For dummy*interval interaction models:

(a) In a basic model, with one interval predictor (X), one dummy predictor (D), and the product term of the dummy times the interval predictor (D*X), how would you interpret the intercept and three variables’ coeffi cients?

(b) How would you write the conditional regression equations for this model within the two levels of the dummy variable?

(c) How would you write the conditional regression equations for this model within levels of the interval predictor variable?

(d) What would be the null and alternative hypotheses, and the corresponding reduced and full models, for the Chow test based on this model?

(e) What would you expect the results to be if you reverse coded the dummy variable, omitting the category that had previously been coded “1”, created a new interaction term based on this new reverse coded dummy, and re-estimated the model with the new dummy and new interaction term?