Alternate Methods of Estimating the Dollar Value of Performance

Abstract

The standard deviation of performance quality measured in dollars, SD$, is critical to calculating the utility of personnel decisions. A popular technique for obtaining SD$ calls for supervisor estimates of the dollar value of performance at different levels. In many cases supervisors can base their estimates on the cost of contracting out the various levels of performance. Estimation problems can arise, however, in contexts where contracting out is not possible, such as in government organizations without private industry counterparts. Estimation problems may also exist where individual salary is only a small percentage of the value of the performance to the organization or of the equipment operated. This paper presents two strategies for estimating the value of performance and for determining SD$ by considering the changes in the numbers and performance levels of system units which lead to increased aggregate performance. One hundred U.S. Army tank commanders provided data about their jobs for these two strategies as well as for the supervisor estimation and salary percentage strategies. The new strategies appear to provide more appropriate and acceptable values of SD$ for those complex, expensive systems where dollar values of performance are less easily estimated.

Performance Utility and Optimal Job Assignment

“…The most common metric is dollar value (e.g., Brogden, 1959;Hunter and Schmidt, 1982), although metrics other than dollar value have also been used (Eaton, Wing, and Mitchell, 1985). In general, the results of this research indicate that substantial productivity gains can result from the use of valid selection procedures.…”

Measuring Performance Improvement

“…Although utility approaches have been criticized on a number of counts (see, for example, Austin et al, 1991;Burke & Pearlman, 1988;Eaton, Wing, & Mitchell, 1985;Guion & Gibson, 1988), they do provide a foundation for establishing the value of performance improvement interventions in bottomline terminology. It is therefore curious that so little has been done with these approaches in performance improvement research.…”

The Utility of Employment Testing for Clerical/Administrative Trades in the Canadian Military

“…Schmidt and Hunter (1983) developed this "40 per cent rule" as an approximate index relating SD, estimates to worker wagedsdaries. Although the "40 per cent rule" is an approximation, it is generally a conservative estimate of SD, (Burke and Frederick, 1984; Eaton, Wing, and Mitchell, 1985).…”