Promoting Psychological Flexibility on Tolerance Tasks: Framing Behavior Through Deictic/Hierarchical Relations and Specifying Augmental Functions

Background and objectives

Acceptance and Commitment Therapy (ACT) has gained considerable empirical support for treating diverse psychological disorders. However, a central limitation in advancing more effective ACT interventions has been the lack of rigorous analysis, from the perspective of Relational Frame Theory (RFT), of the specific mechanisms by which cognitive defusion exercises and self-based exercises operate. RFT conceptualizes psychological flexibility as a generalized repertoire of framing ongoing behavior in hierarchy with the "deictic I" (the observer), which typically reduces the discriminative functions of ongoing behavior and allows derivation of rules specifying appetitive augmental functions (functions that regulate and motivate conduct).

The authors distinguish between two types of defusion exercises: those that only frame aversive private experiences through simple deictic relations (the observer and experience are separated spatially and temporally) versus those combining deictic relations with explicit hierarchical relations (the self contains or is above the behavior) and specifying appetitive augmental functions (providing regulatory functions to that discrimination). Previous research with adolescents and analogue studies suggested that protocols including hierarchical relations and regulatory functions are more effective than those including only deictic relations, but these conclusions relied on self-report measures of frequency of problematic behaviors or experienced distress. The current study aimed to replicate and extend these findings using direct behavioral measures of tolerance to experimentally induced discomfort, providing a stronger test of changes in the discriminative functions of discomfort.

Method

Participants

Thirty-one adult volunteers were recruited. One participant was excluded for reaching maximum pain tolerance during the cold pressor pretest phase (300-second maximum allowed for ethical reasons). The final sample consisted of 30 participants (15 men, 15 women) aged 21-46 years (M = 26.8 years, SD = 5.5). Participants were randomly assigned to one of three experimental conditions with the only restriction of balancing sex (5 men and 5 women per condition). No specific inclusion/exclusion criteria were reported, though the initial evaluation included informed consent screening to exclude participants with medical history incompatible with the cold pressor task (cardiac or circulation problems such as Raynaud's syndrome, blood pressure problems, diabetes, epilepsy, or recent serious injury).

Design

The study employed a randomized independent-groups experimental design (simple randomization, non-blinded). The independent variable was the protocol to which participants were assigned: (a) control (general interview without specific intervention), (b) Defusion I protocol (framing ongoing private experiences through deictic relations), or (c) Defusion II protocol (framing experiences through deictic and hierarchical relations plus specification of appetitive augmental functions). Primary dependent variables were tolerance in time (seconds) on two discomfort tasks (cold pressor and aversive film) measured at pretest and posttest. Secondary dependent variables were self-reports of perceived pain/discomfort in both tasks.

Intervention / Conditions

Control condition (n = 10): Participants received a general interview with the experimenter unrelated to the experimental tasks, including questions about their studies, work, interests, and future plans. The interview lasted approximately 30 minutes.

Defusion protocols (n = 10 per condition, approximately 20-30 minutes total duration): Both defusion conditions were administered in two parts by an ACT-trained experimenter.

In Part I, multiple-exemplar training used neutral private events. The experimenter guided participants through a structured sequence: focus on breathing, observation of comfortable and uncomfortable posture, and observation of thoughts (imagined as written on balloons). At each step, the key deictic question was asked: "Who observes/contemplates this?" In the Defusion II condition, hierarchical instructions were added: "Allow yourself to be much bigger than all these thoughts that are here with you... realize that you have enough room to have whatever thoughts might show up."

In Part II, the procedure was repeated but specifically focused on aversive private events experienced during the pretest tasks (cold water and film). Participants were asked to re-imagine the sensations from the cold pressor task (pin-pricks), to visualize the pain in their hand as if taking a photograph of it. The deictic question was: "Who is contemplating that picture of pain in your hand?" In the Defusion II condition, hierarchical questions were added: "Imagine you let the pain be in charge of what you do... Now imagine that you are in charge of the situation... Would you be bigger than your pain in that case?"

Outcome measures

Behavioral/tolerance measures (primary dependent variables):

• Cold pressor task: Participants submerged their right arm up to the elbow in ice water at 4°C. Instructed to keep the arm in water as long as possible, but reminded they were free to withdraw at any time. Tolerance measured as total time in seconds (maximum 300 seconds for ethical reasons).
• Aversive film task: Participants viewed a 90-second film containing images of surgical procedures (limb amputation) without audio. Instructed to continue watching as long as possible, but free to interrupt. Tolerance measured as total time in seconds of continued viewing.

Self-report discomfort measures:

• After each experimental task, participants responded to Visual Analogue Scales (VAS) indicating a score from 0 (no pain/discomfort) to 100 (unbearable pain/discomfort).

Baseline measures (completed at pretest):

• Acceptance and Action Questionnaire-II (AAQ-II): General measure of experiential avoidance. 7 items, 7-point Likert scale (7 = always true; 1 = never true). Spanish version with adequate psychometric properties (α = .88).
• Cognitive Fusion Questionnaire (CFQ): Measure of cognitive fusion. 7 items, 7-point Likert scale (7 = always true; 1 = never true). Spanish version with adequate psychometric properties (α = .91).
• Depression, Anxiety, and Stress Scales-21 (DASS-21): Measure of negative emotional symptoms. 21 items, 4-point Likert scale (3 = applies to me very much or most of the time; 0 = does not apply to me at all). Contains three subscales: Depression, Anxiety, Stress. Spanish version with adequate psychometric properties (α = .90 for complete scale).

Data analysis

Analysis of variance (ANOVAs) was conducted to explore equivalence of experimental conditions on age, pretest tolerance, and pretest pain/discomfort intensity. Subsequent ANOVAs analyzed change scores (posttest - pretest) to examine differential protocol effects. Effect sizes (eta-squared, η²) were reported and interpreted as small (.01), medium (.06), and large (.14). When equal variances assumption was violated, Welch's robust test was used. Planned one-tailed contrasts (p < .05) tested whether: (a) change scores in Defusion II were greater than in Defusion I and control, and (b) change scores in Defusion I were greater than control. Between-condition effect sizes at posttest were calculated using Cohen's d (interpreted as small: .20-.49, medium: .50-.79, large: > .80). Moderation analyses used nonparametric bootstrapping (PROCESS package; 20,000 bootstrapped samples, 95% bias-corrected confidence intervals) to examine whether cognitive fusion (CFQ), age, and sex moderated protocol effects.

Results

Initial equivalence of experimental conditions: Baseline ANOVAs revealed no significant group differences in age [F(2, 27) = .81, p = .46], pretest cold pressor tolerance [F(2, 27) = .98, p = .39], pretest film tolerance [F(2, 27) = .57, p = .57], cold pressor pain intensity [F(2, 27) = .39, p = .68], or film discomfort intensity [F(2, 27) = .66, p = .53]. No differences appeared on AAQ-II, CFQ, DASS-21, or percentage viewing the complete film (63%: 6, 6, and 7 for control, Defusion I, and Defusion II respectively).

Tolerance - Cold pressor task: Statistically significant differences emerged between conditions on pretest-posttest change [F(2, 16) = 4.94, p < .05, η² = .23]. Planned contrasts revealed that Defusion II participants significantly increased tolerance compared to both Defusion I [t = 2.032, p < .05, d = .91] and control [t = 2.593, p = .02, d = 1.16]. Similarly, Defusion I participants significantly increased tolerance compared to control [t = 2.25, p < .05, d = 1.01].

Tolerance - Aversive film task: Differences between conditions were marginally significant although with almost large effect size [F(2, 27) = 3.21, p = .06, η² = .132]. Planned contrasts showed Defusion II significantly increased tolerance compared to control [t = 2.50, p < .05, d = 1.19]. Differences between Defusion II and Defusion I were marginally significant favoring Defusion II [t = 1.59, p = .06, d = .65]. No significant differences emerged between Defusion I and control [t = .91, p > .05, d = .43].

Perceived pain/discomfort: No statistically significant differences between conditions appeared on pretest-posttest change in perceived pain during cold pressor [F(2, 27) = .10, p > .05, η² = .01] or discomfort during film [F(2, 27) = 2.09, p > .05, η² = .07]. This finding is notable: defusion protocols produced changes in tolerance without corresponding changes in perceived intensity of discomfort.

Moderation analyses: Cognitive fusion as measured by CFQ was a statistically significant moderator of protocol effects on aversive film change (incremental R² = .20, p = .007) and marginally significant for cold pressor (incremental R² = .07, p = .09). Figure 3 shows that differences between conditions were greater as a function of participants' cognitive fusion levels. As expected, age and sex were not significant moderators (all p > .05).

Discussion and conclusions

Results confirm and extend previous research suggesting that defusion protocols including explicit hierarchical relations and specified regulatory functions are more effective than those including only simple deictic relations. The key finding was that Defusion II (hierarchical framing + augmental function specification) produced greater increase in discomfort tolerance than Defusion I (deictic framing only), in both the cold pressor and aversive film tasks. Defusion I was also superior to control on the cold pressor, though not on the film.

A particularly important aspect of the findings was that increases in tolerance occurred without corresponding changes in perceived pain/discomfort intensity. This supports the RFT conceptualization that the change mechanism in defusion exercises is alteration of the discriminative functions of discomfort (reducing its avoidance function) rather than reduction of discomfort itself. In other words, the protocols promoted more flexible responding (greater psychological flexibility) to discomfort, enabling participants to tolerate aversive experiences while remaining engaged with the task. This pattern is consistent with previous ACT research finding that participants continued pain tasks despite reporting high pain levels.

The authors acknowledge several limitations. First, while separating deictic from hierarchical cues, future research could further separate hierarchical cues from questions providing regulatory functions, clarifying which specific interactions drive behavioral change. Second, the aversive film task was not standardized, and many participants reached maximum tolerance at both pretest and posttest, suggesting the task did not produce relevant discomfort for those participants. Third, the study was limited to a non-clinical young adult sample in a laboratory context, limiting generalizability to clinical populations.

Despite these limitations, the authors emphasize this is the first study analyzing effects of different types of relational framing on tolerance to experimentally induced discomfort using direct behavioral measures. Results have practical implications for clinical applications where increasing discomfort tolerance is a therapeutic aim, such as chronic pain or individuals undergoing physiotherapy.

Significance and contribution

This study makes three significant contributions to ACT-RFT research. First, it provides empirical evidence that the RFT conceptualization of psychological flexibility (as hierarchical framing of ongoing behavior with the deictic I, plus specification of regulatory functions) has operational validity in producing more effective change mechanisms than simpler approaches. Second, it is one of few studies using direct behavioral measures (temporal tolerance on discomfort tasks) rather than self-report measures to evaluate defusion intervention effectiveness, providing a more rigorous test of changes in discriminative functions of aversive private experiences. Third, the cognitive fusion moderation finding suggests these defusion exercises may be particularly effective for individuals showing behavioral patterns more fused with aversive stimuli.

---