Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Purpose: Despite substantial contribution of previous research on social worker turnover, few extant studies have attempted to integrate developmental process of burnout and job conditions buffering this process as a single conceptual model of turnover intention among social workers. The present study attempts to develop and empirically test a comprehensive conceptual model explaining 1) the specific relationships between job stressors (i.e., role conflict, role ambiguity, work overloads, lack of skill utilization) and different burnout components (i.e., emotional exhaustion, depersonalization, diminished personal accomplishment), 2) the relationships between burnout components and turnover intention, and 3) the buffering effects of three types of social support (i.e., supervisory support, co-worker support, top managerial support) on these relationships. The conceptual model linking the key constructs in the available data is based on the theoretical frameworks proposed by Cordes and Dougherty (1993) and Koesk and Koeske (1993) and other related empirical findings.

Methods: For the study's cross-sectional study design, 1,500 registered social workers were randomly selected from the total population of California state-registered social workers. From the original sample of 1,500, 398 questionnaires were eligible, yielding a survey completion rate of 27 percent. The proposed conceptual model was empirically tested using structural equation modeling (SEM) techniques. As suggested by Anderson and Gerbing (1988), a confirmatory factor analysis (CFA) was performed to test a measurement model. Once the measurement model had been estimated and a convergent and proper solution arrived at, a structural model based on the conceptual model was assessed to test how accurately this conceptual model fit the present data. In addition, multi-group analyses were performed to test hypothesized moderating effects of social support.

Results: The structural model based on the conceptual model compellingly fit to the social worker data (CFI=0.95, RMSEA=0.054). All the hypothesized paths were significant. Emotional exhaustion was the single most important factor in explaining turnover intention and work overload was a key job stressor explaining emotional exhaustion among social works. The conceptual model overall explained 26 percent of variances of turnover intention. Regarding the hypothesized moderating effects of social support in the model, supervisory support moderated the relationship between work overload and emotional exhaustion (chi-square change(1)=4.74, p<.05), while top managerial support moderated the relationship between emotional exhaustion and turnover intention (chi-square change(1)=4.55, p<.05) and the relationship between diminished personal accomplishment and turnover intention (chi-square change(1)=4.11, p<.05).

Implications: This conceptual model empirically supports proposed specific relationships between job stressors and different burnout components among social workers. It suggests that burnout prevention programs need to focus on different set of job stressors depending on targeting specific burnout component. The findings about social support suggest that effective supervision is important for buffering burnout development, while top managerial practice is important for retaining workers experiencing burnout at work.