Classical Conditioning for Pain: The Development of a Customized Single-Case Experimental Design

Tumal Kumar De,Victoria J Madden

doi:10.36850/e8

Abstract

Single-case experiments are increasingly popular in the behavioral sciences. Due to their flexibility, single-case designs can be customized to test a variety of experimental hypotheses. We were interested in using a single-case experimental approach to test whether pain thresholds can be influenced by Pavlovian classical conditioning. Following the example of earlier studies into this topic, we planned to measure whether participants would more frequently report specific electrocutaneous stimuli as painful when they were presented with specific vibrotactile stimuli that had previously been associated with painful electrocutaneous stimuli. First, we decided on a mean difference effect size measure derived from the Sensation and Pain Rating Scale ratings for the electrocutaneous stimuli provided by the participants. Next, we discussed several possible single-case designs and evaluated their benefits and shortcomings. Then, we ran pilot tests with a few participants based on the possible single-case designs. We also conducted a simulation study to estimate the power of a randomization test to test our hypothesis using different values for effect size, number of participants, and number of measurements. Finally, we decided on a sequentially replicated AB phase design with 30 participants based on the results from the pilot tests and the power study. We plan to implement this single-case design in a future experiment to test our hypothesis

Highlights

Single-case experiments are increasingly popular in the behavioral sciences
In the differential classical conditioning design used here, one neutral cue is paired with a painful electrocutaneous stimulus, while another neutral cue is paired with a non-painful electrocutaneous stimulus
Number of participants: Since the pilot tests revealed a small effect size, and the number of measurements is limited, to ensure high power we considered a large range of participant count at 10, 20, 30, 40, and 50

Summary

Introduction

Single-case experiments are increasingly popular in the behavioral sciences Due to their flexibility, singlecase designs can be customized to test a variety of experimental hypotheses. Following the example of earlier studies into this topic, we planned to measure whether participants would more frequently report specific electrocutaneous stimuli as painful when they were presented with specific vibrotactile stimuli that had previously been associated with painful electrocutaneous stimuli. We ran pilot tests with a few participants based on the possible single-case designs. Our study design builds on two previous experiments (Madden et al, 2016; Traxler et al, 2019), which followed a Pavlovian or “classical” conditioning design (Pavlov, 1928) to test whether neutral but pain-associated cues can bias a participant’s decision about whether a stimulus is painful or nonpainful. Due to the before (conditioning) and after (conditioning) structure of multiple observations required for each participant, SCE designs are perfectly suited for testing this hypothesis

Methods

Results

Discussion

Conclusion