The relation between VR services and employment outcomes of individuals with traumatic brain injury

Journal of Rehabilitation, July-Sept, 2003 by David Gamble, Corey L. Moore

A factorial ANOVA was conducted with weekly earnings for competitively employed clients as the dependent variable and the six VR services as factors. Because certain combinations of services were received by none of the participants, only 29 of the 64 cells in the six-way ANOVA were filled. Type IV sum of squares was specified to fit the model as this is the appropriate option for situations in which there are empty cells (Shannon & Davenport, 2001). The model accounted for 16% of the variance in weekly earnings, F(28, 492) = 3.32, g < .001. A significant main effect emerged for work adjustment services. Weekly earnings were significantly less for consumers who received work adjustment services (M = $196.52) than for their counterparts who did not receive such services (M = $282.83), F(I, 492) = 6.88, p < .01.

More important, however, was the significant three-factor interaction noted for assessment, college, and job placement services on mean weekly earnings for competitively employed clients, F(1,492) = 7.05, p < .009. The finding for the interaction of college with assessment when job placement was not provided is different from the finding for the association between college and assessment when job placement was provided (see Figures 1 and 2). Simple effect results in Figure I indicate that when job placement services were not provided, mean weekly earnings were dependent upon a combination of assessment and college services. Specifically, consumers who did not receive either service (i.e., assessment or college) had higher earnings (M = $308.40, 95% CI = $233.53-$383.27) than consumers who received college services alone (M = $255.60, 95% CI = $93.61-$417.59). Results also indicate that consumers who were provided with college and assessment had higher earnings (M = $276.58, 95% CI = $171.27-$381.90) than consumers who received assessment alone (M = $206.77, 95% CI = $167.50-$246.04). In short, when job placement services were not provided, the effect of college training on mean weekly earnings was different at each level of assessment. Results from Figure 2 indicate that when job placement services were provided, consumers who received college training had higher mean weekly earnings than consumers who did not receive college training for each level of assessment.

[FIGURE 1-2 OMITTED]

Table 3 displays summary statistics on weekly earnings for competitively employed clients by each level of the six factors. At $274, median weekly earnings were highest for clients who received college services. In contrast, median weekly earnings were lowest ($173) for clients who received work adjustment services.

Discussion

College, counseling, and job placement were significantly and positively related to closure status. With an odds ratio of 20.77, the job placement variable contributed most to the prediction of competitive employment for consumers with TBI. Several other studies also identified job placement as a significant predictor of successful closure for con sumers with other types of disabilities (Bolton, Bellini, & Brookings, 2000; Moore, 2001a; Moore et al., 2000; Moore et al., 2002b; Wheaton, Wilson, & Brown, 1996).