Affiliations: [a] University of Oregon, College of Education, Eugene, OR, USA
| [b] Department of Special Education and Clinical Sciences, University of Oregon, Eugene, OR, USA
Correspondence:
[*]
Address for correspondence: Scott H. Yamamoto, Ph.D., Campus Associate, 1215 University of Oregon, College of Education, Eugene, OR 97403-1215, USA. E-mail: [email protected].
Abstract: BACKGROUND:Many people with disabilities gain and retain paid employment through vocational rehabilitation (VR) services. Some empirical research has indicated that self-employment can be a viable employment option for people with disabilities. OBJECTIVE:We examined factors of successful self-employment of people with disabilities who had received vocational rehabilitation (VR) services across a recent five-year span. METHODS:We analyzed extant federal VR data from 2008 to 2012 from the Rehabilitation Services Administration using a special type of Structural Equation Modeling (SEM) technique known as the Multiple Indicators Multiple Causes (MIMIC) modeling to conduct validation and invariance testing. RESULTS:The national model produced reasonable fit, but there was some regional misfit in the invariance models. These results were similar to those from another similar study that used MIMIC modeling on earlier federal VR data from the RSA. CONCLUSIONS:The generalizability of the MIMIC models in this study is limited; no causal inferences were or should be drawn. Using these results, further research should examine modification indices and test alternative MIMIC models of VR self-employment. Implications of this study for VR and policymakers and recommendations include changes to service provision for clients and greater flexibility in supporting self-employment based on individualization.
Keywords: Self-employment, people with disabilities, vocational rehabilitation, structural equation modeling, multiple indicators multiple causes model, validation and invariance testing
