White / Caucasian

Nine new supported employment programs were evaluated by examining supported employment fidelity ratings at baseline and every 6 months thereafter. Site reports based on the nine case studies were used to interpret the fidelity findings.

"Three main sources were used to identify studies" (p. 281). The first source was published literature reviews, the second was to review studies in the Employment Intervention Demonstration Project, and the third was to contact "principal investigators and continuous review of the published literature" (p. 281).

Data were recorded directly from published reports or calculated by hand from the information presented.

Monthly data about job tenure were collected for 82 clients upon beginning competitive employment (prospective sample) and from 60 clients who had begun competitive employment in the preceding six months (retrospective sample). At study entry, information on demographic characteristics, employment history, diagnosis, Social Security entitlements, and current employment was collected. The monthly employment update included information on employment status (employed or unemployed), job losses, job starts, job type, days worked during the past month, changes in hours worked per week, and changes in wage rate.

Four sets of variables: personal history, functional limitations, rehabilitation services and employment outcomes were measured. The Scale of Social Disadvantage, was developed to predict competitive employment for VR clients from information collected on the application form. The FAI, a behaviorally anchored, counselor rated tool was used to measure a client's functional limitations and key environmental factors that are relevant to VR service planning. Three categories of rehabilitation service variables were used. Two discrete services were job placement and personal adjustment training. Three aggregate service variables were vocational training, restoration, and maintenance. Two global measures of service intensity were time in rehabilitation and total service costs. Various service variables were excluded from the analysis like counseling, diagnosis, support and transportation. Two criteria related to employment success, competitive employment versus not working. Weekly salary for competitive employment were used. Clients closed in other work statuses (i.e.. workshop, student, homemaker etc.) were excluded from the analysis.
Hierarchical multiple regression analysis was used to assess predictive relationships between the variable representing the 3 phases of VR services and the two employment outcomes. The first set only included the SDD total score. The second set included the SSD score and six FAI factor scales. The third set consisted of the SDD score, the six FAI scales, and the seven rehabilitation service variables. The three sets of variables were used to predict the employment outcomes (competitive employment and weekly wages).

Data for the study was generated through four sources: consumer and provider questionnaires already in use by the study agencies, data collection forms designed specifically for this research project which were completed by case managers and Supported Employment staff at the study agencies, data from agency administrative records, and IDDT and supported employment (SE) fidelity scores from a state-funded training and consultation center. Prior to the beginning of data collection, approval of data collection procedures and consent forms was received from the Case Western Reserve University IRB.

The statistical consultant compared hours worked by condition across quarters over the course of the two-year study using mixed models, repeated measures ANOVAs. Analyses were performed separately for total hours of community-based work activity (transitional and competitive payment combined), and hours of community-based work activity paid exclusively by the employer (competitive payment). Because of significant skew introduced by non-working participants in each quarter, analyses were repeated using ranked data as a nonparametric check of the results. Except where corrected for multiple comparisons, alpha was set at .05 and all tests were two-tailed.

"Participants were assessed at baseline and again at follow-up 30 to 36 months later; follow-up for clients in the converted programs took place 18 months after the conversion was completed. An independent research team conducted the interviews and evaluation." (p. 352)

Participants in the two original studies who received IPS services were invited to participate in one follow-up interview in 2004.

A modified version of a semi-structured interview developed of a previous study was used. Participants were asked about of all their work activities, including competitive employment, sheltered work, and volunteer work. The pattern of work was coded by the percentage of months worked, excluding recent work history.

Most questions were closed; however, several open-ended questions were included. The analysis was developed using the grounded theory approach developed by Glaser and Strauss.

Aggregate data regarding v-e services offered to patients were obtained from reports submitted monthly state substance abuse office. This provided information on the volume and type of service activities provided throughout the study period. In addition the VRC provided individual service information on each person served. Data on individual service delivery and monthly patient v-e status provided project implementation and outcome data.

In Clinic 1; v-e status data for patients in a comparison clinic (Clinic 2; N=358) allowed assessment of the VRC's impact. Vocational-educational services increased significantly in Clinic 1 and declined in Clinic 2. Logistic regression of factors involved in staying or becoming vocationally involved (vocationally involved refers to working full- or part-time, or being enrolled in education/training) vs. staying or becoming vocationally uninvolved among patients in Clinics 1 and 2 were conducted. Logistic regression showed that pre-treatment/at admission employment heavily influenced positive v-e change and that VRC services contributed significantly to v-e change among patients not working at admission.

For the first and second hypotheses, a repeated measures of motivation to find employment and of eight job-seeking activities for the two intervention groups were compared using linear, mixed-effects models. The third and fourth hypotheses were analyzed using multinomial logistic regression analysis. For the third hypotheses a total motivation score was calculated by summing the motivation scores across the six time points of the 12-week intervention. For the fourth hypotheses, a total job seeking score was calculated by summing the number of job search activities over the 12-week intervention period.