{"id":"W3110959118","doi":"10.1093/jeea/jvaa019","title":"How Much Can We Generalize From Impact Evaluations?","year":2020,"lang":"en","type":"article","venue":"Journal of the European Economic Association","topic":"Evaluation and Performance Assessment","field":"Decision Sciences","cited_by":232,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Exploit; Set (abstract data type); Sample (material); Government (linguistics); Sample size determination; Econometrics; Computer science; Treatment effect; Economics; Statistics; Mathematics; Computer security","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":["metaresearch"],"category_scores_codex":[0.4397657,0.003149829,0.009643949,0.008763031,0.002830533,0.0150814,0.007049141,0.008074475,0.006982554],"category_scores_gemma":[0.7986044,0.001412576,0.009151151,0.008533524,0.01708024,0.02524135,0.009448791,0.01435381,0.001824099],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.006335055,"about_ca_system_score_gemma":0.005641226,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005216292,"about_ca_topic_score_gemma":0.003250214,"domain_scores_codex":[0.521172,0.3888648,0.02381273,0.02327222,0.03933449,0.003543798],"domain_scores_gemma":[0.1939479,0.6649454,0.02948273,0.08495542,0.02444912,0.002219409],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.002674941,0.0007101773,0.08183376,0.02006378,0.03094743,0.0007977054,0.008842325,0.02243625,0.0006235447,0.2338235,0.0937321,0.5035146],"study_design_scores_gemma":[0.0005440893,0.000919694,0.02695257,0.01140514,0.004896136,0.0002739973,0.003278062,0.01186743,0.0009041481,0.8921083,0.04652406,0.0003264615],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08760428,0.1116225,0.382291,0.3121148,0.01992927,0.003901969,0.009055963,0.002365332,0.07111491],"genre_scores_gemma":[0.7949572,0.01515531,0.1021322,0.06967434,0.007449848,0.004581311,0.003396925,0.0006216338,0.002031286],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5602343,"threshold_uncertainty_score":0.6908687,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1830321189311144,"score_gpt":0.4270938186999413,"score_spread":0.2440616997688269,"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."}}