The parameters of logit models are typically difficult to interpret, and the applied literature is replete with interpretive and computational mistakes. In this article, I review a menu of options to
The models are written in the following way in R (example given for model 1a): log(Time_F) ~ Cond + log(STS) + PMS + log(Current_Pract) + FWA + Current_Pract:FWA. The following significance codes appl
NOVA tests whether independent variables significantly affect the dependent variable overall, while the Coefficients table explains each predictor’s direction, magnitude, and significance, enabling in
Normalized Joint and Marginal Shannon information content and normalized mutual information content. Data corresponds to results of Figure 11 in paper.
