Question: You have been hired by a New York (Long Island) real estate firm to assess the determinants of local house prices. You understand that house prices are determined by the interaction of supply and demand and opt to employ a "hedonic regression" framework. This framework specifies the price of a house to be a linear function of the various attributes of the house. Although the house is sold as a bundled commodity, you could estimate the contribution of various attributes in the determination of the final price. To assist this research, you conduct a survey of housing prices for five geographical locations in Nassau county. There are a total of 362 observations on 20 variables. Each variable contains data labels so you can understand how each one is coded. The original data may be found in https://www.wiley.com/college/ashenfelter. BASIC BUSINESS STATISTICS, 5/E by Berenson/Levin el Reprinted by permission of Pearson Education, Inc., Upper Saddle River, NI. The first step in using this data is to create dummy variables for certain qualitative variables.
a. What fraction of the houses in the sample are located in each region? What fraction of the houses in the sample are of each style? (Hint: Consider the means of the dummy variables.)
b. Rim a regression of "value" on the other variables (except taxes, which is a function of value), using the dummy variables you created in part (a). As noted above, a regression of this type is called a "hedonic" regression. Which variables are significantly different from zero at the 5% level and which are not? Interpret your output, responding in particular to the following expert advice: "Location is everything!", "Extra baths mean big bucks !", "Forget swimming pools, they don't pay!"
c. Explain how you could use this model to find "bargains" in the housing market.