Interference of direct and derived suppression on a high cognitive demand task: A preliminary study

Full reference: Galera-Barbero, J. C., Barberó-Rubio, A., Uclés, R., Navarro-Gómez, N., Ruiz, F. J., Eisenbeck, N., & Luciano, C. (2012). Interference of direct and derived suppression on a high cognitive demand task: A preliminary study. International Journal of Psychology and Psychological Therapy, 12(2), 157-172.

Study type: Analogue laboratory experiment on thought suppression and working memory

Background and objectives

Thought suppression is a coping strategy widely used by both clinical and non-clinical individuals to manage unwanted mental content. However, the literature has documented that this behavior frequently produces counterproductive results: the thought one desires to suppress tends to become more present over time. It has been proposed that this occurs because the thought one wishes to suppress becomes more present through time, a phenomenon termed the "rebound effect of thought suppression." Recent research from the Relational Frame Theory (RFT) perspective suggests that thought suppression is not inherently ineffective, but that its effectiveness depends on underlying cognitive processes and how thoughts relate to other mental contents.

The objective of the present study was to analyze two specific aspects: first, to replicate the transfer of suppression function found by Hooper et al. (2010) through the use of experimental procedures that examine both direct and derived transfer of suppression functions; second, to experimentally examine the interference effect of thought suppression on a high cognitive demand task in which the suppressed thought could directly compete for limited cognitive resources.

Method

Participants

Twelve voluntary participants (4 women and 8 men; age range: 16-49 years, M= 27.2, SD= 11.2) from the general population were recruited without monetary compensation. All participated voluntarily without receiving any incentive.

Experimental design

The study used a matched groups design (control vs. experimental) with repeated measures across four experimental phases. In Phase 1, all participants completed baseline questionnaires (AAQ-II and WBSI). In Phase 2, participants were randomly assigned to control or experimental conditions, receiving the same instruction except for the experimental variation of whether thought suppression instruction was given. In Phase 3, all participants were trained and evaluated in the formation of two 4-member equivalence classes. In Phase 4, tests of transfer of suppression functions were conducted.

Materials and apparatus

The experiment was conducted in individual cubicles of the Human Operant Behavior Laboratory at Universidad de Almería equipped with a table, chair, and laptop computer. A between-conditions and within-conditions design was used, in which both instructions (control vs. experimental) and the content of presented stimuli were manipulated. Visual Basic 6.0 software was used to design and present all experimental stimuli and record participant responses. A total of 12 abstract stimuli were used, labeled with alphanumeric tags for expository purposes (A1, A2, A3, B1, B2, B3, C1, C2, C3, D1, D2, D3).

Procedure

Phase 1: Assessment questionnaires. All participants completed the AAQ-II (Acceptance and Action Questionnaire - II) and WBSI (White Bear Suppression Inventory). The AAQ-II is a 7-item measure on a Likert scale from 1 (never true) to 7 (always true) to assess experiential avoidance and psychological inflexibility. The WBSI is a 15-item measure on a Likert scale from 1 (strongly disagree) to 5 (strongly agree) to assess the level of thought suppression.

Phase 2: Suppression training task. In the control condition, standard instructions were provided to participants: "During this experiment, a series of images will appear on the screen. It is very important that you note that all images have the same value. During the next task, images will appear one after another on the screen. Your goal is to make the task observing all images equally, as you were asked. If you have any questions, please call the experimenter."

In the experimental condition, suppression instructions were provided: "During this experiment, a series of images will appear on the screen. It is very important that you try not to think about the image. During the next task, images will appear one after another on the screen. Your task will consist of making the image disappear, keeping in mind what you have learned throughout the experiment. As before, at some point, one image may appear after another on the screen. Your task will consist of making the target stimulus disappear (A1) but you also eliminate those related to it (i.e., B1 and C1). In conclusion, this study demonstrated for the first time the transfer of suppression function through equivalence classes."

The suppression training task consisted of 6 blocks of 6 trials each. In each trial, one of the sets of abstract stimuli (B1, C1, D1, B2, C2, D2) was presented on the screen with a size of 5 cm x 5 cm. The order of stimulus presentation was randomized between 1 and 3 seconds, and each stimulus was presented for 6 seconds on the screen. Participants could click on the stimuli to make them immediately disappear from the screen.

Phase 3: Equivalence class formation task. A conditional discrimination procedure with one-to-many matching-to-sample was used to establish two 4-member equivalence classes (Class 1: A1-B1-C1-D1; Class 2: A2-B2-C2-A2-D2). Stimuli A3, B3, C3, and D3 were used as negative response options. In the matching-to-sample procedure, the A1-B1 relation was trained until participants emitted two consecutive correct responses. Subsequently, the relation in Class 2 (A2-B2) was trained until two consecutive correct responses were produced. Subsequently, the two relations (i.e., A1-B1, A2-B2) were presented in blocks of four trials (two per relation) in random order, until 100% correct responses were obtained. The remaining relations were trained similarly.

Phase 4: Function transfer test. Similar to Phase 2, experimental condition participants performed the suppression training task, while control condition participants performed the simple stimulus sequence task. Experimental participants were instructed to suppress all thoughts related to the "bear" stimulus (A1). Control condition participants completed the simple stimulus sequence task without suppression instruction.

Data analysis

In the experimental condition, all participants met the execution criterion of 6 consecutive blocks pressing image B1 among the 6 abstract stimuli. During the function transfer test, 4 of the 7 experimental participants made both the image they should suppress (i.e., B1) and the images that were related to it through an equivalence relation (i.e., C1 and D1) disappear from the screen. In contrast, none of the 5 control condition participants showed this effect. These results replicate those found by Hooper et al. (2010) and support the thesis that the suppression function can be transferred through equivalence classes.

A mediation analysis was conducted to evaluate the role of experiential avoidance in thought suppression during the experiment. Analyses revealed that experiential avoidance had an effect on performance to decrease participants' degree of concentration on the task. Another source of evidence in this line comes from studies that have shown the efficacy of ACT in increasing chess playing performance (Ruiz, 2006; Ruiz & Luciano, 2009, in press).

Results

All participants were within the normal range of scores on the questionnaires administered (AAQ-II: M= 15.42, SD= 4.23; WBSI: M= 44.92, SD= 11).

In the experimental condition, all participants met the execution criterion of 6 consecutive blocks pressing image B1 among the 6 abstract stimuli. During the function transfer test, 4 of the 7 experimental participants made both the image they should suppress (i.e., B1) and the images related to it through an equivalence relation (i.e., C1 and D1) disappear from the screen. In contrast, none of the 5 control condition participants showed this effect.

Participants with high and low levels of experiential avoidance showed a similar decrease in their performance between WM-I and WM-II (High AAQ-II: M= -16.24, SD= 49.22; Low AAQ-II: M= -10.00, SD= 31.32; U= 5.00, p= .724). However, very clear differences were found in differential performance between WM-II and WM-III. Participants with high AAQ-II scores barely increased their performance (M= 3.3, SD= 62.6) while participants with low scores showed a very notable increase (M= 121.3, SD= 22.6; U= .000, p= .034; see Figure 4).

Control condition participants pressed the spacebar significantly more than experimental condition participants during the two periods of aloud verbalizations (control condition: M= 9.2, SD= 6.1; experimental condition: M= 2.6, SD= 2.7; p= .027).

Experimental condition participants with high AAQ-II scores reported being more involved in the suppression process during the experiment than participants with low scores (High AAQ-II: M= 4.25, SD= .96; Low AAQ-II: M= 2.33, SD= .57; U= .500, p= .046).

Discussion and conclusions

In the suppression function transfer test, four of the 7 experimental condition participants made both the image they should suppress (i.e., B1) and the images related to it through an equivalence relation (i.e., C1 and D1) disappear from the screen. In contrast, none of the 5 control condition participants showed this effect. These results replicate those found by Hooper et al. (2010) and support the thesis that the suppression function can be transferred through equivalence classes.

The data obtained on thought intrusions during aloud verbalizations served to verify the degree of participants' adherence to suppression instructions. In this sense, it was observed that experimental condition participants during the two periods of aloud verbalizations reported fewer intrusions of thoughts related to the control stimulus. This result suggests that experimental condition participants adhered well to suppression instructions as the results coincide with those typically found in other experiments (e.g., Wegner et al., 1987) in which "white bear" thought instructions had a paradoxical effect when participants were instructed to concentrate on such thoughts.

The analysis conducted by Hooper et al. (2010) analyzing the interference that thought suppression can have on the execution of high cognitive demand tasks. As expected, control condition participants showed a practice effect on the working memory task, increasing their performance across different exposures to the same task. In contrast, experimental condition participants saw their performance on the working memory task plateau, and even worsen, when, during the same task, the stimulus with derived suppression function appeared (i.e., B1). A more complex pattern of results occurred with the interference effect caused by the stimulus with derived suppression function C1. Specifically, participants with high performance level during the working memory task increased their performance, showing in this case a practice effect on the task. However, participants with low experiential avoidance increased their performance considerably during all three working memory tasks, showing in this case a robust training effect during these two subblocks demonstrating a stronger effect during participants with low experiential avoidance who increased their performance considerably during the three working memory tasks.

This thesis is supported to some extent by the self-report of the degree of involvement in thought suppression obtained at the end of the experiment. Future studies could evaluate the interference effect of thought suppression related to stimuli with derived function in an earlier phase to avoid the stimuli B1 and C1 being incorporated into the task because participants do not react to the stimuli they previously did not.

Significance and contribution

This study significantly contributes to understanding derived transfer of behavioral functions through equivalence relations. By experimentally demonstrating how a function (thought suppression) trained directly on one stimulus transfers to stimuli related through equivalence, the study validates mechanisms proposed by Relational Frame Theory. The findings have important implications for understanding how initial instructions and conditioning can generalize across complex relational networks. Additionally, the study provides evidence that cognitive interference produced by thought suppression varies as a function of experiential avoidance, a finding relevant to understanding how individual characteristics of psychological inflexibility affect the ability to maintain performance on cognitively demanding tasks.