{"id":"W4413484588","doi":"10.1101/2025.08.21.25334182","title":"A Simulation Study to Advance Human-Centred Artificial Intelligence via Digital Citizen Science: Can Large Language Models Transform Current Approaches to Missing Data Imputation?","year":2025,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Computational and Text Analysis Methods","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"","keywords":"Imputation (statistics); Missing data; Computer science; Artificial intelligence; Data science; Current (fluid); Machine learning; Data mining; Engineering; Electrical engineering","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":[],"category_scores_codex":[0.02289241,0.0008016466,0.0009939086,0.0009416748,0.000857859,0.001829985,0.002323447,0.002410994,0.004280328],"category_scores_gemma":[0.08840274,0.0004674855,0.001714185,0.00143043,0.001465978,0.00203688,0.00214491,0.003208073,0.0003845647],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002300092,"about_ca_system_score_gemma":0.002305747,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01564173,"about_ca_topic_score_gemma":0.01326843,"domain_scores_codex":[0.9891193,0.009339542,0.0002559622,0.0006299579,0.0003201743,0.0003350716],"domain_scores_gemma":[0.792639,0.1927507,0.004140713,0.005319828,0.003922532,0.001227229],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00139144,0.0009307029,0.03452819,0.0004488378,0.0003154241,0.0003797719,0.001305068,0.9128767,0.0003760737,0.03026311,0.002474325,0.01471038],"study_design_scores_gemma":[0.0002847923,0.0003556643,0.001593614,0.00009996885,0.00006012497,0.00004148656,0.000456421,0.9837714,0.0003702202,0.01193213,0.00100477,0.00002938532],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8551546,0.0006099754,0.1313131,0.003449907,0.000155354,0.0008778906,0.002032218,0.0003050638,0.006101992],"genre_scores_gemma":[0.9400965,0.0001895332,0.05650891,0.0004464384,0.00004171811,0.0008848265,0.0009266825,0.00003450218,0.0008709047],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9771076,"threshold_uncertainty_score":0.1210681,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2879845659338913,"score_gpt":0.4615149857511824,"score_spread":0.1735304198172911,"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."}}