Association for Behavior Analysis International

Search for stimulus class formation in animals has been elusive. There is no replication of reported positive findings. An experimental model should allow replication intra and inter subjects, with reduced variability, in any laboratory minimally equipped. True matching to sample - MTS has been demonstrated for identity but not for arbitrary stimulus relations in Capuchin Monkeys. Monkeys with history of generalized identity MTS were trained with success in a repeated shift simple simultaneous discrimination procedure with 3 and 12 stimuli to choose the stimulus of a category. Twelve categories with 5 members each were positive in shifts. Later they showed consistent performance when a new stimulus was introduced in each category. Silhouettes were then introduced and consistent performance was obtained. Transfer of simple discrimination to new stimulus, including silhouettes, is still stimulus generalization. Icons will be inserted and training will be provided if necessary. Arbitrary stimuli will be inserted to form a functional class with 5 pictures, one silhouette, one icon and one arbitrary stimulus in each class. Transfer to category conditional discrimination in the MTS format with 3 and 12 simultaneous comparisons with the pictures will be tested and trained if necessary. Emergence of conditional relations of silhouettes, icons and arbitrary stimuli -trained in the simple discrimination format- to the pictures will be tested. Consistent association of arbitrary members to every baseline member with training of symmetric and transitive relations may facilitate the insertion of the new stimulus into the category set by functionality.

Behavior analysts have been investigating false memories using the stimulus equivalence paradigm. In this experiment, we aimed to investigate if transfer of different function after equivalence class formation would affect false memories. The Experimental Group (EG) learned matching-to-sample (MTS) relations to establish three four-member (Classes 1, 2 and 3) and three 12-member (Classes 4, 5 and 6) equivalence classes. Stimuli were familiar pictures (nodes) and nonsense words. In Classes 4, 5 and 6, the familiar stimuli were pictures of faces portraying emotions of happiness, neutrality, and anger, respectively. Participants who meet criteria in equivalence MTS tests and participants who did not perform any MTS tasks (control group- CG) evaluated two nonsense words from Classes 4, 5 and 6 and the pictures portraying emotions, using the Sematic Differential (SD). A week later, participants from EG saw three lists with nine nonsense words from Classes 4, 5 and 6, in a total of 27 stimuli in the list. After a three minutes distracter task, participants saw three new lists composed by all stimuli from previously list (Targets), the remaining nonsense words form Classes 5, 4 and 6 (Critical Distractors) and the nonsense words from Classes 1, 2 and 3 (Non-related distracters). Participants should indicate witch stimuli were presented on the first lists. The seven participants from EG who finished the experiment evaluated the nonsense words of each equivalence class similar to the evaluation of the faces by the CG. They also recognized significantly more critical than non-related distracters only in the list composed by stimuli from neutral class.

The current study tested if names given by adult participants to the stimuli when asked to talk aloud became a part of the stimulus classes. Training the conditional discriminations and testing for emergent relations was conducted with a Many-to-One training structure in 6-s delayed Matching-to-Sample (DMTS) setup, with potentially three 3-member classes. After the test, the vocal names given by the participants were included as textual stimuli in sorting tests. In the sorting tests, the three first stimuli drawn from the stack of cards were names given by the participant, one name from each class. The next nine stimuli were the stimuli used in the DMTS procedure, with the three last stimuli in the stack being the stimuli corresponding with the names at the beginning of the stack. If the participants put a stimulus on top of one of the names, the card with the name automatically hid the stimulus. Initial results show that the names given by the participants become a part of the stimulus classes (see Table 1). Thus, the three-member classes seemingly also include the names participants give the stimuli.

This experiment investigated whether mixing baseline among symmetry and equivalence test trials enhances the yields of equivalence class formation. Twenty adults were exposed to two test protocols, in counterbalanced order: Simultaneous with Baseline Intermixed and Simultaneous with Post-test of Baseline (Conditions A and B, respectively). The conditions varied in respect to the test procedure and stimuli set. Both started by training AB, AC, AD, and AE relations concurrently. Next, a test assessed the maintenance of baseline and the emergence of symmetry (BA, CA, DA, EA) and equivalence (BC, BD, BE, CB, CD, CE, DB, DC, DE, EB, EC, ED) relations. In Condition A, baseline, symmetry, and equivalence test trials were presented simultaneously. In Condition B, baseline test trials were presented after the test for symmetry and equivalence intermixed. All participants formed equivalence classes in both conditions. Condition A presented slightly lower yields of immediate emergence (16 participants) than Condition B (18). Three out of the four participants with delayed emergence in Condition A initially failed in symmetry trials; the other two in Condition B, failed exclusively in equivalence trials. Condition A produced significantly faster responses to the first five equivalence test trials.

The role of instructions in stimulus equivalence research has often been discussed without being considerably empirically investigated. Generally, as little information as possible are given to participants, to limit the potential effects of instructions on stimulus equivalence responding. In this present stimulus equivalence experiment, instructions are investigated by using a linear series training structure with three 3-member classes. Participants were assigned to either one of three conditions, consisting of: general instructions (e.g., "your task is to find out which stimuli that belongs together"), specific instructions (e.g., "this stimulus is a fish and belongs together with the car which is this stimulus") or a control group without instructions about the task. The depended measures are number of training trials and equivalence responding. As shown in Table 1, the results show that the condition with general instructions gave a higher result of equivalence responding (i.e., 60 %) compared to the other conditions (i.e., 10 % and 0 %, respectively).

Symmetry is an emergent relation for stimulus equivalence. Although human participants are relatively easy to show evidence for symmetry, nonhuman animals typically do not. However, Urcuioli (2008) reported that pigeons demonstrated symmetry in his studies using successive (go / no-go) matching-to-sample procedure, and proposed a theory of pigeons' equivalence-class formation. In the present study, we checked Urcuioli's experiment. Four pigeons received hue-form arbitrary, hue-hue, and form-form identity matching training. In the training, fixed-interval (FI) 5-s schedule was arranged to positive trials and extinction (EXT) was arranged to negative trials. Each training session consisted of 48 positive and 48 negative trials that randomly appeared. Each pigeon was trained until it achieved a discrimination ratio of 0.80 or higher on each matching task for 5 of 6 consecutive training sessions. After the training, form-hue symmetrical relations was tested in non-reinforced trials. The test consisted of 104 trials that included 96 training trials and 8 symmetry probe trials. Figure 1 shows the results of four pigeons. Three of them (P1, P4, P5) responded more to the positive comparison stimuli than to the negative comparison stimuli and demonstrated symmetry. As we reproduced the results of Urcuioli's experiment, we conclude that Urcuioli's theory is correct.

Understanding how symbolic behavior may be enhanced can result in advances in the development of applied technologies. Participants in this study were eighteen university students ranging from 18-40 years (average = 29 years). Participants were asked to group 15 flashcards with black-and-white abstract shapes individually printed on them. Participants were then assigned to one of the two groups: Tact or Mand. Participants in the Tact Group were taught to say "Paf" in the presence of shape 1; "Vek" for shape 2, and "Zog" for shape 3. Participants in the Mand Group were asked to choose three snack items of their choice and were taught to mand for exchanging shape 1 for snack 1, shape 2 for snack 2, and shape 3 for snack 3. After serial training of AB, BC, CD, and DE relations, all probes for stimulus equivalence were presented in one block. Participants were again asked to group flashcards. Results showed no differences between groups suggesting that pre-training stimuli to exert control over tact and mand responses do not produce equivalence class enhancement. We suggest other types of pre-training of verbal operants to clarify the role of stimuli with a trained history to control verbal behavior in equivalence class formation.

Neuroscientific measures such as electroencephalography (EEG) can increase our understanding of complex human behavior. The N400 component is a negative peak elicited 400 ms after stimulus presentation and was originally observed when the end of a sentence was not related with the word presented before. The N400 is not observed if a relation has been established between the stimuli. The N400 effect also applies when the stimuli are not directly trained as in the test for equivalence class formation and can also be elicited by other meaningful stimuli, including pictures. This can be argued to be a measure of the relational strength between two stimuli as a result of the participant's learning history. The current poster presents an experiment with six healthy adult volunteers where we investigated the N400 during a priming procedure after forming three 3-member classes with C-stimuli as familiar pictures. The results showed a reduction of the N400 component over repeated presentations of the unrelated stimulus pairs when the unrelated stimulus pairs were analyzed separately in four consecutive blocks (see Figure 1). Hence, the findings indicate a possible habituation after repeated stimulus pair presentation.

In the present experiment, 12 conditional discriminations were trained as a baseline for three 5-member equivalence classes. All stimuli were abstract shapes. The research question was to study the effect of introducing new stimuli in a sorting test after MTS training or after MTS training and testing. In the experiment, 30 adults participated randomly assigned in two groups. Group 1 had the MTS training and test, followed by a sorting test, while Group 2 had the MTS training, sorting, MTS test and another sorting. Both groups were presented for five new stimuli in the sorting tests. The results show that 13 out of 15 participants in Group 2 sort the new stimuli in the experiment-defined classes (see Figure 1). In Group 1, four out of 15 participants sort new stimuli in the experiment-defined classes, while five participants sort the new stimuli together with three of the experiment-defined classes. In Group 1, three participants do not form equivalence classes. In Group 2, one participant does not respond in accordance with stimulus equivalence. There are higher total scores during MTS test in Group 2 than in Group 1, but the difference is not significant.

The present integrated series of experiments was designed to identify whether the differential order of exposure to test phases (Standard, Inverse, Mixed) respect of a training structure Many to One (MTO), One to Many (OTM) or Lineal Series (SL) has an effect on the emergence on the emergence of equivalence relations. Sixty psychology undergraduates from several universities of Guadalajara participated in the experiments. A Matching To Sample task software was programmed to training the conditionals discriminations (three conditions) and to testing three 3-member equivalence sets of stimuli, symmetry, transitivity and equivalence in four conditions. Overall, for these preparations we observe an effect, it seems has identified an explicit interaction between certain training structure and certain order of exposure to test phases. Hence, the results suggest that the performance in testing for emergence relations depends of order of exposure to test phases respect of the training structure used. Furthermore the accuracy was better with the SL training structure.

Saccade adaptation is a form of motor learning that maintains saccade accuracy in response to new sensorimotor contingencies. We know that reinforcement learning can induce saccade adaptation. Previously, an arbitrary reinforcer was used such as an auditory tone or viewing the target on the fovea to control saccades amplitude. This experiment asks whether changes in saccade amplitude may be induced by the ability to perform a visual discrimination task as a consequence. In a 4AFC task, five subjects were instructed to report which symbol was briefly (60ms) displayed across the whole screen immediately after a saccade. The possibility to perform the discriminative task was contingent on meeting a specific saccade amplitude criterion: when saccades did not meet the criterion, one of four irrelevant symbols was displayed such that the participant could not perform the discriminative task. We observed consistent changes in saccadic amplitude across sessions that followed the criterion manipulations. We conclude that saccades are operant behavior reinforced by the ability to perform visual tasks.

A procedure to study self-controlled behavior is the "refraining-from-eating" situation in which a food dispenser is presented once (SR1) within a repetitive time cycle, and it can be presented again (SR2) after the cycle has elapsed according to the following contingency: If the pigeon tries to consume SR1, the food is immediately withdrawn and the SR2 presentation is cancelled; if, however, the pigeon refrains from attempting to consume SR1, then it can consume SR2. In the present study, the contribution of stimulus control to "refraining" from taking SR1 was assessed. The effects of (a) signaling both SR1 and SR2 with the same color, (b) signaling both the SR1 and SR2 with different colors, (c) adding a houselight during SR1, and (d) adding a response key (which did not have to be pecked) during SR1 were evaluated. Fewer SR1 interruptions occurred when different stimuli signaled each food-dispenser presentation than when the same stimulus signaled both food presentations. When an illuminated response key was presented concurrently with SR1, interruptions were markedly reduced, to near-zero levels in those pigeons that pecked the key. These data are evidence that "resistance to temptation," an example of self-controlled behavior, is affected by stimulus control.

When reinforcement is contingent on a response in the presence of a compound stimulus, the contingency will be satisfied even if the response occurs under the control of only one of the elements. In this study, 12 bees received simple discrimination training followed by probes of stimulus control for isolated elements of the compound. The learning criterion was at least 90% of correct responses in two consecutive blocks of 20 trials. For half of the bees, the positive stimulus (S +) was a circle with blue border and white center and the S- was a circle with black border and yellow center; for the other half, the function of the stimuli was reversed. In the Probes phase, each compound was decomposed into two elements (a circular border and the inner circle) and all combinations of two elements were probed, in 16 trials. Stimulus control by both positive elements was strong for 8 of 12 bees (but somehow higher for the inner element of the circle), while four bees showed evidence of responding almost exclusively to only one of the positive elements. Probing stimulus control after discrimination training with compounds should be a standard protocol to identifying restricted stimulus control. Key word: discrimination learning, stimulus control, compound stimuli, bees

Available procedures to study working memory capacity in rodents generally use manual arena or maze tasks. For example, the odor span task uses an incrementing non-matching- to-sample procedure in which digging in cups scented with novel olfactory stimuli results in reinforcement on each trial, while responses to previously presented stimuli are not reinforced. The present study modified this procedure for use in an automated operant chamber using a 15-channel olfactometer. Rats were trained on a go, no-go procedure to make nose-poke responses in a port through which odorants were delivered. Responses to each odorant were reinforced on an FI 5-s schedule the first time it was presented, but once an odor had been presented, responses to that odor were no longer reinforced. Rats rapidly learned to differentiate between session-novel and session-familiar odors and responded at high rates to new stimuli and much lower rates to repeated stimuli. Determinants of remembering were assessed by manipulating a number of variables including frequency and recency of repeated odors, a mid-session delay, and a distractor task. Accuracy was affected by all these variables, but the greatest disruption was observed after a mid-session delay, with or without a distractor task.