Naming of Stimuli in Equivalence Class Formation in Children

Number of Training Trials

The number of training trials decreased from the first to the second condition. This effect was most pronounced for Joe, starting with the individual naming condition. It is important to emphasize that the individual naming procedure was more time consuming because nine individual names were trained, compared to the common naming procedure, where only three names were trained.

Equivalence Class Formation

The results show that two 4-year-old children did not meet the experimenter-defined criterion in training of the conditional discriminations. However, they established the conditional discriminations after been trained in naming the stimuli. Hence, both Pete and Joe responded in accordance with stimulus equivalence in one or two of the conditions with common or individual naming. The results from the present experiment give some support to previous research that found that naming of stimuli can facilitate equivalence class formation (Bentall et al., 1993; Carp & Petursdottir, 2015; Eikeseth & Smith, 1992; Jennings & Miguel, 2017; Ma et al., 2016; Petursdottir et al., 2015).

Possible Effect of Order

There were differences in responding in accordance with equivalence depending on the naming conditions—common or individual naming—the children were exposed to. Pete, who was exposed to common naming first, did respond in accordance with equivalence in Test 2 and had less than half of the training trials Joe had in the individual naming condition. Pete also responded in accordance with equivalence in Test 1 in the individual condition. It could be that exposure to the common naming condition first made the individual naming condition easier. Joe, who was exposed to individual naming first, did not have intact baseline performance during testing and did not respond in accordance with symmetry and transitivity/equivalence in this condition. However, when Joe was exposed to the common naming of stimuli, he responded in accordance with stimulus equivalence in the first test. The number of training trials was also reduced by 50% in this condition compared to the individual condition; that is, establishing the necessary conditional discriminations demanded fewer training trials.

Naming as a Facilitator for Equivalence Class Formation

Earlier studies support the findings in this article showing that naming could facilitate equivalence class formation. Given the possible effect of order as discussed previously, one interpretation of the data is that common naming facilitated equivalence class formation slightly better than individual naming. In common naming, two similar names are linked together (e.g., FAB-FAB), whereas in individual naming, different names are linked together (e.g., FAB-SEM). Individual naming can facilitate equivalence class formation when the children produce sequences of names (e.g., FAB-SEM-MIP) that are established, and then the child repeats this chain of names from the presentation of a sample to the choice of a comparison, establishing an intraverbal relation between tacted stimuli (Lowe & Beasty, 1987; Ma et al., 2016; Miguel, 2016).

Training a common response to three or more stimuli in a class can establish a stimulus-response-stimulus chain. Saunders (1989) argued that a stimulus-response-stimulus chain is established when using common names, and that the defining characteristics of an equivalence class have not been met. A stimulus-response-stimulus chain can be established when responding to a comparison is controlled by the same topography (here, the same name) as that which is evoked by the sample. In the present study, all of the members in one class were named “SEM.” Therefore, if a stimulus-response-stimulus chain had been established, the comparison would control the response “SEM.” When the three comparisons are presented, the response “SEM” controls the choice of the comparison with the same name, even if the relation is not directly trained. It is essential to emphasize that both Joe and Pete produced incorrect responses during conditional-discrimination training in the common condition, which could indicate that chains were not established just because of the naming training. Pete did not respond in accordance with equivalence before the second test in the common naming condition. We argue that these results suggest that the relations were not established beforehand. Furthermore, when the stimulus names were trained, they were not trained as an experimenter-defined class. The A stimuli were trained separately, as were B and C stimuli, and then mixed. However, if the stimuli had been taught in classes—that is, A1, B1, and C1; A2, B2, and C2; and A3, B3, and C3—then one could argue that the classes were established beforehand.

Sorting of Stimuli

A stimulus-sorting test is a quick way to test for class partitioning. Additionally, stimuli sorting after the MTS test for emergent relations has been shown to correlate with the results from the test. In some cases, the delayed emergence of equivalence classes is also revealed (Arntzen et al., 2017; Arntzen, Norbom, & Fields, 2015; Fields, Arntzen, & Moksness, 2014). In the present study, neither child sorted the stimuli into the correct classes before training and testing. In the condition with individual naming, Joe did not respond in accordance with stimulus equivalence during the test, but he sorted the cards after the MTS test. This finding may be an example of delayed emergence, but there are some limitations in interpreting this as an example of delayed emergence. A sorting test with stimulus cards differs from an equivalence test in an MTS format because the participant has all the stimuli available at the same time and can scan back and forth among the stimuli to figure out which ones belong together. However, the distribution of correct and incorrect responses over time indicated that there was not a delayed emergence during the MTS test. For example, in a study by Spradlin, Cotter, and Baxley (1973), delayed emergence relations appeared after repeated testing, and when a participant does not respond in accordance with equivalence after naming the stimuli, repeated testing should be conducted to control for this.

Joe emitted several systematic errors during Test 1 in the individual naming condition, which implies that classes other than those defined by the experimenter were established (Arntzen, Nartey, & Fields, 2015; Eilifsen & Arntzen, 2009; Mensah & Arntzen, 2017). During Test 2, these errors were not systematic. Although he did not respond in accordance with equivalence, the participant-defined classes were no longer consistent.

Limitations and Further Research

There are several limitations to the present study. First, the sessions in the present experiment were conducted without controlling for session length or the number of trials per day/session. The arrangement in the present experiment was done because the children should decide how long they wanted to participate per session. However, it is important to emphasize that the length of the sessions varied, but this variation did not seem to influence the results.

Second, we did not check the effect of the programmed consequences for the touching responses in the MTS tasks directly. However, the words used as the programmed consequences in this experiment had shown increasing or decreasing effects on behavior in other contexts. The words were spoken out loud by the experimenter, with an enthusiastic (for correct responses) or neutral (for incorrect responses) voice. There is a possibility that the programmed consequences alone had a limited reinforcing effect on establishing the baseline relations. If a token system with items like preferred toys and games had been used contingent on correct responses, the baseline relations probably could have been established faster. Such an arrangement was not employed due to restrictions from the Norwegian Center for Research Data that required that the children receive the same amount of benefits.

Third, the number of trials was quite large in the conditional-discrimination training. Additionally, research with adults using an MTO training structure with abstract shapes has sometimes demonstrated a high number of training trials (Arntzen & Hansen, 2011). Other experiments with older children employing a preliminary training phase have the present study that the number of trials to mastery was lower than in the previous study (e.g., Pilgrim, Jackson, & Galizio, 2000; Smeets & Barnes-Holmes, 2005). Researchers should consider conducting preliminary training to familiarize the children with the task, reducing the large number of errors at the start of the MTS training. Such preliminary training could be identity matching with colors.

Both the short session length and the frequent breaks (after every fifth trial) could have hindered the flow in the establishment of the conditional discriminations. Furthermore, these variables could be responsible for the large number of trials in the present study. It is challenging to arrange procedures as in the present study with 4-year-old children without having short sessions and frequent breaks. Additionally, we had to consider that the children should not be excluded from other activities in the kindergarten.

Fourth, the effect of the naming conditions could be confounded. To control for the effect of the MTS procedure alone, extended sessions with extended training and testing could be used to study the effect of repeated exposure to the procedure. An effect of order should be observed as a reduction of the number of trials to meet the criterion in the second condition. Such a reduction in the number of training trials was observed especially for Joe.

Finally, only two children served as participants in the experiment, and replications should be conducted to control for the differences demonstrated in the present experiment.

Another important line of research is the effect on stimulus class formation of having children make up names themselves, rather than using experimenter-defined names as in the present study.

Summary

Neither Pete nor Joe established the necessary conditional discriminations without any training on naming the stimuli. Pete, who started with common naming, responded in accordance with equivalence in the second test in the first condition and the first test in the individual naming condition. Joe, who started with the individual naming condition, did not respond in accordance with equivalence but did so after the common naming condition.

留言 (0)

沒有登入
gif