5. The data set Apple on Moodle has the following variables:

id = respondent identifier

educ = respondents years of schooling

date = month/day/year of interview

state = home state

regprc = price of regular apples

ecoprc = price of ecolabeled apples

inseason = 1 if interviewed in November

hhsize = household size

male = 1 if respondent identifies as male

faminc = family income in thousands

age = respondents age in years

reglbs = quantity of regular apples consumed in pounds

ecolbs = quantity of ecolabeled apples consumed in pounds

numlt5 = number in household younger than 5

num5_17 = number in household between 5 and 17

num18_64 = number in household between 18 and 64

numgt64 = number in household older than 64

1. Estimate a regression of log(reglbs) on educ, log(regprc), log(ecoprc), inseason, hhsize, and log(faminc). Whats the advantage of specifying the price and quantity variables in log form? Test the overall significance of the model at the 1% level.
2. Create an interaction between inseason and log(regprc) and add it to the regression in (a). Interpret the coefficients on the dummy and the interaction. At the 10% level, is there a significant difference in the in-season vs. out-of-season intercepts? In the slopes?
3. At the 10% level, test the joint significance of the inseason dummy and its interaction with log(regprc). What is this a test for?
4. Add a quadratic in household size to the model in (a). Does apple consumption have a max or min with respect to household size? At what value? Are the linear and quadratic terms jointly significant at the 5% level?