D Question 14 1 pts How does the F-ratio change when you use dummy coding, contrast coding, or post-hoc tests? O It changes based on the normality of the residuals O It doesn't change, because the F-ratio looks at the whole model It can't be used for continuous predictors It's more robust for contrast coding Question 15 1 pts What is the problem with testing many groups from the same dataset against each other? O Increases the likelihood of finding a "significant" difference when there is no real effect (risk of Type 1 error) O Increases the likelihood of getting too many significant results to interpret There is no problem with this O Increases the likelihood of missing a "significant" difference between groups when there is a real effect (risk of Type 2 error)