{"id":"W2606149955","doi":"10.1016/j.jclinepi.2017.04.001","title":"Magnitude and direction of missing confounders had different consequences on treatment effect estimation in propensity score analysis","year":2017,"lang":"en","type":"article","venue":"Journal of Clinical Epidemiology","topic":"Advanced Causal Inference Techniques","field":"Mathematics","cited_by":3,"is_retracted":false,"has_abstract":false,"ca_institutions":"McMaster University; Population Health Research Institute","funders":"Cancer Research UK","keywords":"Confounding; Propensity score matching; Medicine; Inverse probability weighting; Observational study; Statistics; Population; Odds ratio; Internal medicine; Mathematics; Environmental health","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":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.007437641,0.0001685676,0.002670424,0.0001874321,0.00008031584,0.00001471674,0.0001486499,0.0002037328,0.000009038361],"category_scores_gemma":[0.0812901,0.00009568697,0.0003541601,0.00005108016,0.000677385,0.0001405691,0.00003417674,0.0003623832,3.616774e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009957464,"about_ca_system_score_gemma":0.00006118102,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001406778,"about_ca_topic_score_gemma":0.0002680116,"domain_scores_codex":[0.994963,0.002322227,0.002242139,0.0001951189,0.0001198687,0.0001576179],"domain_scores_gemma":[0.9677014,0.02752573,0.004244849,0.0002964021,0.0001287211,0.0001028848],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0005402254,0.0002438713,0.9058064,0.00005605015,0.0004032229,0.00002790765,0.00003834679,0.0001726875,0.0002184807,0.005293669,0.00001457075,0.08718453],"study_design_scores_gemma":[0.0007768541,0.002657009,0.7496063,0.0002608934,0.0004392294,0.00002231701,0.00000612902,0.002491938,0.001199865,0.2424615,0.000004282979,0.00007375008],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9534532,0.00005617557,0.04474183,0.001261755,0.0001261726,0.0002013535,0.00000138112,0.00001100759,0.0001470713],"genre_scores_gemma":[0.9820866,0.0003618586,0.01742035,0.00007139002,0.00003843857,0.000004301507,9.255502e-7,0.000005740904,0.00001040633],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2371678,"threshold_uncertainty_score":0.9264486,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.6695902067594982,"score_gpt":0.6039816511912118,"score_spread":0.06560855556828638,"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."}}