{"id":"W4416260299","doi":"10.48550/arxiv.2506.21744","title":"Federated Item Response Models: A Gradient-driven Privacy-preserving Framework for Distributed Psychometric Estimation","year":2025,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Psychometric Methodologies and Testing","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Social Sciences and Humanities Research Council of Canada; Connaught Fund; University of Toronto","keywords":"Item response theory; Estimation; Field (mathematics); Raw data; Information privacy; Data modeling; Federated learning; Privacy protection","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","metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.006640038,0.0005876529,0.001064015,0.003832245,0.0007227394,0.0007675794,0.003838612,0.0008835367,0.00008757257],"category_scores_gemma":[0.2342378,0.0005751266,0.0006255382,0.01312811,0.000141924,0.0005474816,0.003093256,0.001069615,0.00002497096],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005119304,"about_ca_system_score_gemma":0.0004052806,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008152374,"about_ca_topic_score_gemma":0.000008807527,"domain_scores_codex":[0.9933145,0.001855784,0.001062107,0.002443491,0.0005535533,0.0007705669],"domain_scores_gemma":[0.9127311,0.08140244,0.00151258,0.002489412,0.001512896,0.0003515852],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00172828,0.0001783407,0.01194283,0.0001186512,0.0002012447,0.00002910883,0.0002150269,0.947031,0.0000111813,0.02424964,0.004742166,0.009552512],"study_design_scores_gemma":[0.0005122668,0.00006865149,0.007731095,0.0002242665,0.00007532985,0.000001423148,0.0002492778,0.5729892,0.00001241754,0.4174351,0.0003696512,0.0003313993],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3148933,0.0001565796,0.6813078,0.0003289517,0.001252697,0.0008491473,0.0004565334,0.0002611156,0.0004938939],"genre_scores_gemma":[0.8827434,0.0001072723,0.1152382,0.00008648261,0.00007384055,0.00001342355,0.0001141822,0.0000274478,0.001595671],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5678502,"threshold_uncertainty_score":0.99967,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5639737864270469,"score_gpt":0.3737501621731419,"score_spread":0.1902236242539049,"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."}}