OBJECTIVE: The aim of this study was to evaluate the ability of ChatGPT 3.5 to complete the Indian undergraduate medical examination in the subject of community medicine. We further compared ChatGPT scores with the scores obtained by the students. METHODS: The study was conducted at a publicly funded medical college in Hyderabad, India. The study was based on the internal assessment examination conducted in January 2023 for students in the Bachelor of Medicine and Bachelor of Surgery Final Year-Part I program; the examination of focus included 40 questions (divided between two papers) from the community medicine subject syllabus. Each paper had three sections with different weightage of marks for each section: section one had two long essay-type questions worth 15 marks each, section two had 8 short essay-type questions worth 5 marks each, and section three had 10 short-answer questions worth 3 marks each. The same questions were administered as prompts to ChatGPT 3.5 and the responses were recorded. Apart from scoring ChatGPT responses, two independent evaluators explored the responses to each question to further analyze their quality with regard to three subdomains: relevancy, coherence, and completeness. Each question was scored in these subdomains on a Likert scale of 1-5. The average of the two evaluators was taken as the subdomain score of the question. The proportion of questions with a score 50% of the maximum score (5) in each subdomain was calculated. RESULTS: ChatGPT 3.5 scored 72.3% on paper 1 and 61% on paper 2. The mean score of the 94 students was 43% on paper 1 and 45% on paper 2. The responses of ChatGPT 3.5 were also rated to be satisfactorily relevant, coherent, and complete for most of the questions (>80%). CONCLUSIONS: ChatGPT 3.5 appears to have substantial and sufficient knowledge to understand and answer the Indian medical undergraduate examination in the subject of community medicine. ChatGPT may be introduced to students to enable the self-directed learning of community medicine in pilot mode. However, faculty oversight will be required as ChatGPT is still in the initial stages of development, and thus its potential and reliability of medical content from the Indian context need to be further explored comprehensively.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.007 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".