{"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":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","sts"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.03476872,0.0002631408,0.0005510924,0.0003193672,0.0007436246,0.0001688641,0.0004190655,0.00027,0.0004379062],"category_scores_gemma":[0.1369824,0.0001671073,0.00003734106,0.0005851453,0.002754969,0.0001465738,0.0004584675,0.002016694,0.000001025864],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003794815,"about_ca_system_score_gemma":0.0003435148,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002784778,"about_ca_topic_score_gemma":0.0001977308,"domain_scores_codex":[0.9897584,0.006555946,0.0006215433,0.0006040137,0.001432773,0.001027365],"domain_scores_gemma":[0.9116415,0.08707067,0.0001295788,0.0003526115,0.0003624804,0.000443102],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0005015178,0.000123336,0.0009048196,0.0003313897,0.00007600611,0.0001573852,0.001161834,0.00006562012,0.001079332,0.7937569,0.0004130883,0.2014288],"study_design_scores_gemma":[0.0004610142,0.0004929823,0.00004583966,0.0003022092,0.00003280981,0.00005305753,0.0005602171,0.3494297,0.001251275,0.6469744,0.0002357339,0.0001607622],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.001386898,0.00009111915,0.9962102,0.0009316527,0.00009928077,0.0008136451,0.00001599565,0.0001046747,0.0003464917],"genre_scores_gemma":[0.02449883,0.00004121457,0.9749656,0.00006868248,0.0001849407,0.0001247117,0.00000475955,0.00007457764,0.00003673253],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.3493641,"threshold_uncertainty_score":0.9999589,"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."}}