satisfactory level. Of course, for any particular medical or physical condition, this assumption could be subjected to closer scrutiny.
If there is a correlation between the selection criterion and an important outcome, then specific standards can be set based on some type of utility or cost trade-off analysis. The Department of Defense has developed cost trade-off models to set standards for education levels and cognitive aptitudes (National Research Council, 1994). We describe how these models could be adapted for use with appropriate medical and physical standards.
In summary, the process of establishing a standard for an appropriate physical or medical condition consists of several steps. The first step is to determine a valued outcome that corresponds to some aspect of military effectiveness. The second step is to establish an empirical correlation between the condition and the valued outcome. The third step is to determine a specific standard or a cutoff value in the case of continuous conditions, such as body mass index (BMI), using a cost trade-off analysis of some type.
The outcome used to evaluate an enlistment condition is usually a goal or condition related to military objectives or requirements, such as completion of training, on-the-job performance, or combat effectiveness. The evaluation of education standards, for example, uses the outcome of first-term attrition. The justification is that enlistees who leave before the end of their first term increase training costs, since more recruits must be trained to fill unit manpower requirements. Thus, an enlisted force with 20 percent attrition has to train 125 recruits to fill 100 manpower unit slots, while a force with 40 percent attrition has to train 167 recruits to fill the same unit requirement. This increased recruiting burden represents an increase in training costs of about 40 percent.
Another example is the evaluation of aptitude requirements. In this case, job performance has been the preferred outcome because it is well established that cognitive abilities are good predictors of actual job performance in virtually all military specialties (Armor and Roll, 1994). As the proportion of persons with low cognitive aptitudes increases in a unit, there is a corresponding reduction of combat effectiveness. The concern of recruiting sufficient numbers of people with good cognitive skills has been heightened by the growing technological requirements of many military occupations and activities. Unlike attrition, the assessment of military job performance is an arduous and complex undertaking.