Content area
Full text
INDUCING SENSITIVITY TO DECEPTION IN ORDER TO IMPROVE DECISION MAKING PERFORMANCE: A FIELD STUDY1
Abstract
When an organization's members depend on the data contained in computer-based systems, they become vulnerable to strategic information manipulation. That is, they become susceptible to situations where their decision-making behaviors can be influenced by others able to access and manipulate this data.
This paper describes the results of a field experiment that examines the effects of alternative interventions aimed at inducing sensitivity to the possibility of manipulated data on professionals' task-related decision behaviors: deception detection, false alarms, and task accuracy. While traditional training had no effect on detection success or the issuance of false alarms, warnings about data quality resulted in better detection success. Warnings combined with just-in-time training resulted in better detection success but at the cost of an increased number of false alarms. Higher levels of detection success increased task accuracy and the time spent solving each problem, A higher number of false alarms was associated with lower levels of task accuracy.
Keywords: Information quality, data security, data integrity, error detection, deception detection
ISRL Categories: GB04, HA0901, HB16, HC0201
The Effects of Sensitivity to Deception and Domain Experience on Deception Detection and Task Accuracy
With sustained exposure to and use of information technologies, computer users develop trust in the accuracy of the data stored within a computer system (Parasuraman 1987; Wiener 1985). Yet anywhere from 1% to 10% of data items in critical organizational databases are estimated to be inaccurate (Klein et al. 1997a). In most cases, such inaccuracies are due to errors that occur during data input or data transcription. When an organization's databases are relatively open and visible to the public, such as those for airline reservations systems, errors are regularly found and corrected (Orr 1998). On the other hand, for more-private databases, such as those used by law enforcement agencies, finding and correcting errors are generally neither simple nor regular. Typically, few if any sanctions are applied for submitting invalid data. Moreover, few procedures are in place for finding and correcting data inaccuracies (Laudon 1986). Once errors are introduced, they may remain in these databases for long periods of time.
While invalid data potentially produces devastating results, most database...