The Eye Movement Desensitization and Reprocessing Procedure Prevents Defensive Processing in Health Persuasion

Abstract

In the present study, the method of eye movement desensitization and reprocessing (EMDR) is studied to understand and prevent defensive reactions with regard to a negatively framed message advocating fruit and vegetable consumption. EMDR has been shown to tax the working memory. Participants from a university sample (n = 124) listened to the persuasive message in a randomized laboratory experiment. In the EMDR condition, they were also instructed to follow with their eyes a dot on the computer screen. The dot constantly moved from one side of the screen to the other in 2 seconds. In addition, a self-affirmation procedure was applied in half of the participants. EMDR led to a significant increase in persuasion, only in recipients in whom the persuasive message could be expected to activate defensive self-regulation (in participants with a moderate health value and in participants with low self-esteem). In those with a moderate health value, EMDR increased persuasion, but only when recipients were not affirmed. In addition, EMDR increased persuasion only in recipients with low self-esteem, not in those with high self-esteem. These results showed that EMDR influenced persuasion and in some way lowered defensive reactions. The similarities and differences in effects of EMDR and self-affirmation further increased our insight into the psychology of defensiveness.

Notes

¹ The present moderation analyses were chosen over a sample split for two reasons. First, in the case of a sample split on a continuous measure (health value), an arbitrary cutoff point would be created that depends on the distribution of scores in the specific sample. Second, by applying a sample split the number of participants is reduced and mostly unevenly distributed over a low and a high group. This lowers statistical power of statistical tests in both groups and may result in differences in statistical power of tests between the groups. Both effects are avoided by the present moderation analyses.

² To check to what extent the modeled data reflected the actual data, the analyses were also conducted using the sample split procedure, dichotomizing the health value measure, with a moderate group (n = 49) and a high group (n = 75). The pattern of means was the same as in the modeled data. In addition, the three-way interaction was also significant, F(1, 114) = 6.93, p = .01, η² = .057, and only the two-way interaction in the moderate health value group was significant, F(1, 43) = 9.01, p < .01, η² = .173, as was expected.

³ To check to what extent the modeled data reflected the actual data, the analyses were also conducted using the sample split procedure, dichotomizing the self-esteem measure, with a low group (n = 58) and a high group (n = 66). The pattern of means was the same as in the modeled data. In addition, the two-way interaction between EMDR and dichotomized self-esteem was also significant, F(1, 118) = 7.98, p < .01, η² = .063.