{"id":"W4386000296","doi":"10.31235/osf.io/ewamx","title":"Quality Control for Quality Computational Concepts: Wrangling with Theory and Data Wrangling as Theorizing","year":2023,"lang":"en","type":"preprint","venue":"","topic":"Computational and Text Analysis Methods","field":"Social Sciences","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institute on Governance","funders":"","keywords":"Formality; Quality (philosophy); Craft; Process (computing); Computer science; Control (management); Data quality; Management science; Data science; Epistemology; Artificial intelligence; Political science; Engineering; Law; Operations management","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02797882,0.0002897409,0.000737507,0.0001366123,0.0009581522,0.0004125453,0.0008005119,0.0002305894,0.0001145718],"category_scores_gemma":[0.004556769,0.0002417152,0.0001752539,0.0002594554,0.0006090037,0.0002249367,0.0005849674,0.0003747264,0.000009357394],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008285428,"about_ca_system_score_gemma":0.0007638259,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003722172,"about_ca_topic_score_gemma":0.00249226,"domain_scores_codex":[0.9933065,0.003609448,0.0007505937,0.00108138,0.0008742003,0.0003778939],"domain_scores_gemma":[0.9721897,0.02594189,0.000535548,0.0005273839,0.0006385731,0.0001668815],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0003340768,0.00003927304,0.003114565,0.0001307825,0.0005392621,0.000002301101,0.006047896,0.0194501,0.000003184673,0.9518209,0.00006419294,0.01845351],"study_design_scores_gemma":[0.001089079,0.00002372586,0.003220123,0.0001045636,0.0002084504,5.195794e-7,0.01315577,0.03162065,0.000001554196,0.9487817,0.001339723,0.0004541793],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0236901,0.0004032513,0.9677919,0.002086402,0.0004551845,0.0009113048,0.000537245,0.0003107107,0.003813945],"genre_scores_gemma":[0.9033561,0.00004298646,0.09266279,0.0006397172,0.0008268717,0.00008866529,0.0009818772,0.00003912682,0.001361864],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.879666,"threshold_uncertainty_score":0.9856858,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2420541306345986,"score_gpt":0.5429487259625833,"score_spread":0.3008945953279847,"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."}}