Improving credit evaluation with a weighted application blank

Abstract

Using data from a credit application blank, how can poor credit risks be most effectively identified? 100 records of those who paid for their cars were compared with a like number of records of those who failed to do so. An item analysis was performed and weights assigned to items according to their ability to differentiate between the 2 sets of records. 24 items were chosen as working especially well. "A formula was presented which allows the profit obtaining from the use of any particular cut-off score to be estimated." (PsycInfo Database Record (c) 2025 APA, all rights reserved)