ABSTRACT
This chapter investigates how two extrinsic variables, labeling and overtraining, interact with each other and with an intrinsic variable, age, to determine the degree of activity in the selective encoding module, as measured by the probability of reversing in the test phase of the optional shift procedure. The systematic age-related tendency to stop labeling suggests that to do so produces some cognitive strain in immature systems that is relieved by returning to a more comfortable level of activation. If labeling increases the tendency to attend selectively to the relevant dimension at criterion, the amount of possible increase produced by overtraining should be proportionately reduced. Therefore, the combined effect of overtraining and labeling should be less potentiating than the sum of the potentiation under each condition taken alone. The notion that each functional level operates in a different mode provides for formulating specific, testable hypotheses about how each module functions.
