Emotional disturbance

The study used program evaluation data from 21 agencies that provided supported employment services to individuals with mental illness using the Indiana supported employment model which was based on the Individual Placement and Support Model. Different samples were used for different analyses. An intake form provided information on demographics. The Global Assessment of Functioning scale provided information about a person's current skills. Other information was obtained from monthly records about staff services that an agency had to keep for reimbursement. Because the study related to obtaining work, only services related to this were measured. The study did not look at services that were provided after placement. Notably part of the specific services included an evaluation of a person's employment potential...which is NOT in alignment with best practices in supported employment services. Related to predicting work outcomes and services hours the first discriminant function looked at persons who worked and those who did not as the criterion categories. The second discriminant function looked at a different set of variables measured at intake. For individuals who worked, multiple regression analysis were conducted using the same individual characteristics to predict the total hours of SE services required prior to becoming employed. Related to supported employment services a discriminant function analysis was used to determine the characteristics of services associated with obtaining work.

All data were analyzed using SPSS for Windows (version 15.0). The primary and secondary hypotheses were tested on the whole group. For comparing groups, t-tests and ?2-tests were used to compare means and proportions respectively, unless the data were highly skewed, in which case non-parametric tests were used. Logistic and linear regression models were also fitted including potential confounding variables: (grouped) age, gender, ethnic group, educational level, symptomatology and diagnosis. All data were analyzed in groups as randomized, whether or not receiving an intervention (i.e. intention-to-treat). Data were compared for those followed up with those not followed up, overall and by each treatment arm.

Participants were assessed with psychometric scales and diagnostic interviews before treatment, directly after treatment, at eight months and at 35 months after the end of therapy.

Participants went through a two week intake assessment prior to being randomly assigned to one of the two groups. Afterwards participants were followed for a 24 month study period. Random effects logistic regression was used to assess overall differences between the two groups in month by month employment rates during this time.

The BDI, IES, State-Trait Anxiety Scale, and the WMS-R Digit Span Subtest were administered at pre and post-treatment, and again at 1-month, 3-month, and 6 month follow-ups. SUDS ratings were recorded throughout each treatment session as well as at intake, post treatment, and each of the follow-ups.

Data collection and analysis addressed five key questions: (1) What obstacles would a rural South Carolina setting posed to implementing and sustaining ACT-IVR and IPS model? (2) How do the competitive work outcomes of the ACT-IVR and IPS programs compare to a traditional program providing parallel services in a rural economy? (3) How are overall work outcomes affected by differing strategies for obtaining employment? (4) How do rates of income earned from competitive employment change over time by program? (5) How do the programs' work outcomes compare after controlling for the number of participants' service contacts?

Interviews were audio-recorded and transcribed verbatim.
The computer software program NVivo was used to organize the data for efficient analysis. Data from the key informant interviews and the documents were analyzed simultaneously. Emerging ideas were used to direct and focus later data collection and analysis according to the constant comparative approach. Guided by the principles of constructivist grounded theory, during initial coding we explored the data and considered different possible theoretical and analytical permutations. Later, coding became more focused and selective as we categorized data by grouping codes together and prioritizing some codes over others.

Psychometric assessments were conducted by trained research psychologists at four points in time: one week before the start of treatment, one month after the start, four months after the start and three months after termination.

Data was reviewed for individuals who were still active at the end of the study period. The likelihood that placements retained their jobs was assessment using life tables. Logistic regression was used too. SPSS 12.0 was used to perform all analyses.

Clinical records since the onset of illness were reviewed for each participant.