Affiliations: University of Oregon, Eugene, OR, USA | University of Oregon, Secondary Special Education and Transition, Eugene, OR, USA
Note: [] Address for correspondence: Scott H. Yamamoto, Ph.D., University of Oregon, Eugene, OR 97403, USA. Tel.: +1 541 232 3094; E-mail: syamamo1@uoregon.edu
Abstract: This study examined self-employment within the context of U.S. vocational rehabilitation (VR) to identify significant predictors of successful self-employment case closure, how these predictors changed over time, and whether there were differences in the likelihood of successful self-employment closure across states. To answer the research questions, five fiscal years (FYs) of RSA 911 data from 2003 to 2007, constituting more than a million cases, were analyzed using a two-level Hierarchical Linear Modeling (HLM). These years were selected because they occurred between the two most recent economic recessions. Statistically significant (p < 0.001) predictors of successful self-employment closure across the FYs were: gender, ethnicity, cost of VR services, education attainment, and public supports. The only difference occurred in FY 2004, when significant-disability status was also a significant predictor. Among the significant predictors, ethnicity had the largest effect, followed by education attainment and gender. States were significantly different in their likelihood of successful self-employment closures. Analyses of additional, more recent years of RSA data using HLM with other predictors are warranted to draw more definitive conclusions and develop substantive theoretical explanations. Limitations and implications of this study for researchers and VR agencies are discussed in conclusion.
Keywords: Self-employment, vocational rehabilitation, Hierarchical Linear Modeling
