Examining workplace bullying measurement using item response theory

Purpose The purpose of this paper is to examine aspects of how workplace bullying is identified by the Negative Acts Questionnaire-Revised (NAQ-R), with a focus on the appropriateness of its response scale using item response theory (IRT). Design/methodology/approach IRT, in which the probability of a particular item response reflects an underlying latent variable, was used to examine NAQ-R responses from a sample of 1,173 Australian public servants (Study 1), and a representative UK data set ( n =3,494; Study 2). Findings Results indicated that problems with the response scale appear to be due to the inclusion of the abstract “now and then” option amongst concrete time options (“never”, “now and then”, “monthly”, “weekly” and “daily”). These results were replicated in Study 2, providing evidence of the robustness of the findings, and suggesting the observations are not sample specific. Research limitations/implications This work has implications for methods employed to identify and measure workplace bullying in research and organisational practice. Inconsistent endorsement of response scale options is a concern when total scores are calculated, or when a number of behaviours being experienced with a particular frequency on the response scale are taken as an indication of bullying having occurred. Originality/value Examining and refining the manner in which workplace bullying is indexed is important for identifying and managing workplace risks to health, safety and well-being.