Public defence: Jon Magnus Eilertsen

This thesis consists of three studies that investigate variables and procedures in translational research, based on stimulus equivalence technology and the transfer of stimulus function in stimulus equivalence research.

Study 1

Study 1 applied a stimulus tailoring procedure, participants rating six images of needle injections in different locations on a human hand. The images rated as most and least painful were later employed in a class expansion phase as stimuli D1 and D2 respectively.

Another image, D3, showed a human hand where the a cotton swab was exchanged with the hypodermic needle. Fifteen participants trained and tested for the formation of three three-member equivalence classes followed by D->A class expansion training.

After a new test for the emergence of three 4-member equivalence classes, participants were asked to pick one out of three water bottles that were identical except for the abstract B1, B2, and B3 stimuli attached to them. Finally, participants were asked to rate the abstract D stimuli on a similar scale they had initially rated the needle injection images with.

Results show that participants avoided picking the water bottle labeled with stimulus B1, equivalent to the image rated as most painful. In addition, there were no significant difference in how the participants rated the D1, D2 stimuli and the abstract B1 and B2 stimuli equivalent to them.

Study 2

Study 2 employed a tailoring procedure as a sorting task in a procedure based on stimulus equivalence technology. The sorting task was presented to participants before conditional discrimination training and testing.

Participants, in a stimulus equivalence technology based procedure, sorted stimulus cards with the written names of different food items into three carbohydrate ranges “less than 20”, “20–40” and, “more than 40”. An OTM training structure was used to train and test for the emergence of three 3-member (Conditions 1 and 2) and three 5-member equivalence classes (Condition 3).

During training, the carbohydrate ranges were the A stimuli and the different food items incorrectly sorted in served as comparison stimuli (B, C, D, and E).

Main findings were that out of 22 participants 21 of them responded in accordance with stimulus equivalence. In addition, post-class formation sorting tests found a 100 percent correspondence between MTS test and sorting tests.

Study 3

Study 3 employed delayed matching-to-sample (DMTS) procedures with either DMTS 3 s or DMTS 6 s delay for two experimental groups. Previous studies have found that training with DMTS 2 s increase the generalization of stimulus control properties of facial expressions to equivalent class members beyond the generalization derived from training with simultaneous MTS.

In Study 3, a control group rated three facial expressions (happy, neutral, and angry) and abstract stimuli on a Semantic Differential Scale. Participants in two trained conditional discriminations with the LS training structure using either DMTS 3 s or DMTS 6 s. The A1, A2, and A3 stimuli showed the angry, neutral, and happy facial expressions respectively, the B, C, D, E, and F stimuli were abstract.

Participants rated the abstract D stimuli after training, and testing for B->F transitivity and F->B equivalence relations. Finally, participants in the experimental groups were asked to rate the abstract D stimuli on a Semantic Differential Scale.

Results show that DMTS 3 s produces lower yield, but training with DMTS 6s requires more training trials. In addition, the generalization of stimulus control properties from the facial expressions (A stimuli) to the D stimuli is, however, more in line with how the control group rated the facial expressions for the DMTS 3 s group than for the DMTS 6 s group.

Participants furthermore overrated the positive valences of the abstract D3 stimuli equivalent to the happy faces in the DMTS 3 s group. No such over-rating was seen for the D1 (equivalent to angry faces) in the same group, or for D1 or D3 in the DMTS 6 s group.