{"id":"W3038781317","doi":"10.1093/ije/dyaa087","title":"Recognizing, reporting and reducing the data curation debt of cohort studies","year":2020,"lang":"en","type":"article","venue":"International Journal of Epidemiology","topic":"Health, Environment, Cognitive Aging","field":"Environmental Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Economic and Social Research Council; Medical Research Council; Department of Health and Aged Care, Australian Government; Canadian Institutes of Health Research; National Institute for Health and Care Research; European Commission; Connected Health Cities; Wellcome Trust","keywords":"Debt; Data curation; Context (archaeology); Technical debt; Data science; Computer science; Business; Finance; Software; Software development; Biology","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.008341838,0.00007766281,0.0003105921,0.0000272319,0.00005826842,0.000005187821,0.0004419281,0.00003779732,0.0001171068],"category_scores_gemma":[0.04592113,0.00005305458,0.00004000508,0.00005411889,0.0003030781,0.0003867163,0.0004789765,0.0002571035,0.000008527138],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008600554,"about_ca_system_score_gemma":0.00002220627,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001176714,"about_ca_topic_score_gemma":0.00001488846,"domain_scores_codex":[0.9972818,0.0004950399,0.001590161,0.0002659404,0.0002317247,0.0001352722],"domain_scores_gemma":[0.9945583,0.00193006,0.003195567,0.0001758274,0.00006300453,0.00007718211],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00006034196,0.00002402454,0.9015519,0.00001201971,0.0004409056,0.00004630344,0.001543508,0.001934871,0.004617538,0.0001725758,0.00453598,0.08506004],"study_design_scores_gemma":[0.000727355,0.0003408648,0.9396167,0.0002964157,0.0002302492,0.001021509,0.00225338,0.02169825,0.001488699,0.009807407,0.02226986,0.0002493001],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9514191,0.00111857,0.007878271,0.03833928,0.0004006441,0.000127136,0.000008565305,0.0000046237,0.0007038694],"genre_scores_gemma":[0.9871299,0.002521118,0.006591846,0.003457396,0.000274473,0.000001550185,0.000007597789,0.000006518799,0.000009564478],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.08481074,"threshold_uncertainty_score":0.9621155,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3123556779361549,"score_gpt":0.4488741108666015,"score_spread":0.1365184329304465,"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."}}