|
| Alternative Paradigms for Studying Equivalence Class Formation and its Neural Correlates |
| Sunday, May 28, 2006 |
| 1:30 PM–2:50 PM |
| International Ballroom South |
| Area: EAB; Domain: Basic Research |
| Chair: Lanny Fields (Queens College, City University of New York) |
| Abstract: This symposium will consider alternative training paradigms and testing for studying the formation of equivalence classes that is more compatible with exploring the neural correlates of equivalence relations than a training and testing conducted in a typical MTS format. The alternative training paradigm, the stimulus pairing yes no procedure (SPYN), however, to date has not proven to be reliable. Two methods will be described that enhance class formation using the SPYN format for training and testing. In addition to its’ utility for studying the neural correlates of equivalence relations, it also extends demonstrations of equivalence class formation in different training and testing formats. A second study will describe a testing strategy that can be used to isolate the neural correlates of emergent relations in equivalence classes, nodal function, and the effects of nodal distance. In addition, the data obtained from this procedure also shows that equivalence relations are resistant to disruption by the intermixing of other conditional relations that are unrelated to each other. The last presentation is an fMRI study that describe refinements in the identification of the neural substrates of the baseline and emergent relations in equivalence classes. |
| |
| Behavioral Designs for Identifying the Neural Correlates of Equivalence Relations. |
| LANNY FIELDS (Queens College, City University of New York) |
| Abstract: The presentation will consider the advantages and disadvantages of using trials presented in a MTS format for the measurement of the neural correlates of equivalence class formation, an alternative to the MTS format studying the neural correlates of equivalence, and a testing strategy that can isolate the neural correlates of nodal function and nodal distance. One alternative to MTS trials that will be considered is the trace stimulus pairing yes no (SPYN) procedure in which one stimulus is presented at any time during training and testing. Thus, patterns of neural activation can be correlated with individual stimuli used during training and testing. These will be described in the presentation. When measuring the neural correlates of equivalence class formation, it is necessary to subtract the pattern of activation by an equivalence probe from a comparator task that matches the equivalence probe in all regards but the operation that characterizes equivalence. One strategy is to use pseudo-equivalence probes. Their construction and potential utility will be described in the presentation. The behavioral effects of testing with pseudo-equivalence probes has not yet been investigated. Procedures that produce reliable equivalence class formation with the stimulus pairing yes-no procedure have yet to be developed. The effects of these probes on the stability of equivalence classes on the behavioral level has not been explored. These matters require resolution at the level of behavior before the procedures can be used to investigate the neural correlates of equivalence class formation, nodal function and the effects of nodal distance. |
| |
| Preliminary Training that Enhances Equivalence Class Formation with a Trace Stimulus Pairing Yes-No Procedure. |
| MICHAEL E. MARROQUIN (Queens College and The Graduate Center, City University of New York), Lanny Fields (Queens College, City University of New York) |
| Abstract: Equivalence classes are typically established using trials presented in a matching to sample format that contains a sample from one classes, one comparison from the same class and other comparisons from other classes. Equivalence classes, can also be established where training and testing are conducted in a stimulus pairing yes-no format. Subjects are presented with only one stimulus at a time during training and testing. In addition, reinforcement is presented for responding “YES” when the two stimuli on a trial are from the same potential class and “NO” when the stimuli are from different classes. Without preliminary training, 50% of subjects form 2-node 4-member equivalence classes. In most cases, the failure of class formation involve failures of transitivity. Thus, the induction of a generalized transitivity repertoire prior to equivalence class formation should enhance the percentage of subjects who form equivalence classes. We studied the effects of two transitiivty induction procedures on the subsequent formation of equivalence classes. All procedures were conducted in the trace SPYN format. In Multiple Exemplar Training (MET), many sets of nonsense syllables were used as A, B and C stimuli. For each set, AB and BC relations were trained, after which transitivity was assessed with AC probes. If passed, the cycle was repeated with a new set of stimuli. If failed, the AC relation was directly trained and then the cycle was repeated with a new set of stimuli. The cycle was repeated until transitivity tests were passed with three consecutive new stimulus sets. 6 of 34 subjects showed the emergence of the transitivity repertoire and all then formed new equivalence classes. The remaining 28 subjects did not complete transitivity induction in the allocated time: 2 hours. While effective, then, the MET approach was quite inefficient.In Programmed Trial Unique Training (PTUT), AB and BC were trained after which AC was also trained. AC probe trials were never presented. This cycle was repeated with new sets of stimuli that changed gradually from an initial set that consisted of semantically related terms to a final set that consisted on nonsense syllables. The content of each stimulus set changed gradually in 6 steps from semantically related stimuli to nonsense syllables. All 14 subjects completed this preliminary procedure in short time. All relations were acquired in a minimal number of trials, and subjects typically responded accurately on the initial AC |
| |
| Effect of Pseudo-transitivity and Pseudo-equivalence Probes on the Stability of Equivalence Classes. |
| DANIELLE TITTELBACH (Queens College and The Graduate Center, City University of New York), Lanny Fields (Queens College, City University of New York) |
| Abstract: When exploring the neural correlates of equivalence class formation, the neural correlates of transitivity can be studied by measuring BOLD fMRI activations in the presence of transitivity probes and pseudo transitivity probes. After training AB and BC, AC is a true-transitivity probe. After training EF and GH, EH and GF are pseudo-transitivity probes. The match true transitivity probes because each contains a sample stimulus that was also a sample in training and a comparison that was also a comparison in training. They differ with regard to the in the presence and absence of a nodal stimulus that links the A and C stimuli in the true transitivity probe and the absence of such a nodal linkage between the stimuli in the pseudo transitivity probes. Thus, the neural correlates of transitivity ought to be adduced by the subtraction of the activation produced by WZ and YX from that produced by AC. To date, the behavioral effect of presentation of pseudo transitivity probes on the stability of true transitivity probes has not been explored. That will be described in this presentation. After training AB, and BC, AC emerged. That was followed by training EF and GH. Thereafter, pseudo transitivity probes are EH and GF were presented. Finally, EH and GF were presented along with AC. The transitivity indicative performances occasioned by AC trials remained stable when AC trials were intermixed with the EH and GF probes. These results showed that pseudo-transitivity probes do not disrupt transitive performances. A subsequent experiment explored the effects of intermixing pseudo-symmety, pseudo-transitivity, and pseudo-equivalence probes with test performances occasioned by the symmetry, transitivity, and equivalence probes used to document the presence of equivalence classes. Given the behavioral stability of performances occasioned by these probes, the used of pseudo probes can be used to measure the neural correlates of the emergent relations in equivalence classes. The behavioral data also show that equivalence classes resistant to disruption by the intermixing of conditional discriminations that are unrelated to each other. |
| |
| Where Does Stimulus Equivalence Occur in the Brain? Some Methodological Determinants of ‘Where’. |
| MICHAEL W. SCHLUND (Kennedy Krieger Institute, Johns Hopkins School of Medicine, and University of Pittsburgh), Michael F. Cataldo (Kennedy Krieger Institute) |
| Abstract: Matching-to-sample tasks (MTS) are commonly used to teach individuals with and without neurological dysfunction important skills. However, we have a limited understanding of the brain regions recruited during MTS tasks, especially when MTS involves stimulus equivalence (SE). Human neuroimaging and nonhuman neurophysiological studies on learning and memory suggest discriminating SE relations may recruit inferior frontal-subcortical and inferior parietal lobe regions. To investigate this issue, twelve adult subjects completed matching to sample (MTS) training designed to establish two three-member stimulus classes: A1®B1, B1®C1 and A2®B2, B2®C2. Next, conventional SE testing was performed while brain activation was measured using functional MRI. We will present results that show distinct activation patterns for each derived relation. However, we will emphasize that while such results highlight a potential role for frontal-subcortical and parietal regions in SE responding, activation attributable to MTS task variables (e.g., feedback, visuospatial features, saccadic eye movements, delayed stimulus control) cannot be disregarded. This point is illustrated by brain activation data showing how frontal and parietal activation is modulated by use of multiple comparison stimuli and the type of comparison condition used in imaging analyses. |
|
| |