T-distributions are wider than Z-distributions, we use T-distributions when we do not know the population standard deviation.

for interaction plots: if there is no interaction, the lines on your plot will follow the same shape / appear parallel. vs if there is interaction they will differ in shape