{"id":"W4413033493","doi":"10.2196/78320","title":"Impact of Prompt Engineering on the Performance of ChatGPT Variants Across Different Question Types in Medical Student Examinations: Cross-Sectional Study","year":2025,"lang":"en","type":"article","venue":"JMIR Medical Education","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Preprint; Medical education; Psychology; Engineering; Medicine; Computer science; World Wide Web","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001322256,0.0001085927,0.0002100455,0.000189371,0.00006526169,0.0000131744,0.0001592252,0.0001640633,0.0003806432],"category_scores_gemma":[0.002451466,0.00007111477,0.00005505704,0.0004388068,0.00009346157,0.00006731275,0.00003258265,0.0003737506,0.000005817475],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000337044,"about_ca_system_score_gemma":0.001716572,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007408697,"about_ca_topic_score_gemma":0.0000779009,"domain_scores_codex":[0.9979671,0.000102655,0.0006866553,0.0001972991,0.0008765128,0.0001698287],"domain_scores_gemma":[0.998842,0.0003722839,0.0001291457,0.0002289282,0.0003062692,0.0001213995],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001056169,0.004508999,0.9663879,0.0001565839,0.00003633734,5.607685e-7,0.00371091,0.00004726345,0.00008671272,0.0005089047,0.00004136303,0.02440888],"study_design_scores_gemma":[0.0001537016,0.0006817117,0.9916425,0.0007568938,0.000009303043,0.000004088224,0.0009424039,0.005133453,0.0005796094,0.00004259449,0.000007062473,0.00004666591],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9966499,0.00006906259,0.00003790742,0.001303471,0.0008413782,0.001007585,0.00000172529,0.00001465265,0.00007438356],"genre_scores_gemma":[0.9991342,0.00005256107,0.000007500172,0.0001157857,0.0001823114,0.000393242,0.00002164059,0.000006964779,0.00008582336],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02525464,"threshold_uncertainty_score":0.4167776,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05309610483663259,"score_gpt":0.5138825237555716,"score_spread":0.4607864189189391,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}