{"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":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001724729,0.0004278168,0.000941219,0.0005055073,0.0001531404,0.0001920799,0.0002557439,0.0003708148,0.002553814],"category_scores_gemma":[0.01638321,0.0003248524,0.0001059047,0.001412763,0.00004240403,0.0001407928,0.000134802,0.0006859389,0.00002852725],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004382996,"about_ca_system_score_gemma":0.001437839,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003523878,"about_ca_topic_score_gemma":0.00007386926,"domain_scores_codex":[0.9969352,0.000169242,0.000758685,0.0006699584,0.0009935284,0.0004733111],"domain_scores_gemma":[0.9908524,0.006396555,0.0003966713,0.0005213975,0.001707406,0.0001255907],"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":[0.00006337062,0.0002612292,0.0001171926,0.004582587,0.0005402979,0.00001318356,0.00003974544,0.0001266332,0.00004632706,0.8193082,0.1577169,0.01718428],"study_design_scores_gemma":[0.0002725488,0.0001435031,0.00001469097,0.0009972079,0.000438716,0.00005220694,0.0000233581,0.02738319,0.0001269256,0.9518085,0.01819492,0.0005442147],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.00001441508,0.00005045364,0.630459,0.000006792085,0.0004249462,0.0003545168,0.00008841493,0.0001672777,0.3684343],"genre_scores_gemma":[0.0001875365,0.0002314585,0.9020127,0.00005236149,0.0001791941,0.0001029319,0.00006601463,0.00005747435,0.09711033],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.2715538,"threshold_uncertainty_score":0.9999204,"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."}}