Many applied economic investigations, for modelling different types of food consumption/ purchase, use a range of explanatory variables. Such models usually are conducted at the household level and include three independent variables income and education of household members and the price of the food item. (Some models also include the prices of the substitutes and complements.) Consider the following simple linear model to explain monthly beer consumption:

beer=0+1income+2price+3education+4income*education+u

E(u | income, price, education)=0,

Var(u | income, price, education)= (income/education)^2

Can the results of the regression in scenario one be used to make an inference on coefficients? Explain your answer in no more than 100 words. Referring to question above, write the transformed and simplified equation with a homoscedastic error term. Also, mathematically show the new transformed error term is now homoscedastic.