This study assessed whether tact training would establish analogies as measured by equivalence-equivalence relations. In Experiment 1, six college students were trained to tact "same" or "different" in the presence of AB and BC compounds based on component class membership (e.g., A1B1 as "same", and A1B2 as "different"), and then tested on emergent tacts (BA, CB, AC, CA) and equivalence-equivalence relations. Only one of six participants passed all tests without remedial training. In Experiment 2, six college students were trained to tact only compounds belonging to the same class as "same". Three of six participants passed all tests without remedial training. In Experiment 3, six college students were trained to tact stimuli belonging to the same class with a common name prior to exposure to relational tact training. All participants passed tests without remedial training. In Experiment 4, eight college students were trained to tact stimuli belonging to the same class with a common name. Six participants passed without remedial training, while two, who did not tact the relation of the compounds, did not. Results from these studies suggest that simple discrimination of individual components and their relation in the form of tacts is related with equivalence performance.

Traditionally, behavior analysts have studied stimulus equivalence using a matching-to-sample (MTS) preparation. Although researchers have shown the utility of MTS to yield equivalence classes, the procedure requires several prerequisite skills for a learner to accurately respond. Previous research with humans and nonhumans has shown that relational responding can be produced via compound stimulus discrimination and successive matching-to-sample (S-MTS). We conducted four experiments with college students to further evaluate the effectiveness of S-MTS in the establishment of stimulus relations. S-MTS trials consisted of the presentation of a single sample stimulus followed by one comparison in a fixed location on a computer screen. Depending upon the sample-comparison relation, participants touched (i.e., go) or did not touch (i.e., no-go) the comparison stimulus. Following training of the baseline relations (AB/BC), we assessed the emergence of symmetry, transitivity, and equivalence performances (i.e., BA/CB and AC/CA). Results support the utility of the S-MTS procedure as a possible alternative to traditional MTS. This study has direct implications for participants for whom traditional three-array MTS procedures may be challenging.

No abstract