Exploring Patterns of Employee Psychosocial Outcomes Among Child Protection Workers

Purpose: Child protection workers continue to suffer from increased levels of job stress and burnout, often resulting in an intention to leave.  Although the literature investigating these psychosocial employee outcomes is extensive, much of the literature is limited in that it has failed to determine if variability exists across these constructs in a way that can be systematically categorized.  The purpose of the current exploratory study was to conduct a cluster analysis to determine if subtypes of job stress, burnout and intent to leave exist among public child protection workers.  In computing a cluster analysis, we may unearth possible associations and group structures in the data not previously known that will help us to better understand workers’ needs. 

Method: This study employed a cross-sectional research design and utilized a statewide purposive sample of respondents (N = 199) from a public child protection organization in a New England state.  Information was collected regarding demographic characteristics (e.g., age and years working at DCFS), perceptions of employment-based social capital dimensions (i.e., trust, organizational commitment, supervisory support, coworker support, fairness, communication, and influence), and psychosocial employee outcomes (i.e., job stress, burnout, intent to leave).  A hierarchical cluster analysis using Ward’s method was computed, which yielded three different clusters.  Since three clusters were identified, Analysis of Variance (ANOVA) was used to determine if significant group differences emerged, while Scheffé post hoctests were examined to determine which groups differed.

Results: Cluster 3 was the largest (n = 76), followed by cluster 1 (n = 74). Cluster 2 was the smallest (n= 49).  Cluster 1 was the youngest mean age and reported the highest levels of job stress, burnout, and intent to leave. Cluster 2 reported the oldest mean age, but reported the most favorable perceptions of all of the variables examined.  Cluster 3 reported similar responses to Cluster 1 in terms of length of time working for the agency, age, levels of cooperation and coworker support, but varied significantly on all of the employee outcomes and most of the dimensions of employment-based social capital.  Results of the post hoc test suggest that all three clusters significantly varied on the employee psychosocial outcomes: job stress, burnout, and intent to leave.  Additionally, all three clusters varied significantly on levels of inclusion, communication, organizational commitment, fairness, and supervisory support.  Yet, they did not vary significantly on length of time working for the agency and coworker support.  Clusters 1 and 3 did not significantly vary on age and levels of cooperation within the agency; however, both significantly varied with Cluster 2 on these variables. 

Conclusion: Our findings suggest that child protection workers represent a heterogeneous group with dissimilar psychosocial needs.  To meet those employee needs and maximize their work output, child welfare administrators should make a concerted effort to better understand the unique needs of this strained workforce.  Doing so may provide opportunities for developing more accurate group-specific prevention interventions, which may in turn lead to increased job satisfaction and retention of child protection workers.