Understanding the Impact of Coworker Conflict in an Unionized U.S. Public Sector Workforce and Comparative Psychometric Review of the Negative Act Questionnaire

Abstract

Background: Workplace violence is an enormous problem worldwide. A widespread and costly segment of this problem includes those actions which are perpetrated by a current or former fellow employee, so-called Type III violence, defined as negative acts and bullying in this study. Purpose: The study's primary purpose is to describe the individual and organizational impact of coworker conflict in a large public sector workforce. A secondary purpose was to conduct a comparative psychometric review of the 21-item, 6-item, and 4-item Negative Acts Questionnaire-Revised (NAQ-R). Methods: A cross-sectional web-based survey of coworker conflict was conducted in 2009 among 12,546 state government workers (response rate of 72%,) and a pilot sample of 420 workers (94% response rate). Multinomial logistic regression was used to describe the individual and the organizational impact of coworker conflict as well as demographic and work organization factors. Cronbach's alpha and separation reliability were used to assess the internal consistency of the different versions of the NAQ-R. Validity was assessed using 1 parameter Item Response Theory Model (IRT), and logistic regression with theoretically related constructs (individual impact and supportiveness of the organizational climate). Results: The Multinomial logistic regression model results suggest high individual impact for regular and severe coworker conflict, female gender, job tenure of 2-20 years, working in a low supportive organizational climate and membership in a support/administrative bargaining unit. While regular and severe coworker conflict, females, age 45-55, providing healthcare and field service, with a low supportive organizational climate was associated with high organizational impact. The NAQ-R different items version (21,6, and 4) demonstrated high internal consistency (Cronbach's alpha 0.93-0.7, and separation reliability 0.96-0.74) and validity. The comparative psychometric was supportive for utilizing less number of items guided by IRT and PAR. Conclusion: Co-worker conflict is associated with both individual and organizational level impacts. The identification of worker and organizational risk factors for impact should lead to programs and interventions to reduce the problem. Longitudinal studies are needed to demonstrate the reliability and validity of the variation of the NAQ-R number of items over time to provide more supportive evidence for their usefulness.