To develop a credit scoring system or model, a creditor or insurance company selects a random sample of its customers, or a sample of similar customers, and analyzes it statistically to identify characteristics that relate to risk. Each of the characteristics then is assigned a weight based on how strong a predictor it is of who would be a good risk. Each company may use its own scoring model, different scoring models for different types of credit or insurance, or a generic model developed by a scoring company.

Under the Equal Credit Opportunity Act (ECOA), a creditor’s scoring system may not use certain characteristics — for example, race, sex, marital status, national origin, or religion — as factors. The law allows creditors to use age in properly designed scoring systems. But any credit scoring system that includes age must give equal treatment to elderly applicants.