{"id":"W105148950","doi":"10.1002/j.1839-4655.2014.tb00322.x","title":"Evidence for the use of an algorithm in resolving inconsistent and missing Indigenous status in administrative data collections","year":2014,"lang":"en","type":"article","venue":"Australian Journal of Social Issues","topic":"Data Analysis and Archiving","field":"Social Sciences","cited_by":162,"is_retracted":false,"has_abstract":true,"ca_institutions":"Centre for Global Health Research","funders":"NSW Ministry of Health","keywords":"Indigenous; Missing data; Identification (biology); Life expectancy; Project commissioning; Consistency (knowledge bases); Data quality; Linkage (software); Publishing; Computer science; Data science; Data mining; Demography; Sociology; Political science; Operations management; Engineering; Machine learning; Artificial intelligence; Population","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":["metaresearch"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.3371085,0.001475809,0.001480223,0.004464267,0.003619337,0.009294737,0.004797504,0.00507597,0.003107482],"category_scores_gemma":[0.5934406,0.001673116,0.003611199,0.008673054,0.005230995,0.01166936,0.004705593,0.004948627,0.00149186],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003466898,"about_ca_system_score_gemma":0.008840763,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009924577,"about_ca_topic_score_gemma":0.008185443,"domain_scores_codex":[0.5757947,0.3531676,0.02529727,0.01457638,0.02946805,0.001695971],"domain_scores_gemma":[0.2343768,0.6444618,0.02685468,0.04387884,0.04914094,0.001286987],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002885888,0.001169811,0.2705529,0.005335732,0.004078642,0.0001972583,0.01037125,0.03828086,0.001293518,0.04782929,0.009989378,0.6080155],"study_design_scores_gemma":[0.001767924,0.003058396,0.2433782,0.01213633,0.003945258,0.002958447,0.008680397,0.5055962,0.01254859,0.1327883,0.07221424,0.0009276868],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1696497,0.006614413,0.7899539,0.009954488,0.0007260006,0.003183175,0.001157528,0.00146454,0.01729628],"genre_scores_gemma":[0.1654256,0.001189443,0.829162,0.001353281,0.0001156931,0.0009764954,0.0006025298,0.0004845361,0.0006903211],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6628915,"threshold_uncertainty_score":0.8174633,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3896094153131164,"score_gpt":0.4706682045091592,"score_spread":0.08105878919604281,"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."}}