TY  - RPRT
TI  - Treatment Effect Bias from Sample Snooping: Blinding Outcomes is Neither Necessary nor Sufficient
AU  - Aaron Fisher
PY  - 2022
UR  - https://arxiv.org/abs/2007.02514
ID  - 2007.02514
ER  -