The Gaslighting at Work Scale: Development and initial validation

Introduction: Despite the growing literature on bullying, mobbing, harassment and abuse at work, the measurement of workplace gaslighting has received scarce attention. There is an absence of valid tools to measure workplace gaslighting. Given that workplace gaslighting is an alarming issue, there is a need for valid tools to measure this construct. Aim: To develop and validate a tool to measure workplace gaslighting. Methods: First, we performed a complete and thorough literature review to create a pool of items that could measure workplace gaslighting. Second, we employed an expert panel to assess the content validity of the initial set of our items. Third, we examined the face validity of our items by performing cognitive interviews with employees, and calculating the item-level face validity index. Fourth, we performed an item analysis for the items that were produced after the initial development phase of the Gaslighting at Work Scale (GWS). Fifth, we examined the construct validity of the GWS by performing exploratory and confirmatory factor analysis. Sixth, we examined the concurrent validity of the GWS by using the Patient Health Questionnaire-4 (PHQ-4), the Quiet Quitting Scale (QQS), the Utrecht Work Engagement Scale (UWES-3), and the Single Item Burnout (SIB) measure. Finally, we examined the reliability of the GWS by calculating Cronbach’s alpha, McDonald’s Omega, Cohen’s kappa, and intraclass correlation coefficient. Results: Initially, we developed a pool of 32 items and, after expert panel review and item analysis, we deleted 15 items. Then, we conducted exploratory factor analysis, and we removed six items. Confirmatory factor analysis verified the results from the exploratory factor analysis. Thus, we found that the final version of the GWS includes 11 items. Moreover, we identified two factors; loss of trust (five items), and abuse of power (six items). Concurrent validity of the GWS was excellent since we found statistically significant correlations between GWS, PHQ-4, QQS, UWES-3 and SIB. Cronbach’s alpha and McDonald’s Omega for the GWS was 0.939 and 0.940, respectively. Cohen’s kappa for the 11 items ranged from 0.840 to 0.914. Intraclass correlation coefficient for the GWS was 0.997. Therefore, the reliability of the GWS was excellent. Conclusions: The GWS is a two-factor 11-item scale with robust psychometric properties. The GWS is a short and easy-to-use tool that measures gaslighting behaviors of supervisors in a valid way. Valid identification of harmful supervisor behaviors is crucial to promote subordinates’ mental health and work productivity.