highly significantp<0.001
To summarize these results in a single conditional logit model, we added all highly significant (i.e. p<0.001) interactions to a backward-stepwise method of estimating the conditional logit model (Model 2).
To summarize these results in a single conditional logit model, we added all highly significant (i.e. p<0.001) interactions to a backward-stepwise method of estimating the conditional logit model (Model 2).