{"id":"W2742988910","doi":"10.21767/1791-809x.1000510","title":"Methodological Challenges to Collecting Valid Research Data: Researching English Language Barrier Populations in Canada","year":2017,"lang":"en","type":"article","venue":"Health Science Journal","topic":"HIV, Drug Use, Sexual Risk","field":"Medicine","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Language barrier; Open access publishing; Computer science; Data science; World Wide Web; Natural language processing; Linguistics; Philosophy","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":["metaresearch"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.335543,0.0009719557,0.002468995,0.007406204,0.02065441,0.01397831,0.009899396,0.003971504,0.003038432],"category_scores_gemma":[0.5481225,0.001438259,0.00181424,0.01350453,0.01348549,0.007932089,0.01138282,0.006924774,0.0004913208],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.07706001,"about_ca_system_score_gemma":0.2795091,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.872136,"about_ca_topic_score_gemma":0.9072924,"domain_scores_codex":[0.6634369,0.2194151,0.02944736,0.009762196,0.06597684,0.01196167],"domain_scores_gemma":[0.4740947,0.2992438,0.05161416,0.02502286,0.1354837,0.01454065],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"qualitative","study_design_scores_codex":[0.0004637181,0.000453673,0.3392995,0.03000686,0.001675897,0.001707456,0.2958484,0.002186241,0.0006938265,0.05541937,0.0480557,0.2241893],"study_design_scores_gemma":[0.0003603568,0.0007727584,0.3160172,0.1121827,0.001299867,0.001097615,0.3763771,0.005747623,0.002160517,0.0399351,0.1435668,0.0004824147],"study_design_candidate":"qualitative","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.3856497,0.104594,0.1132098,0.2925817,0.004308274,0.04189867,0.00714996,0.0002969509,0.05031088],"genre_scores_gemma":[0.7798471,0.02766668,0.1032195,0.04383804,0.000532502,0.04022814,0.001915173,0.0001676512,0.002585187],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.664457,"threshold_uncertainty_score":0.8193938,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.8479707906009973,"score_gpt":0.6553598181891445,"score_spread":0.1926109724118529,"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."}}