{"id":"W2800160739","doi":"10.1177/0962280218772065","title":"Understanding and diagnosing the potential for bias when using machine learning methods with doubly robust causal estimators","year":2018,"lang":"en","type":"article","venue":"Statistical Methods in Medical Research","topic":"Advanced Causal Inference Techniques","field":"Mathematics","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"Hôpital du Sacré-Cœur de Montréal; Université de Montréal","funders":"Canadian Institutes of Health Research; Université de Montréal; Natural Sciences and Engineering Research Council of Canada; McGill University","keywords":"Propensity score matching; Resampling; Estimator; Causal inference; Average treatment effect; Computer science; Parametric statistics; Statistics; Econometrics; Mathematics","routes":{"ca_aff":true,"ca_fund":true,"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"],"consensus_categories":[],"category_scores_codex":[0.07292441,0.0009516353,0.001317752,0.002430845,0.0008555985,0.002698865,0.002377919,0.003140124,0.001923151],"category_scores_gemma":[0.3782891,0.0006969211,0.001216135,0.002022529,0.003131405,0.003508426,0.003695856,0.003012285,0.0002642588],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001117079,"about_ca_system_score_gemma":0.002022205,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002014792,"about_ca_topic_score_gemma":0.001261738,"domain_scores_codex":[0.9531012,0.03809416,0.002050429,0.002188676,0.004005,0.0005603943],"domain_scores_gemma":[0.6205645,0.3436551,0.01375098,0.01541167,0.006102091,0.0005156869],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005049084,0.0001957041,0.04841305,0.001093714,0.0007741065,0.000818419,0.001784344,0.238063,0.003523044,0.4822849,0.002287293,0.2202575],"study_design_scores_gemma":[0.0001048475,0.0001793757,0.005784141,0.000289007,0.0001358533,0.0004363876,0.0002137649,0.6333652,0.003942187,0.3514607,0.004008524,0.00008009294],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01988193,0.0005145323,0.9780963,0.0006460855,0.00003789253,0.00007733984,0.00005149596,0.0001130576,0.0005812405],"genre_scores_gemma":[0.4442809,0.000602021,0.552982,0.0005770903,0.000143792,0.0005363322,0.0002080711,0.000120208,0.0005496183],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9270756,"threshold_uncertainty_score":0.3856658,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.7613136427473822,"score_gpt":0.6566960719608058,"score_spread":0.1046175707865764,"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."}}