Thesis Title

Predictability of Staff Risk in a Mental Health Agency

Date of Graduation

Spring 2006


Master of Science in Counseling


Counseling, Leadership, and Special Education

Committee Chair

Joseph Hulgus


Persons entering the mental health field are often unaware of how dangerous their work can be. Oftentimes they are sent into undesirable environments and are frequently sent alone. Workers are placed in vulnerable situations with clients who have lost power and control over their lives. The workers may become the symbol of this disempowerment for the client or the agent of a care system the client feels is not meeting their needs adequately. Thus, they may become a target for misplaced or displaced frustrations. The purpose of this study was to predict employee safety within a community mental health setting. Specifically, predictors for both experienced risk (hereafter referred to as real risk) and perceived risk were sought. De-identified archival data was retrieved from a local Midwestern mental health agency. A census sample of 500 employees from 17 locations were surveyed, resulting in 267 completed surveys utilized in this study. Multinomial logistic regression analysis and standardized canonical discriminant function analysis were utilized to develop prediction equations; cross-tabulations and other qualitative analysis were utilized to better describe results obtained. Results indicated that there are significant predictors of employee risk. These results suggest that workers who have experienced harm were more likely to have previously received harassing phone calls from clients, to have been victims of property damage, or have had property stolen, felt unsafe not due to a client and were community based Implications for current practice and future research are discussed.


violence, workplace, predictability, risk, mental health, employee

Subject Categories



© Aaron B. Rimel