3 Next you will perform a forward stepwise regression on the model with interactions (Model 2). For models with interactions, it is important to remember that if an interaction between two predicting variables is included in the model, then the main effects of the predicting variables should be included automatically. Thus for this variable selection, the minimum or starting model is the model with no interactions (Model 1) and the full model is the one with interactions (Model 2). Based on the output of this stepwise regression analysis, a The first three interaction terms entering the model are interactions between ac- cess and education factors. b The first two interaction terms entering the model are interactions between access and income. c There are interactions terms in the final selected model that are not statistically significant at significance level a = 0.05. d All interactions terms in the final selected model are statistically significant at significance level a = 0.01