{"id":"W6921167955","doi":"10.6084/m9.figshare.29755804.v1","title":"Additional file 1 of Using a large language model (ChatGPT) to assess risk of bias in randomized controlled trials of medical interventions: protocol for a pilot study of interrater agreement with human reviewers","year":2025,"lang":"en","type":"article","venue":"Figshare","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan","funders":"","keywords":"Protocol (science); Inter-rater reliability; Randomized controlled trial; Calibration; Measure (data warehouse); Data collection","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.04304839,0.002083931,0.003043355,0.003922522,0.001802943,0.002875203,0.001996219,0.002552671,0.8015149],"category_scores_gemma":[0.3027093,0.002408332,0.003773222,0.004738977,0.001242675,0.002795298,0.001983146,0.002849316,0.06837232],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003101665,"about_ca_system_score_gemma":0.008259133,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002266739,"about_ca_topic_score_gemma":0.007239447,"domain_scores_codex":[0.9759292,0.01535122,0.004984848,0.001675485,0.001474197,0.0005852191],"domain_scores_gemma":[0.5703613,0.3827077,0.01649916,0.01438146,0.01419004,0.001860439],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.01613422,0.0007569172,0.001981536,0.09538727,0.001354936,0.0001265919,0.0008087258,0.002442066,0.0003546162,0.008659474,0.8133516,0.05864201],"study_design_scores_gemma":[0.3115838,0.004534912,0.02090148,0.05951981,0.004334355,0.0006087057,0.001174021,0.0159497,0.002238471,0.08115876,0.4970866,0.0009094083],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.001758566,0.0003469962,0.01722348,0.001137101,0.0005063482,0.157689,0.8122483,0.002940777,0.006149447],"genre_scores_gemma":[0.008784841,0.0002032504,0.04753009,0.0008846888,0.0001598938,0.900777,0.03411924,0.0008520283,0.006688983],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.9569516,"threshold_uncertainty_score":0.2831149,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.6404749868466397,"score_gpt":0.5838550379586818,"score_spread":0.05661994888795796,"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."}}