Transfer of Training and Interpolated Activity

Abstract

This chapter is concerned with two very simple experimental paradigms. In the first the experimental group learns task A and then task B and is compared with a control group that learns task B without any prior training. From this comparison we can discover whether learning task A has any effect on the learning of task B. In the second experimental paradigm task B is practiced twice, separated by an interval during which task A is practiced. Here we can discover whether the learning of task A effects the already learned task B. These both appear simple and straightforward designs. However, results from them are open to an enormous variety of interpretations. Furthermore although apparently very neutral, both experimental procedures are closely tied to learning theory and are of little use outside this framework. The most basic assumption behind these designs is that the learning of one task will indeed effect the learning of another. In the case of motor skills such related tasks have been surprisingly difficult to find. When the two tasks are related it is assumed that they have certain components in common. This notion of components is closely linked to the theory that complex activities consist of chains of stimuli and responses, and also to the theory that performance at any time will depend on a combination of underlying variables such as habit strength, reactive inhibition, conditioned inhibition, and so on. In practice it usually turns out to be very difficult to identify any of these components for a particular task. The major exceptions to this rule are tasks involving verbal learning.