Pridit Is a Useful Technique for Detecting Consumer Fraud When No Training Sample Is Available

Marketing researchers and managers often face situations with incomplete information for decision-making. For example, when information needed for classification into strategic customer groups is lacking because of a disclosure social desirability bias. Consumers misbehaving through fraud are unlikely to self-disclose those actions. This is an increasing global problem for services and retailers. Detection of consumers misbehaving can be methodologically more difficult than studying other customer behaviors, since there may be no known observable dependent variable from surveys or observation for training standard statistical models.