Academic journal article The Journal of Rehabilitation

Factors Involved in Time Taken in Returning to Work after an Industrial Injury

Academic journal article The Journal of Rehabilitation

Factors Involved in Time Taken in Returning to Work after an Industrial Injury

Article excerpt

Data on a population of 51,944 injured workers in the State of Alabama were examined to determine the relationship of 13 predictor variables to a single criterion variable, the number of days taken in returning to work after an industrial injury. Although the predictor variables showed a statistically significant relationship to the criterion variable, none accounted for a meaningful proportion of variance. When all available predictor variables were included in the equation, they accounted for less than one percent of the variance in the criterion variable. Thus, it is apparent that statistical significance had been achieved only because of the size of the analysis group.

Rehabilitation is a field concerned with helping individuals lead meaningful and rewarding lives (Wright, 1980). As a result, the search for factors that impact on successful rehabilitation is ongoing and predicting rehabilitation outcomes is a commonplace activity. Typically, research takes the form of attempting to predict rehabilitated versus non-rehabilitated status (Raleeh, 1986) or an examination of one or more issues related to a specific disability group such as an examination of productive functioning two years after spinal cord injuries (Herron, 1987).

Researchers in the specific area of vocational rehabilitation have also been actively trying to develop models which would accurately predict outcomes. Typical of their research would be investigations into factors impacting on return to work after burn injuries (Helm, Walker, and Peyton, 1986), predicting earnings of clients with visual impairments (Gandy, 1986), predicting return to work after a cerebral infarction (Howard, Till, Toole, Matthews, & Truscott, 1986), predicting dropouts from a vocational rehabilitation program (Tseng, 1972; Worrall & Vandergroot, 1980), and predicting which employees with disabilities will return to work (Hester, Decelles, & Gaddis, 1986). Other related topics include such issues as the cost-benefit of vocational rehabilitation (Bellante, 1972; Worrall, 1978), and the vocational status of older workers (Growick & McMahon, 1983; Morrison & Magel, 1984; Rasch, 1979).

There are a number of reasons for the importance placed on prediction of rehabilitation outcomes. One is that there are economic benefits to vocational rehabilitation (Hester & Faimon, 1985; Jedrziewski & Morrison, 1986; LaFon, 1989; Roberts, 1989). In addition, there are social benefits to having people in the workforce and therefore it is important to identify key factors impacting on the time necessary to return from injury (Morrison & Magel, 1984). Finally, there are those who maintain that prediction is fundamental to the human service fields (Wong, Gay, & Wainwright, 1987). In the words of Bolton (1972): "It can be reasonably argued that the central professional activity in the provision of rehabilitation services is prediction" (p. 16).

There are two major types of predictions: (a) clinical prediction, or professional judgment. and (b) statistical prediction. Typically, statistical prediction makes use of multiple regression analysis (Marascuilo, 1971) and allows one to make inferences on a criterion variable based on what is known about one or more predictor variables. In contrast, clinical prediction is less objective and tends to be vague, subjective, and intuitive (Goldman, 1971).

It remains a matter of some controversy as to which type of prediction is more effective. Numerous early studies demonstrated the value of statistical prediction over clinical prediction (Goldman, 1971; Holtzman & Sells, 1954; Lewis & MacKinney, 1961; Meehl, 1956; Pierson, 1958; Rosen & Van Hom, 1961). But perhaps the classic study on clinical and statistical prediction was that conducted by Meehl (1959) who examined a number of studies comparing clinical versus statistical prediction. In half of the studies, both types of prediction were equally effective, while in the other half, statistical prediction was more accurate in forecasting behavioral outcomes. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.