{"id":"W4407369950","doi":"10.47004/wp.cem.2025.0725","title":"Point-identifying semiparametric sample selection models with no excluded variable","year":2025,"lang":"en","type":"report","venue":"","topic":"Statistical Methods and Inference","field":"Mathematics","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Sample (material); Variable (mathematics); Selection (genetic algorithm); Semiparametric model; Semiparametric regression; Computer science; Econometrics; Statistics; Feature selection; Mathematics; Artificial intelligence; Nonparametric statistics; Chemistry; Chromatography","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01845056,0.0009853392,0.002508293,0.001366806,0.0008259136,0.001832738,0.003170091,0.001600928,0.006081224],"category_scores_gemma":[0.07592187,0.0008758678,0.00194969,0.001720501,0.002459241,0.003370014,0.004193266,0.00308271,0.00136927],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00082939,"about_ca_system_score_gemma":0.001434675,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001599252,"about_ca_topic_score_gemma":0.001599772,"domain_scores_codex":[0.9841733,0.01224412,0.0004601415,0.001262389,0.001472298,0.0003878196],"domain_scores_gemma":[0.9340304,0.05206167,0.005272119,0.006036991,0.002187331,0.0004114586],"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":[0.0002828684,0.0002895369,0.02164339,0.0004837612,0.0005087139,0.0006075609,0.001220897,0.1602647,0.00120582,0.6398872,0.004867275,0.1687383],"study_design_scores_gemma":[0.0001077635,0.0002398488,0.003403736,0.00009416294,0.0001066648,0.0001903461,0.0002005041,0.5396634,0.001081261,0.4500788,0.004776442,0.00005713444],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.009149664,0.00008494352,0.9891737,0.0002228231,0.00001784026,0.0001165171,0.0001053912,0.0001539668,0.0009751035],"genre_scores_gemma":[0.6297898,0.0005796532,0.3583314,0.0005195825,0.0002042906,0.001753181,0.0009684816,0.0001737531,0.007679803],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01845056,"threshold_uncertainty_score":0.09757704,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2071953127708966,"score_gpt":0.4136875120752775,"score_spread":0.2064921993043809,"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."}}