Methodological aspects may be direct or indirectly responsible for the difference between
data obtained with non-human subjects and human participants when submitted to studies
on class formation. The present study investigated the effect of the use of repeated
reversals of simple discriminations on the acquisition of behavior compatible with
functional class formation with captive Cebus apella. In Experiment I, two capuchin
monkeys (Cebus apella) were given simultaneous simple discrimination training and
reversals with six pairs of visual stimuli in the context of foraging. Responses to the stimuli
(differently colored wooden boxes) assigned as S+ were reinforced with food available in
the boxes. One different kind of food for each potential stimulus class was used as
reinforcer. After repeated reversals in the contingencies between the two stimulus groups,
tests for functional class formation were carried out. Each test consisted of a contingency
reversal with four (two of each potential class) out of the six stimuli (with the remaining
two stimuli kept absent) and later reinsertion of the absent stimuli, after highly accurate
performance on the reversed discriminations. The first choice response to the reintroduced
pair of stimuli was analyzed: if this response was consistent with the reversed
contingencies, this was considered as an evidence of class formation. Multiple tests were
carried out, each time with one different stimulus pair being removed. Evidence of class
formation was found in all tests. In Experiment II, one of the subjects was given
simultaneous simple discrimination training with shifts of the stimulus positions between
sessions and class-specific reinforcer was removed: a third reinforcer was employed in
both classes. The same tests described in Experiment I were conducted but with shifts of
stimulus positions between sessions and no class-specific reinforcer. Evidence of class
formation was found, again, in all tests.
|
