Suppose that the estimated coefficients b1 b2 and b3 are

Question

Consider the two regression models:

Pizza=β₁+β₂Income+β₃Age+β₄(Age×Income)+u

(1)

Pizza=β₁+β_2Income+β₃Age+u

(2)

where    Pizza   = Individual's annual expenditure on pizza

Income   = Individual's annual income

Age     = Individual's age in years

Suppose that the estimated coefficients b₁, b₂, and b₃  are all positive in both Model (1) and (2) and b₄  is also positive in Model (1).

By including the interaction term (Age×Income) in Model (1), what can be examined about the relationship between Pizza and Income and Age beyond what is examined in the simpler Model (2)?