=+4. It may be worthwhile to focus on whether the intercepts are also different for the two groups. Given that the interaction was statistically significant, determine whether the associated with the categorical variable (with the continuous variable and the interaction term already in the regression equation) is statistically significant.

The most general method for doing so is to do a simultaneous regression of the outcome on the continuous variable and the interaction, sequentially adding the categorical variable to the equation. With only two categories, and thus one dummy variable, this same information can be garnered from the statistical significance of the b of the categorical variable with all three terms in the regression equation. If the categorical variable is statistically significant, you should next focus on the graph of the interaction. If it makes sense to interpret the differences in intercept, do so. If the categorical variable is not statistically significant, or this difference is not interpretable, focus only on the slope bias. Stop.