A fidelity study of the Indiana Supported Employment Model for individuals with severe mental illness

Kathleen Mary Vogler, Purdue University

Abstract

A 40-item fidelity scale, the Scale for the Indiana Supported Employment Model (SISTEM), was developed to measure implementation of the Indiana model for Supported Employment. This scale was based in part on a shorter 15-item fidelity scale (IPS Fidelity scale). Staff from 24 Supported Employment (SE) programs for individuals with severe mental illness participated in telephone or on-site interviews. The SISTEM and IPS scales were both administered in a one-hour interview. Data regarding program characteristics, and program outcome were obtained for each site. Outcome data included measures of placement rates, weeks of job retention, wages, and job satisfaction. The results indicated that the SISTEM fidelity scale had good interrater reliability (Total Scale Score intraclass correlation = .97). Agreement between Supervisors and Employment Specialists was generally adequate. However, Follow-along Specialists presented a quite divergent view, suggesting that interview data source is important. Internal consistency of the scale was lower than anticipated but was approaching acceptable levels (Cronbach's alpha = .59). The SISTEM scale was correlated with the IPS subscales, suggesting that there is construct validity for SISTEM. However, there were few correlations found between program fidelity and outcome. In summary, the scale is an improvement over an existing fidelity scale. It was easier to administer and score and provided greater coverage of issues related to interagency and intraagency coordination. However, the scales low internal consistency indicates that the scale needs further development. Implications and suggestions for future use of SE fidelity scales are discussed. ^

Degree

Ph.D.

Advisors

Major Professor: Gary R. Bond, Purdue University.

Subject Area

Health Sciences, Mental Health|Psychology, Clinical|Sociology, Public and Social Welfare|Psychology, Psychometrics

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS