Use of the Copenhagen Burnout Inventory with Social Workers: A Confirmatory Factor Analysis

Background & Purpose: Burnout among social workers has the potential to lead to serious consequences for clients, social workers, and agencies. Burnout is associated with mental and physical health problems among social workers, a deterioration in services provided to clients, and increased staff turnover for agencies. The Copenhagen Burnout Inventory (CBI) has been used in other professions to identify burnout. However, little research has been done to investigate the use of this scale among United States-based social workers. The objectives of this study are (1) to evaluate the validity of the CBI with a multi-state sample of social workers in the United States; (2) to examine the discriminant validity between scores on two subscales; and (3) to examine the use of the CBI with social workers in varying organizational roles.

Methods: A convenience sample of practicing social workers in the United States was obtained via four social media sites (N = 1774). Participants completed an anonymous survey which included the Copenhagen Burnout Inventory and demographics such as organizational role. Confirmatory factor analysis was used to determine the factorial validity of scores obtained from the CBI with social workers. Given that large sample sizes may lead to excess power for confirmatory factor analyses, cross validation was applied.

Results: The hypothesized three-factor model achieved adequate fit with the data (CFI = .908, RMSEA = .093, RMSEA CI [.088, .097], and TLI = .894) and was a better fit than other models tested. Analyses supported measurement invariance for direct service status as well as supervisor status. Post-hoc analysis was performed in an exploratory effort to improve the model. Correlations between error terms were then considered by first reviewing the conceptual and theoretical development of the CBI and then, the modification indices to re-specify the model. this resulted in adding four error correlations which improved the model fit significantly (CFI = .956, RMSEA = .065, RMSEA CI [.060, .070], and TLI = .948).

Conclusions & Implications: Considering the expansion of social workers’ roles in integrated health settings, businesses, and nonprofits, assessing and managing stress levels of employees can improve outcomes for all stakeholders. The CBI is an effective tool to measure burnout among social workers. Given its public domain status along with demonstrated reliability and factorial validity among a national sample of practicing social workers, researchers should consider it as an instrument in future studies of burnout among social workers. The CBI could also be useful within human service organizations to better understand the factors associated with burnout, successfully manage challenges, and provide solutions to social workers. Further research should examine the use of the CBI among a racially diverse sample of social workers as well as within specific agency types. The longitudinal invariance of the CBI should also be assessed so that the CBI could considered for longitudinal study designs.