Skip to main content

Table 2 Estimates of \(\theta _1\) obtained using various missing data methods in the case study (n=953)

From: Should multiple imputation be stratified by exposure group when estimating causal effects via outcome regression in observational studies?

Method

\(\theta _1\) estimate

Standard error

P-value

95% confidence interval

CCA

0.282

0.136

0.038

0.016, 0.549

MI-NI

0.296

0.124

0.018

0.051, 0.540

MI-E\(\times\)O

0.301

0.123

0.015

0.059, 0.544

MI-E\(\times\)C

0.301

0.127

0.018

0.051, 0.551

MI-E\(\times\)OC

0.301

0.122

0.014

0.062, 0.540

MI-E\(\times\)I

0.294

0.121

0.015

0.057, 0.532

MI-EG

0.273

0.122

0.026

0.033, 0.514