Skip to main content

Table 3 Covariates included in the simulation as a function of their association with treatment allocation and outcome

From: How to perform prespecified subgroup analyses when using propensity score methods in the case of imbalanced subgroups

Outcome

Treatment allocation

 

Absent

Moderate

Strong

Absent

X3

 

X1

Moderate

\(X9|S=1\)

X2 ; X5

X7 ; \(X9|S=2\)

Strong

\(X4|S=2\) ; X6

X8

\(X4|S=1\) ; X10

  1. Strong, moderate and absence of impact were set by parameter values of \(\log 2\), \(\log 1.3\) and \(\log 1\), respectively
Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us