Logistic Regression with Moderation and Control Variable

Question

I perform a Logistic Regression in R which looks the folllowing:

logit <- glm(result~Condition+Strength, data = combined, family = binomial)

I am mainly interested in the effect of Condition on result while controlling for Strength. So far the effect is significant. Furthermore I want to include another variable called Focus which I expect to be a Moderator for the effect of Condition on result. Can anyone tell me how to do this, while still controlling for Strength, as without this controlling variable the effect completely disappears?

Answer 1

Can anyone tell me how to do this, while still controlling for Strength

logit <- glm(result ~ Condition*Focus + Strength, data = combined, family = binomial)

R will automatically calculate the main effects of Condition and Focus for you (assuming that you have set both as.factor()), so you don't have to specify the Condition + Focus part if you don't want to.

as without this controlling variable the effect completely disappears?

Do not simply include a covariate because it yields significant results; this can inflate Type I error and may be considered p-hacking unless you have a solid theoretical reason to include the covariate. Including it simply for statistical significance is frowned-upon.