{"id":"W4307407759","doi":"10.1162/99608f92.13e1d60e","title":"The Use of Data Science in a National Statistical Office","year":2022,"lang":"en","type":"article","venue":"Harvard Data Science Review","topic":"Data Analysis with R","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Statistics Canada","funders":"","keywords":"Data science; Computer science","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.1577007,0.0008676118,0.001429736,0.007379442,0.003691574,0.01169575,0.00355019,0.005429953,0.004681365],"category_scores_gemma":[0.2483169,0.0008169844,0.001041713,0.01120803,0.02353781,0.01587445,0.00681858,0.01504665,0.003556926],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.009824232,"about_ca_system_score_gemma":0.02497109,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008457003,"about_ca_topic_score_gemma":0.003574512,"domain_scores_codex":[0.7951872,0.138755,0.01498232,0.01210322,0.03707491,0.001897259],"domain_scores_gemma":[0.5702175,0.2780346,0.01969665,0.05236619,0.07450867,0.005176491],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00002865022,0.00002464159,0.001012372,0.001137051,0.00006332695,0.00007714754,0.0007566739,0.0007008445,0.0002400652,0.7596975,0.1072438,0.1290179],"study_design_scores_gemma":[0.00001160094,0.00003078152,0.0008432032,0.00205377,0.00002833327,0.0001153144,0.0003703076,0.0007384562,0.0002734263,0.1117066,0.8837913,0.00003701291],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"commentary","genre_gemma":"empirical","genre_scores_codex":[0.002094476,0.1620041,0.2050534,0.5019432,0.02115439,0.0003209406,0.001622518,0.001073619,0.1047333],"genre_scores_gemma":[0.1353939,0.2499857,0.3646327,0.1696845,0.04383723,0.001572199,0.002706324,0.002333509,0.02985372],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.1577007,"threshold_uncertainty_score":0.8340108,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2546039632704132,"score_gpt":0.3974377244224109,"score_spread":0.1428337611519977,"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."}}