RT Article
T1 Using recursive partitioning to find and estimate heterogenous treatment effects in randomized clinical trials
JF Journal of experimental criminology
VO 17
IS 3
SP 519
OP 538
A1 Berk, Richard
A2 Olson, Matthew
A2 Buja, Andreas
A2 Ouss, Aurélie
LA English
YR 2021
UL https://krimdok.uni-tuebingen.de/Record/176789614X
AB When for an RCT heterogeneous treatment effects are inductively obtained, significant complications are introduced. Special loss functions may be needed to find local, average treatment effects followed by techniques that properly address post-selection statistical inference., Reanalyzing a recidivism RCT, we use a new form of classification trees to seek heterogeneous treatment effects and then correct for “data snooping” with novel inferential procedures., There are perhaps increases in recidivism for a small subset of offenders whose risk factors place them toward the right tail of the risk distribution., A legitimate but partial account for uncertainty might well reject the null hypothesis of no heterogenous treatment effects. An equally legitimate but far more complete account of uncertainty for this study fails to reject the null hypothesis of no heterogeneous treatment effects.
K1 Decision trees
K1 Post-selection statistical inference
K1 Heterogeneous treatment effects
K1 Randomized experiments
DO 10.1007/s11292-019-09410-0