Transitivity as Skinnerian problem solving controlled by self‐constructed relational stimuli
Having learners spin a one-sentence story that links stimuli together sharply boosts their later transitive relations.
01Research in Context
What this study did
Adults learned five-member equivalence classes through match-to-sample.
Half the group only named each picture. The other half made up a short story that linked all five pictures together.
Everyone then took tests for transitive relations (A-C, B-D, etc.).
What they found
Story writers derived far more transitive relations than the name-only group.
Even on the very first test, the tale group scored higher.
The stories acted like glue that held the class together.
How this fits with other research
Fields et al. (2018) predicted this. Their theory paper says adding meaningful content speeds class formation. Moustakis gives the live data.
Doran et al. (2012) also trained five-member classes and saw that learners liked transitive links best. The new twist: self-made stories create those strong links in the first place.
Lazareva (2025) warns that ‘transitivity’ tasks may not all measure the same thing. The 2018 tale effect is still positive, but future work should spell out which transitivity test was used.
Why it matters
If you run equivalence lessons, have the learner invent a quick story that weaves the stimuli together. One sentence per picture is enough. This tiny add-on can unlock more derived relations without extra drills.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Before the next match-to-sample block, ask the learner to tell a mini-story that uses all the pictures in order.
02At a glance
03Original abstract
In tests of the derivation of complex relations such as transitivity, extant cues might fail to evoke effective responding, necessitating the construction of supplemental stimuli prior to their solution. The significance of this process was investigated by within-subject manipulation of an instructional variable designed to produce different levels of construction of supplemental stimulation concerning relationships among stimulus elements of concatenated conditional discriminations. In two experimental sessions, serial training of three 5-member stimulus classes occurred, either with the instruction to simply name the component stimuli or to both name them and generate a tale serially linking the stimuli named; such constructed stimuli might be spontaneously reconstructed by precurrent acts in subsequent tests of "emergent" relations. Participants whose supplemental stimulus construction at the first session was limited to name-giving derived significantly more relations when, in training at session two, they generated tales linking stimulus elements; this same near-errorless derivation was obtained at the first session whenever relational stimuli beyond bidirectional naming were constructed. In some cases the uninstructed construction of supplementary relational stimuli occurred at the first session, to equivalent effect; such construction might constitute a typically unobserved component of the derivation of relations among stimulus elements entailed in multiple conditional discriminations.
Journal of the Experimental Analysis of Behavior, 2018 · doi:10.1002/jeab.473