{"id":"W2030541374","doi":"10.1017/s0954394512000129","title":"Models, forests, and trees of York English:<i>Was/were</i>variation as a case study for statistical practice","year":2012,"lang":"en","type":"article","venue":"Language Variation and Change","topic":"Linguistic Variation and Morphology","field":"Social Sciences","cited_by":753,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta; University of Toronto","funders":"","keywords":"Multicollinearity; Plural; Variation (astronomy); Inference; Causal inference; Random forest; Econometrics; Interpretation (philosophy); Verb; Moderation; Computer science; Statistics; Linguistics; Variable (mathematics); Argument (complex analysis); Natural language processing; Artificial intelligence; Mathematics; Linear regression","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":[],"consensus_categories":[],"category_scores_codex":[0.001206386,0.00009034845,0.0001584476,0.00007149506,0.0002285878,0.00004935279,0.00004388306,0.0001039238,0.00008979395],"category_scores_gemma":[0.002495836,0.00008658177,0.00001778493,0.0001064716,0.00004603012,0.0003317946,0.00002672866,0.00006131561,0.000001580072],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002913347,"about_ca_system_score_gemma":0.00004531485,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.04961462,"about_ca_topic_score_gemma":0.01111332,"domain_scores_codex":[0.9989375,0.0002836393,0.0002134548,0.0001677948,0.0001870784,0.0002104962],"domain_scores_gemma":[0.9986803,0.0007072826,0.0001855125,0.0001002536,0.0001944298,0.0001322302],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"qualitative","study_design_scores_codex":[0.00004864035,0.0002556634,0.00109728,0.00001614358,0.00003191676,0.00001940494,0.8728898,8.261636e-7,0.0000157703,0.1234778,0.00006835645,0.002078451],"study_design_scores_gemma":[0.005702847,0.0009740703,0.1114744,0.00002818882,0.0008755313,0.0002095277,0.8528892,0.005740423,0.00001883015,0.007341122,0.0141079,0.0006379932],"study_design_candidate":"qualitative","study_design_consensus":"qualitative","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9308868,0.001370326,0.05775229,0.001167529,0.001631484,0.002785317,0.000244175,0.00015058,0.004011496],"genre_scores_gemma":[0.9940102,0.00004166248,0.00432484,0.0003642518,0.0009646175,0.0001143348,0.0000317426,0.00001055868,0.00013778],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1161366,"threshold_uncertainty_score":0.9567141,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06490848551197312,"score_gpt":0.3640252859992587,"score_spread":0.2991168004872855,"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."}}