Computational Approaches for Analyzing Tradeoffs Between Training and Aiding

Abstract

A methodological framework is presented for representing tradeoffs among alternative combinations of training and aiding for personnel in complex situations. In general, more highly trained people need less aid, and those with less training need more aid. Balancing training and aiding to accomplish the objectives of the system in a cost effective way is the concern. A wide variety of methods, tools, and models is reviewed. These approaches are evaluated in terms of their advantages and disadvantages when used to analyze training/aiding tradeoffs. The use of the proposed framework and its component methods, tools, and models is illustrated by ;-,n analysis of a realistically complex example involving the design of a head-up display for use by truck drivers in long-haul transport. Results demonstrate that the tradeoff issue can be involved in other than an ad hoc manner. Research needed in predictive models, learning processes, and intelligent systems is reviewed. Four tables and nine figures illustrate the discussion. (SLD)

Job Aiding/Training Decision Process Model

“…Thus, to a great extent, the decision making is highly proceduralized, e.g. if situation x, then employ training type y and/or aiding type z (Rouse and Johnson, 1989).…”

“…This approach requires that the relevant trade-offs be explicitly idrntified prior to the predictive analysis (Rouse and Johnson, 1989).…”

“…This approach, however, also requires epicit trade-off formulation prior to the simulation execution (Rouse and Johnson, 1989).…”

“…Technical training, for example, serves as the source of knowledge and skills essential to task performance. In other words, training "creates the potential to perform" while job aiding, in contrast, directly augments the performance (Rouse and Johnson, 1989). Job aids collectively refer to those devices with the capacity to store and retrieve the "How", "What", and "When" information pertinent to the performance of a particular task.…”

“…Typically, these types of analyses have required many person-years of effort. Often, the result has been a time-consuming and expensive effort that provided insights which were too late to be implemented in any substantial way (Rouse and Johnson, 1989). Whether for evaluating current AFS job performance, selecting among new system design alternatives, or ensuring flight-line personnel are task qualified, a methodology for efficiently producing consistent, timely, and supportable A/T decisions is a must.…”