{"id":"W6926586816","doi":"10.25384/sage.20109174.v1","title":"sj-csv-1-las-10.1177_00238309221097978 – Supplemental material for Using Mahalanobis Distances to Investigate Second Dialect Acquisition: A Study on Quebec French","year":2022,"lang":"en","type":"dataset","venue":"Figshare","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Mahalanobis distance; Statistical analysis; Training set","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.001323182,0.002931273,0.002060264,0.004692456,0.001977761,0.003113546,0.004144736,0.002645619,0.2059572],"category_scores_gemma":[0.008734491,0.0008353919,0.001559645,0.00692929,0.0005538548,0.001214735,0.002008195,0.001758068,0.1438484],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004393183,"about_ca_system_score_gemma":0.007243752,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.5397753,"about_ca_topic_score_gemma":0.7201124,"domain_scores_codex":[0.9990827,0.0001311367,0.00007362151,0.000225206,0.0002667273,0.0002206002],"domain_scores_gemma":[0.9945109,0.001724842,0.000256817,0.0008873823,0.00217801,0.0004420847],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"observational","study_design_scores_codex":[0.00003198842,0.00001503715,0.0009708357,0.0003485367,0.00002996541,0.00001544066,0.00002736023,0.0001873674,0.00005883291,0.0001435046,0.9964846,0.001686467],"study_design_scores_gemma":[0.000445884,0.00002565609,0.01711654,0.0005522174,0.00007840011,0.00006064707,0.0002063979,0.000838643,0.0003866756,0.0009984076,0.979214,0.00007649047],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.00009541506,0.00004616748,0.00003742713,0.00003048485,0.00001377312,0.000008317604,0.9991493,0.0002604147,0.000358721],"genre_scores_gemma":[0.0004549595,0.00003020724,0.0002234232,0.00002583732,0.000006004731,0.00006975292,0.9981647,0.00009079962,0.0009343369],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.4602247,"threshold_uncertainty_score":0.9258702,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06862342982769716,"score_gpt":0.3262850243798612,"score_spread":0.2576615945521641,"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."}}