In simulations of randomized experiments with 500 people per arm, using the randomization when imputing missing covariates gives only small accuracy gains, while ignoring the outcome can bias heterogeneous treatment effect estimates.
write newline
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2019 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Leveraging Random Assignment to Impute Missing Covariates in Causal Studies
In simulations of randomized experiments with 500 people per arm, using the randomization when imputing missing covariates gives only small accuracy gains, while ignoring the outcome can bias heterogeneous treatment effect estimates.