How often is Employee Anger an Insider Risk I? Detecting and Measuring Negative Sentiment versus Insider Risk in Digital Communications

Abstract

This research introduced two new scales for the identification and measurement of negative sentiment and insider risk in communications in order to examine the unexplored relationship between these two constructs. The inter-rater reliability and criterion validity of the Scale of Negativity in Texts (SNIT) and the Scale of Insider Risk in Digital Communications (SIRDC) were established with a random sample of email from the Enron archive and criterion measures from established insiders, disgruntled employees, suicidal, depressed, angry, anxious, and other sampled groups. In addition, the sensitivity of the scales to changes over time as the risk of digital attack increased and transitioned to a physical attack was also examined in an actual case study. Inter-rater reliability for the SNIT was extremely high across groups while the SIRDC produced lower, but acceptable levels of agreement. Both measures also significantly distinguished the criterion groups from the overall Enron sample. The scales were then used to measure the frequency of negative sentiment and insider risk indicators in the random Enron sample and the relationship between the two constructs. While low levels of negative sentiment were found in 20% of the sample, moderate and high levels of negative sentiment were extremely rare, occurring in less than 1% of communications. Less than 4% of the sampled emails displayed indicators of insider risk on the SIRDC. Emails containing high levels of insider risk comprised less than one percent or the sample. Of the emails containing negative sentiment in the sample, only 16.3%, also displayed Journal of Digital Forensics, Security and Law, Vol. 8(1) 40 indicators of insider risk. The odds of a communication containing insider risk increased with the level of negative sentiment and only low levels of insider risk were found at low levels of negative sentiment. All of the emails found to contain insider risk indicators on the SIRDC also displayed some level of negative sentiment. The implications of these findings for insider risk detection were then examined.