From data maintained by the multiple listing agency in Albuquerque, the file contains prices of home resales along with descriptive data on the home, including square feet, age, number of certain features, custom built or not, and corner location or not. Another dummy variable is whether the home was located in the Northest sector of the city. Realtors can use this kind of data to find an expected selling price for a home with given characteristics. The Northeast sector of the city is the largest residential area. It is more Anglo and more Republican than the rest of the city. An initial stepwise regression (F inclusion at 4.0) selects square feet, custom built, and age for an r-square of 82.2%. However, the question can be put whether the changing price expectation based on square feet, whether custom built, and age is the same in the Northeast sector as for the remainder of the city. Product variables with Northeast can be established and regres- sions including these variables run. The result is some alternative models that include these interactions at levels of significance exceeding 0.10 Since there are 49 missing values for AGE, it is useful to build a model as in the previous paragraph, but excluding age so that more observations will be used.