If an organism is explicitly taught an A->B association, then might it also spontaneously learn the symmetrical B->A association? There is only a small amount of evidence that attests to the detection of emergent symmetry in nonhuman animals (e.g., one chimpanzee and two pigeons). This report examines the necessary and sufficient conditions for finding emergent symmetry in pigeons while attempting to control for the problems of spatial and temporal location found in previous symmetry and stimulus equivalence experiments. Using a successive go/no go matching-to-sample procedure, which showed all of the training and testing stimuli in one location, four experimental manipulations were examined. In Experiment 1 temporal location was controlled without the inclusion of identity matching intermixed with arbitrary matching; Experiment 2 contained identity matching with stimuli different from arbitrary matching; in Experiment 3 identity matching was trained to criterion and then intermixed with arbitrary matching; and in Experiment 4 two sets of arbitrary matching were trained (e.g., AB and CD) but only one of those stimulus sets was trained in identity matching (e.g., AB). No evidence of emergent symmetry was found in Experiments 1 and 2. In Experiment 3, two pigeons showed moderate evidence of emergent symmetry, one pigeon showed suggestive evidence of emergent symmetry, and one pigeon did not show any evidence of emergent symmetry. In Experiment 4, two pigeons showed moderate evidence of emergent symmetry with the AB Stimulus Set (one of those pigeons also showed suggestive evidence of emergent symmetry with the CD Stimulus Set) and one pigeon did not show any evidence of emergent symmetry with either stimulus set. These data suggest that intermixing identity matching with the same stimuli used in arbitrary matching is a necessary, but not sufficient condition to obtaining emergent symmetry in pigeons.

Pesquisas apontam que o DMTS tem se mostrado mais eficaz no estabelecimento de classes de equivalência e maior grau de relacionamento entre estímulos de uma mesma classe do que o SMTS. Uma explicação possível para essas diferenças é que o aumento no intervalo de atraso entre a retirada do estímulo modelo e apresentação dos estímulos comparação em procedimentos de DMTS leva a uma maior exposição dos participantes ao estímulo modelo. Essa hipótese se alinha a resultados obtidos em experimentos que avaliam parâmetros de observação no estudo de discriminações condicionais, que demonstram que maiores durações de fixação do olhar sobre estímulos modelo estão associadas a maiores porcentagens de acerto no aprendizado em tarefas de MTS de identidade com modelos múltiplos. O presente estudo teve como objetivo avaliar os efeitos do emprego de atraso sobre a formação de classes e sobre parâmetros de fixação do olhar sobre os estímulos modelo em tarefas de MTS a partir do rastreamento dos olhos. Nove estudantes de graduação ou pós-graduação foram expostos a um treino de discriminações condicionais com estrutura OTM para estabelecer as classes A1B1C1, A2B2C2, A3B3C3 e A4B4C4. Cada uma dessas classes foi associada a uma condição de atraso durante o treino: simultânea, atrasos 0s, 2s e 4s respectivamente. Posteriormente, foram submetidos aos testes das relações emergentes de equivalência CB e BC, e simetria BA e CA em tentativas de DMTS com atraso 0s. Sete participantes atingiram o critério de aprendizagem durante o treino e conseguiram formar classes. Apesar de todos os participantes que atingiram os critérios terem aprendido primeiro relações de DMTS, não foram observados padrões ou diferenças significativas no aprendizado das discriminações condicionais ou formação de classes em nenhuma das condições. A análise dos parâmetros de observação também não revelou diferenças significativas entre frequência e duração de fixações do olhar sobre o modelo em nenhuma nas tentativas de SMTS ou DMTS com qualquer valor de atraso. Todavia, foi observado um efeito de prática, com valores maiores de duração de fixação nos blocos iniciais do treino. Esses resultados sugerem, portanto, que o maior tempo de exposição aos modelos não é suficiente para explicar as diferenças observadas no estabelecimento de classes de equivalência em tarefas de DMTS. Entretanto, alinham-se com a perspectiva de que um maior engajamento em comportamento de observação pode estar relacionado a um maior controle de estímulos dados os resultados do efeito de prática / Researches have indicated that the DMTS has been shown to be more effective in establishing equivalence classes and a higher degree of relationship between stimuli of the same class than SMTS. One possible explanation for these differences is that the increase in the delay interval between the withdrawal of the stimulus model and the presentation of the comparison stimuli in DMTS procedures leads to greater exposure of the participants to the stimulus model. This hypothesis aligns with results obtained in experiments that evaluate observation parameters in the study of conditional discriminations, which demonstrate that longer fixation durations on model stimuli are associated with higher percentages of learning success in multiple-models identity MTS tasks. The objective of the presente study was to evaluate the effects of the use of delay on the formation of classes and on parameters of fixation of the look on the model stimuli in MTS tasks from the eye tracking. Nine graduate or postgraduate students were exposed to conditional discrimination training with OTM structure to establish classes A1B1C1, A2B2C2, A3B3C3 and A4B4C4. Each of these classes was associated with a delay condition during training: simultaneous, delays 0s, 2s and 4s respectively. Subsequently, they were submitted to the tests of the emergent equivalence relations CB and BC, and symmetry BA and CA in attempts of DMTS with delay 0s. Seven participants reached the learning criterion during the training and were able to form classes. Although all participants who met the criteria first learned DMTS relationships, no significant patterns or differences in learning conditional discrimination or class formation were observed in any of the conditions. The analysis of the observation parameters also did not reveal significant differences between frequency and duration of fixations of the look on the model in any in the attempts of SMTS or DMTS in any with any value of delay. However, a practical effect was observed, with higher fixation duration values in the initial training blocks. These results suggest, therefore, that the greater time of exposure to the models is not enough to explain the differences observed in the establishment of equivalence classes in DMTS tasks. However, they are aligned with the view that greater engagement in observation behavior may be related to greater control of stimuli given the results of the effect of practice

Submitted by Aelson Maciera (aelsoncm@terra.com.br) on 2017-09-27T18:38:57Z No. of bitstreams: 1 DissSDCP.pdf: 2530109 bytes, checksum: 969acf4fbd844c709316d1eb4bc99426 (MD5) / Approved for entry into archive by Ronildo Prado (producaointelectual.bco@ufscar.br) on 2017-10-10T18:02:51Z (GMT) No. of bitstreams: 1 DissSDCP.pdf: 2530109 bytes, checksum: 969acf4fbd844c709316d1eb4bc99426 (MD5) / Approved for entry into archive by Ronildo Prado (producaointelectual.bco@ufscar.br) on 2017-10-10T18:03:00Z (GMT) No. of bitstreams: 1 DissSDCP.pdf: 2530109 bytes, checksum: 969acf4fbd844c709316d1eb4bc99426 (MD5) / Made available in DSpace on 2017-10-10T18:11:07Z (GMT). No. of bitstreams: 1 DissSDCP.pdf: 2530109 bytes, checksum: 969acf4fbd844c709316d1eb4bc99426 (MD5) Previous issue date: 2017-03-29 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / False memories can be defined as recollections that do not correspond to the facts experienced by subjects. Cognitive research have pointed out that the referred phenomenon is, to some extent, determined by semantic relations among stimuli. Based on these studies, behavior analysts have proposed experimental analogous of false memories using the stimulus equivalence paradigm to establish symbolic relations. The current master’s thesis was composed by one preliminary study and two main studies. The preliminary study aimed to assess the valence and arousal of pictures depicting facial expressions, and to select the meaningful stimuli that were employed on the class formation procedures of Studies 1 and 2. Study 1 aimed to analyze possible effects of emotional stimuli with different valences (positive, neutral and negative) on the occurrence of false memories. For this purpose, a systematic replication of Aggio’s (2014) Study 3, using lists of symbolically related stimuli through the stimulus equivalence paradigm was carried out. The main experimental manipulations of Study 1, in the present thesis, were the control of valence and arousal magnitudes attributed to the emotional stimuli; the use of emotional stimuli in the class formation procedure, which contained the unrelated distracting stimuli; and the randomization of list order presentation. Study 1 was structured in four phases, which encompassed 1) the establishment of six equivalence classes; 2) a false memories test with a memorization task and a recognition test, held one week after the beginning of Phase 1; 3) class merger and maintenance test, and 4) stimuli assessment through a semantic differential. Results showed that the false memories effect, under symbolic control, occurred only in relation to the positive stimuli list. It was hypothesized that the stability of the equivalence classes was an intervening variable. In order to verify the aforementioned hypothesis and search for new evidences regarding the effects of emotional valence on the establishment of false memories, a second study was proposed. The Study 2 had the same procedure as Study 1, with the exception of two changes: the reduction of the time interval employed between the first and the second phase, and the insertion of baseline review blocks of the six equivalence classes before the false memories test. These changes were adopted in order to increase the probability that the equivalence relations were stable and strengthened at the moment of the false memories test. Results showed the occurrence of the false memories effect in positive and neutral lists. These results suggest that stimuli with negative emotional valence can reduce the probability of false memories occurrence. / Falsas memórias podem ser definidas como lembranças não fidedignas aos fatos vivenciados pelos sujeitos. Pesquisas de base cognitivista têm apontado que o referido fenômeno é, em alguma medida, determinado por relações semânticas entre estímulos. A partir dessas evidências, analistas do comportamento propuseram análogos experimentais de falsas memórias com o emprego do paradigma da equivalência de estímulos como recurso para o estabelecimento de relações simbólicas. A presente dissertação foi composta por um estudo preliminar e por dois estudos principais. O estudo preliminar objetivou avaliar as propriedades de valência e alerta de imagens de expressões faciais, além de selecionar os estímulos significativos que seriam empregados na formação das classes de equivalência estabelecidas nos Estudos 1 e 2. O Estudo1 visou analisar os possíveis efeitos de estímulos emocionais com diferentes valências (positiva, neutra e negativa) sobre a ocorrência de falsas memórias. Para tanto, foi proposta uma replicação sistemática do Estudo 3 de Aggio (2014), o qual fez uso de listas de estímulos simbolicamente relacionados por meio do paradigma da equivalência de estímulos. As principais manipulações experimentais propostas, no Estudo 1, da presente dissertação, foram o controle das magnitudes das propriedades de valência e alerta atribuídas aos estímulos emocionais; a utilização de estímulos emocionais na formação das classes que continham os distratores não relacionados e o balanceamento das ordens de apresentação das listas. O Estudo 1 foi estruturado em quatro fases que envolveram: 1) estabelecimento de seis classes de equivalência; 2) teste de falsas memórias, compreendendo uma tarefa de memorização e um teste de reconhecimento, realizados uma semana após o início da Fase 1; 3) teste de fusão e de manutenção de classes e 4) avaliação de estímulos por meio de instrumento de diferencial semântico. Os resultados demonstraram que o efeito de falsas memórias, sob controle simbólico, ocorreu apenas na lista de estímulos de valência positiva. A estabilidade das classes de equivalência foi apontada como uma possível variável interveniente na determinação dos referidos resultados. A fim de verificar a hipótese supracitada e buscar novas evidências em relação aos efeitos da valência emocional, sobre o fenômeno das falsas memórias, foi proposto um segundo estudo. O Estudo 2 teve um procedimento idêntico ao adotado no Estudo 1, com exceção de duas modificações: a redução do intervalo de tempo empregado entre a primeira e a segunda fase e a inserção de blocos de revisão de relações de linha de base das seis classes de equivalência, antes da realização das tarefas de teste de falsas memórias. As alterações citadas foram adotadas a fim de aumentar a probabilidade de que as relações de equivalência estivessem estáveis e fortalecidas no momento de realização das tarefas de teste de falsas memórias. Os resultados indicaram a ocorrência do efeito de falsas memórias nas listas de valência positiva e neutra. Os referidos resultados sugerem que estímulos com valência emocional negativa podem reduzir a probabilidade de ocorrência de falsas memórias.

The present studies evaluated the effect of training structures on the development of equivalence classes in college students and individuals with intellectual disabilities. Experiment 1 evaluated the effects of two types of training structures, One-To-Many (OTM) (AB, AC, AD), and Many-To-One (MTO) (BA, CA, DA), on the establishment of equivalence classes in college students. A between group comparison was used in Experiment 1. Forty-two participants were randomly assigned to two different groups. Twenty-one were assigned to the OTM group and twenty-one to the MTO group. Participants in both groups were taught 3 four-member stimulus classes. Participants in both groups were exposed to conditional discrimination training, mixed training, symmetry and equivalence test. Response accuracy and response latency were measured in both groups. The results showed that the MTO training structure was slightly more effective in establishing equivalence classes in college students. In the Experiment 2, six young adults with intellectual disabilities were taught mathematical relations using the MTO training structure which was the most effective training structure in Experiment 1. All participants were taught three 3-member stimulus equivalence classes using the MTO training structure. The experimental sequence consisted of a generalization probe and pretest followed by conditional discrimination training, symmetry test, equivalence test, and posttest. Upon the completion of the training and testing phases a generalization probe was evaluated. Five participants demonstrated equivalence relations. The results show that the MTO training was superior to the OTM in the Experiment 1. Response latencies were faster in the MTO group during the training phases and slower in the testing conditions. Experiment 2 showed that only five participants demonstrated equivalence relations and transferred untaught relations to new setting. Results and implications are discussed in light of the research on equivalence and training structures in both adults and individual with intellectual disabilities.

The present study evaluated the use of stimulus equivalence to train novel money relations with adults with developmental disabilities. Three programs from the PEAK Relational Training System: Equivalence Module (PEAK-E) were used: Reflexivity: Money, Symmetry: Money to Monetary Value and Equivalence: Monetary Exchange. The overall goal of the study was to evaluate how stimulus equivalence could help the subjects improve money skills, a socially significant behavior for all adults. The study used a multiple probe design across programs to evaluate skill acquisition following the train/test procedure for each program. Results demonstrated a significant increase in PEAK scores from baseline probes for two of the three programs for all subjects. Additionally, results showed that all subjects learned untrained skills. The results of this study demonstrated the value of stimulus equivalence for teaching money skills as well the efficacy of the PEAK-E training system to teach adults with developmental disabilities.

The following studies examined the effectiveness and the efficiency of individual and small group equivalence-based instruction on English and math relations with elementary school children. The first experiment implemented a multiple-probe design across three typically developing third graders using an automated procedure to evaluate the stimulus equivalence paradigm (SEP) on establishing English and math equivalence classes. Since some researchers have proposed that covert behavior facilitates in the formation of equivalence classes (Horne & Lowe, 1996; Stromer, Mackay, & Remington, 1996), the second experiment examined the possible role of covert behavior, more specifically, a visual imagining strategy, on correct responding after remedial instructional sessions. To further extend SEP into more academic environments, the third experiment used an observational learning procedure to evaluate the formation of English and math symmetry relations as well as the formation of equivalence classes in a small group format. The emergence of topography-based responding was also assessed after instruction for all three experiments.

The purpose of this study was to evaluate the efficacy of the procedures described in the PEAK-E curriculum in generating derived intraverbal categorization responses in both a vocal categorization context and written problem-solving context, replicated across three children with disabilities. Six four-member equivalence classes were taught, including three class member stimuli (A, B, and C) as well as one function class name (D), using a match-to-sample arrangement. These classes were divided into two stimulus sets (i.e., classes 1-3 and classes 4-6) and trained using pre-class (D-C training followed by mixed A-B/B-C training) and post-class formations (Mixed A-B/B-C training followed by D-C training). The procedures were efficacious in generating derived intraverbal categorization responses for one participant in a vocal context, and additional exemplar training was required for the emergence of vocal categorization responses in the other two participants. None of the participants were able to solve the written problem-solving tasks following training and testing of all target relations.

The PEAK Relational Training System's Direct Training Module was examined to determine age appropriate development norm ages in which to compare to persons with disabilities. Fifty-one typically developing children between ages of 2-18 were scored for the PEAK direct training assessment module by novel implementers (parents, caregivers) as well as professional educational teachers and administrators. Assessment implementation brought scores from fourteen different adult individuals and fifty-one different children's assessment forms for this study. The inter-rater reliability showed consistency of results with the relationship between Peak and age. Inter-rater agreement had a high degree of agreement through the same age, male or female, grade levels with student assessments on the PEAK rating pyramid. These subjective reports may be related and have one advantage that data results may be displayed conveniently in graphical form. The results suggest that there was no correlation between PEAK Total Scores and sex with R-Squared of .49, and produced a significant fit to the data (p < .001). Total PEAK by Age Group with an One Way Anova being: F(4,45) = 23.57, p < .001 suggest that the differences between ages in terms of PEAK total Score was significant.

The present study evaluated the utility of the methods outlined in the Promoting the Emergence of Advanced Knowledge Relational Training System Equivalence Module (PEAK-E) through a single-case evaluation. Validity, reliability, and effectiveness were the variables explored to assess the degree to which the assessment was able to identify appropriate skills for targeted intervention, and the degree to which the programs were efficacious in teaching the targeted skills. Baseline results suggested that the programs identified through the PEAK-E assessment were not within the participants’ repertoires prior to the intervention. Following the implementation of 9 programs across three participants with autism, mastery was achieved for all of the directly trained relations, and all targeted derived relations emerged for 8 of the 9 programs

The present study evaluated the validity and efficacy of the Promoting the Emergence of Advanced Knowledge Relational Training System Equivalence Module (PEAK-E) using a single-case, multiple baseline across skills design with 3 adolescents with autism. Total scores from the PEAK-E assessment (PEAK-E-A) and programs were used to evaluate the degree to which the assessment was able to identify language skills to teach using the PEAK-E curriculum, and how effective the PEAK-E curriculum was at teaching those target skills. Baseline results suggested that the programs selected using the PEAK-E-A were not currently in the participant’s language repertoire. Upon completion of 9 total programs across three participants (3 programs per participant), mastery was achieved and all derived relations emerged for all 9 programs. The PEAK-E pre-assessment, assessment, and training programs were proven to be valid and efficacious at improving participant’s language skill repertoire. Keywords: autism, applied behavior analysis, verbal behavior, derived relational responding, stimulus equivalence, PEAK