Nast Stores has derived the following consumer credit-scoring model after years of data collecting and model testing:
Y = (0.20 × EMPLOYMT) + (0.4 × HOMEOWNER) + (0.3 × CARDS)
EMPLOYMT = 1 if employed full-time, 0.5 if employed part-time, and 0 if unemployed
HOMEOWNER = 1 if homeowner, 0 otherwise
CARDS = 1 if presently has 1–5 credit cards, 0 otherwise
Nast determines that a score of at least 0.70 indicates a very good credit risk, and it extends credit to these individuals.
1c. Your boss mentions that he just returned from a trade-association conference, at which one of the speakers recommended that length of time at present residence (regardless of homeownership status) be included in credit-scoring models. If the weight turns out to be 0.25, how do you think the variable would be coded (i.e., 0 stands for what, 1 stands for what, etc.)?
1d. Suggest other variables that Associated might have left out of the model, and tell how you would code them (i.e., 0, 1, 2 are assigned to what conditions or variables?).