Effects of Limited Hold on Equivalence Class Formation

In an attempt to limit the opportunity to engage in mediating behavior, two groups of adult participants received preliminary training in identity matching with limited hold levels (LH) for responding of 0.7 s for the sample and 1.2 s for the comparisons. The two groups were subsequently trained to form three 5-member classes, using the same LH levels, where the A, B, D, and E stimuli were abstract stimuli, and the C stimuli were meaningful pictures. In two tests for emergent relations, the LH for Group Short was unchanged, whereas 5 s were added to the LH for the comparisons for Group Long. None of the participants in Group Short responded in accordance with stimulus equivalence in either of the two tests. In Group Long, one participant responded in accordance with stimulus equivalence in the first test, and an additional eight participants formed equivalence classes in the second test.

Naming of Stimuli in Equivalence Class Formation in Children

In the present study, two typically developing 4-year-old children, Pete and Joe, were trained six conditional discriminations and tested for the formation of three 3-member equivalence classes. Pete and Joe did not establish the AC relation within 600 trials and were given two conditions of preliminary training, including naming of stimuli with two different stimulus sets. Pete started with preliminary training with common naming of stimuli, followed by conditional-discrimination training and testing for emergent relations, and continued with preliminary training on individual naming of stimuli, followed by the same training and testing as described previously. Joe experienced the same conditions but in reversed order. Pete responded in accordance with equivalence in the second round in the condition with common naming. In the first round of testing in the condition with individual naming, he responded in accordance with equivalence. In the condition with individual naming, Joe did not respond in accordance with stimulus equivalence but established all of the directly trained relations during training. In the condition with common naming, he responded in accordance with equivalence in the first round of testing. The results from the experiment support earlier findings that both common and individual naming could facilitate the emergence of equivalence classes.

Mediated Generalization and Stimulus Equivalence

From the 1930s to the 1970s a large number of experimental studies on mediated generalization were published, and this research tradition provided an important context for early research on stimulus equivalence. Mediated generalization and stimulus equivalence have several characteristics in common, notably that both traditions seek to experimentally investigate derived responding among arbitrarily related stimuli in human participants. Although studies of stimulus equivalence are currently being regularly published, few studies investigate mediated generalization in humans today, and the research tradition is mainly of historical interest. The current article will give an account of the origin, the development, and the demise of research on mediated generalization, including a presentation of publication trends, experimental methodology, and the conceptual context research on mediated generalization took place within. Finally, some thoughts on the demise of mediated generalization and its relevance for modern research on stimulus equivalence and other types of derived responding are presented, including reflections on the observability of explanatory variables and the use of inferential statistics.

Enhanced Equivalence Projective Simulation: A Framework for Modeling Formation of Stimulus Equivalence Classes

Formation of stimulus equivalence classes has been recently modeled through equivalence projective simulation (EPS), a modified version of a projective simulation (PS) learning agent. PS is endowed with an episodic memory that resembles the internal representation in the brain and the concept of cognitive maps. PS flexibility and interpretability enable the EPS model and, consequently the model we explore in this letter, to simulate a broad range of behaviors in matching-to-sample experiments. The episodic memory, the basis for agent decision making, is formed during the training phase. Derived relations in the EPS model that are not trained directly but can be established via the network's connections are computed on demand during the test phase trials by likelihood reasoning. In this letter, we investigate the formation of derived relations in the EPS model using network enhancement (NE), an iterative diffusion process, that yields an offline approach to the agent decision making at the testing phase. The NE process is applied after the training phase to denoise the memory network so that derived relations are formed in the memory network and retrieved during the testing phase. During the NE phase, indirect relations are enhanced, and the structure of episodic memory changes. This approach can also be interpreted as the agent's replay after the training phase, which is in line with recent findings in behavioral and neuroscience studies. In comparison with EPS, our model is able to model the formation of derived relations and other features such as the nodal effect in a more intrinsic manner. Decision making in the test phase is not an ad hoc computational method, but rather a retrieval and update process of the cached relations from the memory network based on the test trial. In order to study the role of parameters on agent performance, the proposed model is simulated and the results discussed through various experimental settings.