Arikian, Aimee Jo2011-11-142011-11-142011-10https://hdl.handle.net/11299/117812University of Minnesota Ph.D. dissertation. October 2011. Major: Educational Psychology. Advisor: Thomas J. Hummel, Ph.D., 1 computer file (PDF); ix, 131 pages, appendix A.Attrition from the treatment care pathway is common in clients with eating disorder diagnoses. The purpose of this study was to test which variables are predictive of individuals who attend intake evaluation and return for treatment and those who do not start treatment. Participants were 462 clients from a community based eating disorders specialty treatment clinic whose archived records were used for analyses. Both client characteristics and situational variables were studied. Logistic regression and probability modeling statistics demonstrated that 10 independent variables increased the probability of non-treatment start. Four factors in particular: higher difference between current and ideal weight, lower body mass index, shorter distance from clinic, and absence of psychosocial and environmental problems demonstrated noteworthy contributions to the probability of non-treatment start. Findings were discussed within the context of the tertiary prevention of chronic eating disorders. Future suggestions were geared toward providing more appropriate services for clients who typically would not start treatment following initial evaluation. It was also suggested that the probability model used in analyses be transferred into everyday use within the clinic setting.en-USEating disordersTreatment startThesis or Dissertation