Research That Matters (January 17 - 20, 2008)

Background: Workers in child welfare, social work, and mental health settings are subject to psychosocial stressors related to their work (Mor Barak et al, 2001). Supervision plays a vital role in buffering the stressors' effects and in promoting worker effectiveness (Austin & Hopkins, 2004). In the short-term, effective supervision facilitates workers' capacity to perform their jobs. In the long-term, supervision relates to service effectiveness by enhancing workers' knowledge, skills, and experience (Kadushin & Harkness, 2002) and, can, ultimately, improve client outcomes (Glisson & Hemmelgarn, 1998; Poertner, 2006). Study Goals and Relevance to Conference Theme “Research That Matters”: This study's three goals were: (1) to assemble the disparate research on the impact of supervision vis-ŕ-vis worker outcomes providing a state-of-the-art review; (2) to statistically quantify the relationship between supervision categories and worker outcomes using meta-analysis; and, (3) to highlight significant findings and to provide evidence-based policy and practice recommendations. All three goals are relevant to the conference theme because the aggregate results from different studies can facilitate the implementation of evidence-based policies and training programs for supervisors and administrators. Method: This study presents a state-of-the-art review and meta-analysis of research on outcomes of supervision. Our systematic literature search resulted in 32 qualified articles published in academic journals between 1990 and 2007 with a combined sample size of 11,937 workers in child welfare, social work, and mental health settings. Utilizing an interdisciplinary theoretical framework we created a typology of three supervisory dimensions: (a) emotional/social support, (b) task-assistance, and (c) quality of interpersonal interaction. We then created two categories of beneficial and non-beneficial worker outcomes. The meta-analytic procedures used were based on techniques from Rosenthal (1991) and Peterson and Brown (2005) to calculate applicable effect sizes. Results: We found that all supervisory dimensions (emotional/social support, task-assistance, and interpersonal interaction) were positively and significantly related to beneficial worker outcomes with moderate level effect sizes. Specifically, task-assistance (i.e., supervisory guidance pertaining to job-related tasks) had the largest positive effect size on beneficial outcomes (r=.38, k=6, p<.001) as compared to both support (r=.32; k=12; p<.001) and interpersonal interaction (r=.32; k=8; p<.001). We also found that support and interpersonal interaction were negatively and significantly related to non-beneficial outcomes with moderate effect sizes (support: r=-.28; k=6, p<.001; interpersonal interaction: r=-.30; k=5; p=.007). Note: no studies assessed relationships between task-assistance and non-beneficial outcomes. Conclusions/Implications: These findings underscore the importance of effective supervision in fostering beneficial worker outcomes and in limiting non-beneficial outcomes. Uniquely, this study found that supervisor's task-assistance had the strongest positive relationship to beneficial worker outcomes and that the interpersonal interaction dimension had the strongest negative relationship to non-beneficial outcomes. Based on these findings, leaders should devote attention to providing supervisors with tools to support and interact well with their supervisees as well as skills to provide task-related guidance aimed at assisting workers in effectively providing client services. We draw suggestions for policy and practice, particularly for supervisory training, and for future research.