Events, instruments, and reporting errors

Social scientists frequently lack methods for assessing the accuracy of sur vey responses about events and behaviors. And often they do not have a gene ra framework for understanding or systematically analyzing the characterist ics of experience, if any, that influence errors in reports about those eve nts. In this validation study, structural features or characteristics of ex changing child support payments are examined to predict errors in reports a bout those events. The analysis compares court records about child support payments with answers from two telephone surveys that use parallel samples of divorced parents. Results show that indicators for the complexity, clari ty, and affective intensity of the events predict reporting errors.