The effect of two training sets on transposition in adults

Abstract

Transposition in the intermediate size problem as a function of training on two sets with the same relationally defined element reinforced was explored with 96 adults. The data indicated that the relationship between the dual training sets was critical in determining whether or not relational transfer occurred in test. ProblemWhile young children transpose with test sets one step removed from the training set and show first absolute and then random response as the test inc reases in distance from the training set (Zeiler, 1963) and somewhat older children display relational transfer at all distances (Kuenne, 1946), a recent study with adults (Zeiler, 1964) revealed quite different effects.

SOURCES OF STIMULUS CONTROL DURING MULTIPLE DISCRIMINATIONS1

“…An additional probe with a new set produced responses to the stimulus closest in size to the last S+. Zeiler and Paalberg (1964) found that two sets distinctly different from each other caused adult humans to respond to the middle-sized member of a new set. Humans generalized control by the middlesize relation to new situations, whereas chimpanzees either learned to respond to the middle-sized member only of sets in which such responses had been reinforced, or, like pigeons, were controlled by absolute stimuli.…”