{"id":"W7130675486","doi":"10.1109/swc65939.2025.00307","title":"Generating Synthetic Categorical Data for Causal Modelling of Accessibility Barriers","year":2025,"lang":"","type":"article","venue":"","topic":"Disability Rights and Representation","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"","keywords":"Categorical variable; Statistical model; Bayesian network; Generative grammar; Synthetic data; Probabilistic logic; Copula (linguistics); Causal model; Data modeling; Bayesian probability","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007323458,0.0005131564,0.0004350921,0.001272882,0.00041238,0.0007707172,0.00118392,0.0008907482,0.002175485],"category_scores_gemma":[0.0373424,0.0002975604,0.0008676291,0.001364729,0.0007059621,0.0009757566,0.00112722,0.001332683,0.0003079183],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001072232,"about_ca_system_score_gemma":0.0008218027,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004847672,"about_ca_topic_score_gemma":0.004952477,"domain_scores_codex":[0.9964675,0.002662231,0.0001064523,0.0003513446,0.0003172416,0.00009524112],"domain_scores_gemma":[0.9469588,0.04608256,0.001921824,0.003044814,0.001733024,0.0002589948],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002146761,0.0003193192,0.036504,0.0002768166,0.0001414342,0.0002251172,0.0007098012,0.8708378,0.001057346,0.04211539,0.002690475,0.04490773],"study_design_scores_gemma":[0.00001842738,0.0000529504,0.002557397,0.00002904452,0.00001247194,0.00004380513,0.0001736927,0.9761881,0.0006902281,0.01906159,0.001158051,0.00001421371],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3421374,0.0002367198,0.6484346,0.0008484059,0.0001283572,0.0004645347,0.004710848,0.0006523262,0.002386879],"genre_scores_gemma":[0.8716196,0.000123318,0.1216714,0.0001396344,0.00002577544,0.0005344265,0.005119285,0.00005269907,0.0007139281],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007323458,"threshold_uncertainty_score":0.03873062,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1279443924268044,"score_gpt":0.4034514060749615,"score_spread":0.2755070136481572,"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."}}