Optimal static and dynamic training schedules: State models of skill acquisition (Q1916545)

Training research has a long history in engineering psychology. This research has focussed on three related questions: (1) What behaviors need to be trained? (2) How should the behaviors be trained? (3) How should the training be evaluated? A primary objective of many training programs is to maximize the proficiency of the individuals involved in the limited time available for training. Meeting this objective requires answers to each of the above three questions. In particular, the tasks to be trained must be selected, the methods to train each task must be identified (one method might be appropriate at one level of proficiency, another method at another level), and the tests for evaluating the end results of training must be chosen. We will describe a framework for conceptualizing the optimization of training schedules. We want to begin by summarizing various of the different theories of skill acquisition that have been proposed over the years. The literature is voluminous and necessarily requires us to be selective. As such, we want to focus on what we think are two of the main foci of research in the area of skill acquisition, one focus starting with the mathematical learning theories proposed in the 1950s and 1960s, the other focus beginning with research on attention (and, in particular, automaticity) first extensively discussed in the 1970s. The learning models can be used to predict the rise in proficiency with increases in training. The attention models can be used to predict which skills will benefit most from training. The learning and attention models were developed in different contexts. Below, we show that the learning models can be applied not only in the rather limited context in which they were developed, but also in the richer context that the research on attention models suggests bearing the most fruit. Furthermore, these quantitative learning models are sufficiently general to characterize a range of training, from tasks requiring simple responses to tasks demanding complex procedures.