{"id":"W4405906449","doi":"10.2196/59298","title":"Assessing the Data Quality Dimensions of Partial and Complete Mastectomy Cohorts in the <i>All of Us</i> Research Program: Cross-Sectional Study","year":2024,"lang":"en","type":"article","venue":"JMIR Cancer","topic":"Global Cancer Incidence and Screening","field":"Medicine","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Preprint; Cross-sectional study; Quality (philosophy); Psychology; Gerontology; Medicine; Computer science; Statistics; World Wide Web; Mathematics; Physics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.05051852,0.0003633705,0.0008366901,0.003345614,0.001291915,0.001928664,0.001514867,0.00107765,0.0009862431],"category_scores_gemma":[0.09736465,0.0005883455,0.002183835,0.003619576,0.001296267,0.002155568,0.003293552,0.001686324,0.0001565769],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001777987,"about_ca_system_score_gemma":0.00218906,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008031376,"about_ca_topic_score_gemma":0.007965231,"domain_scores_codex":[0.9657859,0.01466888,0.006912159,0.004400598,0.007047084,0.001185285],"domain_scores_gemma":[0.8832359,0.03271409,0.05792368,0.00947857,0.01303279,0.003614983],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00008122821,0.00006305105,0.9980557,0.00003557316,0.0002346087,0.00001243943,0.0004979369,0.00003411729,0.00003945487,0.00006293636,0.0001358882,0.0007472],"study_design_scores_gemma":[0.00002232492,0.0003128706,0.9963573,0.00006783635,0.0001695289,0.0001595235,0.001450654,0.0006055009,0.0001296928,0.0000831085,0.0006244079,0.00001733005],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9958662,0.0004405865,0.001459158,0.000224678,0.00001796893,0.0003678204,0.001209412,0.00001054694,0.0004035941],"genre_scores_gemma":[0.9966208,0.00008376395,0.00162829,0.0001483035,0.00001315231,0.0003830167,0.001048916,0.000008461125,0.00006533043],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9494815,"threshold_uncertainty_score":0.2671706,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.609466669380634,"score_gpt":0.6237969928738044,"score_spread":0.01433032349317043,"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."}}