Show simple item record

dc.contributor.authorYamamoto, Scott Hiromi
dc.date.accessioned2011-09-01T16:28:24Z
dc.date.available2013-06-04T17:45:42Z
dc.date.issued2011-06
dc.identifier.urihttp://hdl.handle.net/1794/11534
dc.descriptionxv, 111 p. : ill.en_US
dc.description.abstractDespite numerous legislative and programmatic efforts, individuals with disabilities continue to experience greater difficulties gaining employment and poorer outcomes of employment than individuals without disabilities. These disparities negatively impact society. My review of the U.S. empirical research literature suggests, however, that self-employment could improve employment opportunities and outcomes for individuals with disabilities, and their success is most influenced by individual characteristics, level of supports, and accountability systems. In this dissertation study, I used a nonexperimental research design to investigate six research questions with Hierarchical Linear Modeling (HLM) and Structural Equation Modeling (SEM) statistical analyses. Extant data on more than a million clients of vocational rehabilitation (VR) agencies from the 50 states and District of Columbia for fiscal years 2003 to 2007 were obtained from the Rehabilitation Services Administration. Results of the HLM analysis indicated that among the significant (<italic>p</italic><.001) predictors of self-employment closure across the fiscal years, ethnicity had the strongest effect. The initial SEM analysis produced an inadmissible solution; the respecified model of individual characteristics, level of supports, and accountability systems produced a reasonable model fit in each fiscal year. The model invariance testing across the four U.S. Census Regions indicated a reasonable fit in each fiscal year when model parameters were freely estimated for each region, but very poor fit and significant differences were indicated when some parameters were fixed to be equal across the regions. The major limitations of this dissertation study are model misspecification in HLM and SEM and the small number of RSA fiscal years that were analyzed; causal inferences cannot be made. The primary implication of this study for researchers is using the results of the statistical analyses to develop and test theories about self-employment of individuals with disabilities through VR. The primary implication for VR is using the results to make decisions about services and agency policies. Recommendations for further research include (a) using Laplace estimation in HLM, (b) analyzing other HLM random effects and predictors, (c) testing a SEM model of different indicators and factor structure with Bayesian estimation, and (d) conducting empirical longitudinal studies given the complex developmental processes of self-employment.en_US
dc.description.sponsorshipCommittee in charge: Richard Albin Chair; Deanne Unruh, Member; Deborah Olson, Member; Lauren Lindstrom, Member; Patricia Gwartney, Outside Memberen_US
dc.language.isoen_USen_US
dc.publisherUniversity of Oregonen_US
dc.relation.ispartofseriesUniversity of Oregon theses, Dept. of Special Education and Clinical Sciences, Ph. D., 2011;
dc.subjectSpecial educationen_US
dc.subjectHierarchical linear modelingen_US
dc.subjectIndividuals with disabilitiesen_US
dc.subjectSelf-employmenten_US
dc.subjectStructural equation modelingen_US
dc.subjectVocational rehabilitationen_US
dc.subjectPeople with disabilities -- Employment
dc.titleIndividuals with Disabilities in Self-Employment through Vocational Rehabilitation Agencies across the United Statesen_US
dc.typeThesisen_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record