Initially we only considered multiple linear models without interactions.However, it is often useful to consider the possibility that two different predictive factors will also have a significant interaction.

Consider a model which predicts a consumer's interest in a product based on the consumer's age and the amount of advertising they have observed.Assuming both of these predictors are significant and positive in their effect, what would this mean?

(1) Do older or younger people have more interest? (1point)

(2) Does advertising help or hurt? (1point)

(3) Now suppose there is a statistically significant interaction between these two terms. What does this mean about the conclusions above? [3 points]