Professional Documents
Culture Documents
Formed The Same Classes
Formed The Same Classes
Formed The Same Classes
was needed to make a decision about the class membership of the sample and
comparison, as with the TSP-2R-RW trials. In contrast, however, working memory was
not needed to make decisions about the response to be produced since the responses
occurred in the presence of the comparison stimuli in these TSP-2R trials. Thus, less
working memory was needed to form classes when trials were presented in the TSP-2R
format than the TSP-2R-RW format. The results of these two experiments, then, suggest
that likelihood of class formation is directly related to the amount of working memory
required to respond correctly during training and testing. Additional research is needed
to assess the validity of this interpretation of the relation between working memory and
The 4-MIX test contained probes of the three-and four-member classes, and mastery
levels of responding were maintained by all probes. Thus, class expansion did not
disrupt the intactness of the previously formed three-member classes. When the four-
D-based probes were maintained for only four of the participants. Thus, the emergence
of the four-member relations did not require the presence of the three-member relations.
By contrast, the D-based probes presented alone did not maintain class-indicative
responding in their initial administration for the seven remaining participants. For them,
the maintenance of the four-member classes in the 4-MIX test depended on the presence
of the three-member probes. Since these decrements returned to mastery with test
Historically, class expansion has been assessed in the context of probes of previously
formed classes. The present experiment suggests that their inclusion in prior
experiments may have supported the observed class expansion in the prior experiments.
If so, how many and which types of probes served such a supporting role? This question
can be answered with additional research, which could inform a better theoretical
As noted in the introduction, sorting tests have been used to document the formation of
equivalence classes. These tests take much less time to complete than the typical MTS
test, and produce results that are consistent with the outcomes of MTS tests. In all prior
experiments, however, the only stimuli used in a sorting test have been the members of
classes were being probed, sorting the stimuli into three piles could indicate the
emergence of three classes or the emergence of two classes with the remainder not
functioning as members of either of the two classes, that is, a “discard” pile of cards that
presented as a class.
The present experiment avoided that interpretive ambiguity by including stimuli that
were not members of any of the experimenterdefined classes. All six participants who
were tested placed the cards into three stacks, one with stimuli from Class 1, one with
stimuli from Class 2 and a third with the novel stimuli that were not members of either
class. Thus, the sorting test modified in this manner documented the formation and
Sidman (1987) argued that two or more comparisons should be used to form
being generated by reject relations rather than select relations. R. R. Saunders et al.
(2005), however, demonstrated that the likelihood of class formation did not vary with
the number of comparisons presented during the training or testing trials. In the present
experiment and in Fields et al. (2009), participants formed equivalence classes and then
expanded class size in the context of trials that contained single comparisons per trial. In
addition, the