Suppose you have data from a local firm that has been selling computers over the last 12 months. Load this dataset into Stata with the following command: use http://www. ssc. wisc. edu/-mckelvey/ec410/pcsales. dta, clear All units are in thousands of dollars, aside from sales (which is simply the number of computers sold) and income (which is in millions of dollars).

a) The number of computers sold (sales) may depend on factors aside from computer price (price), such as the price of a similar machine sold by a competitor (comp), TV advertising expenditure (tvadexp), radio advertising expenditure (radioadexp), and income of the purchasing community (income). Run the regression: sales = 60 + 01 price + 02comp + 03income +04tvadexp + 0radioadexp +E Interpret your estimate of 05.
b) Total advertising expenditure is the total amount spent across both TV and radio advertising (totadexp=radioadexp+tvadexp). Consider the following equation: sales = m + 1 price + 712comp + 73income + 74totadexp + 75radioadexp + 5 Interpret your estimate of 75. How does it differ from your interpretation of 05? c) Now run this regression: sales = lo + 1 price + l2comp + zincome + litotadexp+n What assumption does this regression make about the effect of TV versus radio advertising on sales? d) Finally, run the regression: price = Mo + Micomp + Mzincome + Mztotadexp+v and save the residuals (calling them rprice). Then run the regression: sales = 0o + Qirprice + 5 What is the relationship between 14 and 01. Explain why this is.