Association for Behavior Analysis International

Recent stimulus equivalence research has utilized simple-discrimination training procedures with three-term contingencies. However, these procedures have not been effective from the outset for some participants. Crossmodal identity training with compound-consequence images has been shown to facilitate early probe performances (Sheehan, 2018). The current study examined which stimulus-control topographies are important for success on initial probe performances. Specifically, the impact of the crossmodal component of identity training and whether it helps establish a correspondence between the computer-presented images and reinforcer tokens was evaluated. Seven typically developing children (ages 3-4) received one of three variations of identity training (i.e., crossmodal identity training, computer-based identity training, or tabletop identity training) prior to simple discrimination training with compound discriminative stimuli and class-specific reinforcers. Data to date suggest signs of emergence for one participant, who received crossmodal identity training, although these data do not indicate complete emergence. These data raise the question of which variables are responsible for the lack of emergent performances in the current study.

Selection-based instruction (i.e., multiple-choice format) might not be enough for the emergence of topographical responses such as writing. This study examined the emergence of topographical responses after participants read notecards, which presented descriptions of the trained relations, and answered selection-based questions, and the formation of equivalence classes. Stimuli involved 4 four-member classes [textual label (A), textual definition (B), graph (C), vignette (D)] of single-subject designs. Fifteen undergraduates were submitted to a pretest-training-posttest design. Pre and posttests presented AB, AC, AD, BA, CA, DA, BC, BD, CB, DB (selection-based), CA tact and AB and BA intraverbal (topographical) relations. AB, AC, AD relations were trained and BA, CA, DA, BC, CB, BD, DB emergent relations were tests. Ten participants increased pre- to posttests scores on the selection-based and topographical tests, indicating that the training comprised of notecards and selection-based questions taught participants to select the correct answers on the selection-based questions. Moreover, the training taught some of the participants to write the names and definitions of the single-case designs, even without explicit training. Five participants demonstrated the emergence of equivalence classes. This study showed the emergence of topographical responses after participants were exposed to a training comprised of notecards and selection-based questions.

Delayed matching to sample (DMTS) improves equivalence classes formation when compared to simultaneous matching (SMTS). However, this differential outcome was not observed for classes established by SMTS when the reorganization of classes was tested after reversal of conditional discriminations using both procedures. We aimed to complete this investigation evaluating the effects of conditional discrimination reversals, carried out by DMTS or SMTS, over the reorganization of classes initially established by DMTS. Undergraduate students learned AB, AC and AD conditional discriminations through DMTS (2s). After the emergence of equivalence relations between B, C and D stimuli, participants were divided into two groups: DMTS Group continued with DMTS (2s) and SMTS Group was exposed to SMTS. A reversal training modified AD relations (A1D2, A2D1) and reorganization tests verified whether baseline equivalence relations were altered. Out of 20 participants, 15 reorganized the classes. No differences were observed when accuracy on reorganization tests was compared between the groups. However, choice response speed was significantly lower for SMTS group in reorganization tests when compared to DMTS or to both groups in equivalence tests. The results suggest that additional measures of classes reorganization (e.g., response speed) should be useful to investigate the modification of equivalence relations.

In pursuit of a behavior analytic account of symbolic meaning, researchers often investigate conditions that influence the probability that derived stimulus relations will emerge. One common preparation for such investigations is premised on a well-replicated experimental outcome termed “The Meaningful PIC Effect”: In potential stimulus classes of otherwise non-representational (meaningless) forms, ABCDE, a higher percentage (yield) of participants demonstrate equivalence relations in conditions where the C stimuli are pictures than in conditions where C stimuli are non-representational. Further, within this paradigm, pretraining a stimulus function to non-representational C stimuli through successive-discrimination procedures enhances equivalence outcomes. However, such pretraining typically has been conducted with keyboard responses involving letters of the alphabet, themselves meaningful stimuli - a potential confound. The present study asked whether pretraining successive discrimination of non-representational C stimuli using a simple one-button go/no-go procedure would similarly enhance subsequent equivalence yield with ABCDE non-representational stimuli. College-student participants were randomly assigned to a keyboard-response successive-discrimination condition, a one-button go/no-go successive-discrimination condition, or two control conditions. Across 29 participants to date, both variations of successive-discrimination pretraining enhanced equivalence yields, suggesting another way in which meaningfulness emerges from acquired stimulus function. Additional data collection is ongoing to increase group sizes and statistical power.

The comprehension of socially agreed codes (arbitrarily related to other events) plays important roles in everyday tasks. The present study investigated the acquisition of arbitrary relations in typically developing children, using two procedures for fading the sample stimulus in a matching-to-sample procedure (MTS). In Static Fading, gradual transformations in the sample stimulus occurred in a sequence of trials; in Dynamic Fading, transformations occurred within each trial. Twelve children aged between three and six years participated. We used four sets of three pairs of stimuli, each pair consisting of a picture and a letter. The computerized procedure began by teaching letter-letter identity matching (IDMTS), which gradually morphed into arbitrary visual-visual picture-letter matching. Each participant was exposed to the four stimulus sets, which alternated according to the fading procedure, and the order of alternation was balanced between participants (DSDS x SDSD). Six children learned all relations, two learned the relations for two stimulus sets, one child learned one stimulus set and three children did show any learning. The amount of training and the number of learned stimulus sets relations varied with age: older children tended to learn more relations with less exposure to the procedure, regardless of the type of fading procedure.

The N400 is an event-related brain potential sensitive to the semantic relations in naturally occurring language. To illustrate, when words are unrelated (e.g., grease-bread) the N400 will be more negative than N400 evoked by related words (e.g., butter-bread). The peak amplitudes can be used to differentiate the types of experimentally-defined stimulus relations established by Matching-to-Sample (MTS) procedures. That is, the N400 will be more negative for the “non-equivalent” than to the “equivalent” stimulus-stimulus relations. Research reported here concerns the impact of explicit training of name-object relations, in the N400 waveforms. In this regard, a between groups design was used, in which both groups were exposed to the same experimental procedures, except for the stimuli used, so that Group 1 used only abstract figures, and Group 2 used both pseudowords and abstract figures, during MTS training. The waverforms obtained from participants shows a clear N400 effect for Group 2, but not for Group 1. These results are discussed considering the possibility that this effect reflects a difference between groups in the degree of relatedness between stimuli in equivalence classes.

Research outside of behavior analysis has explored the use of emotion induction procedures. Previous research has shown that depending on the emotion induced, it is possible to alter discounting when an emotion of “gratitude” was induced, compared to the induction of happiness or sadness. The current research explores the effects of emotion induction of these three emotions on an equivalence test. 120 undergraduate students were separated randomly into four groups according to the emotion they were induced: gratitude, happiness, sadness and control. After the emotion induction procedure, participants went through a respondent conditioning training of four stimuli classes and were later presented with a learning test of the relations trained and an equivalence test with a multiple matching-to-sample test. Results so far show that most of the participants (independent of their group) scored well in the learning test but didn’t achieve equivalence, however, the participants of the control group had better scores in the equivalence test. Higher scores on the learning test are observed in the gratitude and the control groups. Results are discussed with respect to the effectiveness of emotion induction procedures and their value in behavior analytical research.

Vascular Dementia is a type of neurocognitive disorder caused by a disruption in the blood supply to the brain. Such disruption can damage the brain cells, leading to problems in cognitive functioning such as remembering. The participant, Charlie, was a 74 year-old male diagnosed with Vascular Dementia. He lived in a nursing home at the time of the study. The study was a replication of an earlier experiment conducted with the same participant seven months earlier. In that study, a conditional discrimination training was presented by the use of either simultaneous (A-phases) or the simple-to-complex (B-phases) protocol. These phases were replicated in the current study. The stimuli, pictures of Charlie’s significant other, were presented using the Many-to-One training structure. Based on the findings from previous study, the inter-trial-interval was set to 5000 ms throughout the experiment. The results showed that the simple-to-complex protocol lead to higher accuracy of responding (see Figure 1). Additionally, the results also show how conditional discrimination may be sensitive to changes in cognition caused by additional disruptions of blood flow to the brain (a new stroke).

Training structures have been observed to influence yields in equivalence class formation. The Linear Series (LS) training structure provides the lowest yields. The current study investigates the influence of a LS training structure on the transfer of stimulus function and preference tests. Sixteen college students rated six images showing needle injections to different parts of a human hand. The images were rated on a 5-point Likert Scale ranging from least to most painful (1¬–5). The images rated as most and least painful were later employed as D1 and D2 stimuli, respectively. The D3 image showed a Q tip being pressed against a hand. Next, the participants trained six conditional discriminations in a LS training structure and tested of forming three 3-member equivalence classes. Six out of the 16 participants formed three 3-member classes, whereas five out of six participants successfully expanded the classes to three 4-member classes. Class expansion was done by training the needle injection images (D-stimuli) to the A-stimuli. In a preference test, three bottles were labelled with the B-stimuli. Three participants chose bottle B3, whereas one participant chose bottles B1 and B2. One out of five participants rated the B and D stimuli the same.

Relational Density Theory (RDT) provides a behavioral account of higher-order and self-organizing properties of equivalence networks. Similar to behavioral momentum theory, RDT posits that a first self-organizing property is the relative resistance of relational behavior as a product of behavioral mass. In RDT, mass is a function of the competing molar state volume (class size and nodal distance) and density (response strength) of a given equivalence class. The purpose of the present study was to extend upon prior research on RDT with 13 college student participants. In phases 1 and 2, 4 classes were established that differed in size (3-member, 4-member, 6-member, 8-member) in a one-to-many training structure (max. nodal distance = 1 for all classes). In phase 3, a single derived relation within each class was counter conditioned (-x), and changes in relational responding for each class were evaluated in phase 4. We predicted that classes with greater volume would be more resistant given similar obtained density values in phases 2 and 3. Results suggest that greater distance from the counter-conditioned relation was predictive of greater resistance to counter-conditioning, consistent with a first model prediction within RDT.

The effects of linguistic morphology of saying and reporting, and delay between the say-do-report moments, have not been systematically assessed in the say-do correspondence studies. In the present study adults were exposed to task, differentiated by the moment: 1) Say, i.e. answer questions; 2) Do, i.e. perform do-moment-related actions; 3) Report, i.e. answer questions about previous moments. In Experiment 1 groups of adults were exposed to three different linguistic morphologies (e.g. writing, vocalizing, pointing); whereas, for Experiment 2 two groups were exposed to different delay values (e.g. 0 and 24hrs). Results showed little difference between linguistic morphologies and correspondence, and less correspondence under the 24hr delay. These results are discussed in terms of the factors that modulate the establishment of correspondence (i.e. task dependence, existing temporality between the Say, Do and Report moments).

Research by Bland, Cowie, Ellfie, and Podlesnik (2018) demonstrated that a negative discriminative stimulus (i.e., stimulus correlated with the absence of reinforcers, S-) could be used to punish behavior maintained by positive reinforcers with pigeons. We replicated Bland et al. using the Portable Operant Research and Teaching Lab (PORTL). In the first phase, college students card touching was reinforced in the presence of one card (S+) and extinguished in the presence of an S-. In subsequent phases, we evaluated whether the S- functioned as a punisher when the S- was placed atop the S+ for 1-s intervals when presented (a) after an average of four responses to the S+ and (b) after each response to the S+. In both phases, participants’ behavior changed following the presentation of the S- when compared to a phase with the S+ only and a control stimulus. Most participants responding decreased in at least one of the punishment conditions. These results replicated Bland et al. because the presentation of the S- decreased the rate of responses maintained by positive reinforcement.

A recurring issue in the literature on resurgence of operant responding in human participants is lack of sensitivity to reinforcement schedules and extinction. In the present experiment, to increase the sensitivity to the schedule, a vigilance task was used. In the 1950s, Holland found that these tasks are useful to demonstrate systematic effects of different reinforcement schedules on human operant responding. Participants pressed a button to observe the state of a pointer on a dial (vigilance response). The pointer moved on a variable interval (VI) schedule and served as reinforcement of the vigilance response. When a change in the state was detected, pressing a red collect button produced points that were exchanged for money. A console with three buttons was built to establish two vigilance responses on a three-phase resurgence procedure. During Phase 1, presses on a green button were reinforced. During Phase 2, alternative presses on a blue button were reinforced and presses on the green button were extinguished. In Phase 3 presses on both buttons were extinguished. Resurgence of the vigilance response established in Phase 1 was found on the four participants. Alternative responses, however, were observed during most of the extinction sessions.

Autoshaping is a phenomenon that produces steady rates of responding, despite the lack of contingency between reinforcer and responding (Brown & Jenkins, 1968). Attempts to determine the processes responsible for the establishment and maintenance of autoshaped responding have produced conflicting results (e.g. Williams & Williams, 1969; Barrera, 1974). These experiments employed negative automaintenance procedures to determine the role of operant and respondent processes. A comparison of the procedures of each study revealed a discrepancy in these negative automaintenance procedures. This study attempted to determine the effect of this discrepancy on the persistence of autoshaped responding in each of the two procedural variations. This study was conducted using human participants in the Portable Operant Research and Teaching Laboratory (PORTL) ( Rosales-Ruiz & Hunter 2014). A forward pairing preparation was used to establish responding to two different manipulanda (see Brown & Jenkins, 1968). Once established, each manipulandum was assigned to one of the two different negative automaintenance conditions. Each participant was exposed to both conditions consecutively. Results suggest that the procedural variation may account for the discrepancy in the previous findings. Implications for the role of conditioned reinforcers and shaping in autoshaping and automaintenance are discussed.