Association for Behavior Analysis International

In the present symposium, three papers on additional measures within research on emergent relations will be presented. Steingrimsdottir and Arntzen present a paper on equivalence class formation in six healthy older adults and electroencephalography (EEG) measures. The preliminary results showed some correlation between the likelihood of formation of equivalence classes and the likelihood of being at the first stages of developing dementia, recorded by the EEG (see Figure S_A). In the second paper, Hansen and Arntzen present an experiment in which the was to explore the differential outcomes among eye-response topographies, when using head-mounted eye-tracking technology during the concurrent matching-to-sample formation of five 3-member as well as three 5-member stimulus equivalence classes, respectively. Collected data suggest different eye response pattern in the many-to-one training structure when compared to the one-to-many as well as linear series structures (see Figure H_A). In the third paper, Vie and Arntzen asked if a shift was observed when a one-to-many training structure is used, where each sample is matched to at least two comparisons in the training. The results, so far, does not show the large shift in talking from the last part of training to the test for emergent relations (see Figure V_A).

It has been suggested that studying responding in accordance with stimulus equivalence might provide important information about changes in cognition as a function of age. Importantly, such studies could also be important contribution to our understanding of the cognitive changes that happen when a person gets a neurocognitive disorder (NCD) such as Alzheimers disease. It has been difficult to identify those who are at the earliest stages of the disease and efficient methods have been called for. In that regard, it has been suggested that testing for responding in accordance to stimulus equivalence may be an important contribution when looking for those early signs. In this presentation, the results from conditional discrimination learning and testing for responding in accordance with stimulus equivalence in six healthy older adults (6070 years of age) were compared to electroencephalography (EEG) results that are particularly designed for diagnosing dementia. The preliminary results showed some correlation between the likelihood of responding in accordance with stimulus equivalence and the likelihood of being at the first stages of developing dementia, recorded by the EEG. The results are important contribution for further studies in this area.

Applying eye-tracking technology during Matching-to-Sample performance, additional dependent measures (i.e., observing behavior topographies) add to the evidence that observing behavior vary among different conditional discriminations procedures (e.g., Dube, Balsamo, Fowler, Dickson, Lombard, & Tomanari, 2006). Moreover, initial laboratory work suggests that dependent molecular variables, such as fixation duration, rate, and variability, contribute with supplementary knowledge on stimulus equivalence class formation when generated with different training structures (i.e., Hansen & Arntzen, 2013, October). In a systematic replication, therefore, the purpose of the following investigation is to explore the differential outcomes among eye-response topographies, when using head-mounted eye-tracking technology during the concurrent matching-to-sample formation of five 3-member as well as three 5-member stimulus equivalence classes, respectively. As of now, collected data suggest different eye response pattern in the many-to-one training structure when compared to the one-to-many as well as linear series structures. Hence, in our group design, we plan to recruit a total of 60 university-college students to participate. Furthermore, in accordance with Hansen & Arntzen (2013, October), we expect to note differential results in fixation variability when introducing a more solid (i.e., mixed) training format.

In a recent study on the use of a talk-aloud protocol by Vie and Arntzen (2013), showed that several participants had a large shift from almost no talking in the last part of training when the comparison array was present, to over 80 percent talking when the comparison array was present in the test for emergent relations. A many-to-one training structure was employed, and it was suggested that the finding was a result of the training structure. Hence, it was possible for the participants to predict the correct comparison since the samples were matched to one comparison only. The research question in the current study is if the same shift emerges when a one-to-many training structure is used, where each sample is matched to at least two comparisons in the training. The parameters are identical to Vie and Arntzen's (2013) study, except the training structure used. The results, so far, does not show the large shift in talking from the last part of training to the test for emergent relations.