Detecting data errors in organizational settings: Examining the generalizability of experimental findings

Abstract

Experiments show that giving users incentives helps them detect errors in data and that expectations about the base rate of errors in data improve error detection. However, experimental findings are often criticized as not being generalizable to organizational settings. The study reported here examines the generalizability of experimental findings on user detection of data errors. Afield interview study was conducted to examine this question. Twenty interviews were conducted with users of information systems in a variety of organizational settings. The findings of the field interview study show that strong informal incentives, perceptions about the materiality of data errors, and perceptions about the base rate of errors in data affect the detection of data errors in organizational settings.