Measuring experiential avoidance and psychological inflexibility: The Spanish version of the Acceptance and Action Questionnaire-II

Full reference: Ruiz, F. J., Langer Herrera, Á. I., Luciano, C., Cangas, A. J., & Beltrán, I. (2013). Measuring experiential avoidance and psychological inflexibility: The Spanish version of the Acceptance and Action Questionnaire - II. Psicothema, 25(1), 123-129. doi: 10.7334/psicothema2011.239

Study type: Cross-cultural validation and psychometric adaptation

Background and objectives

Experiential avoidance and psychological inflexibility are central constructs in Acceptance and Commitment Therapy (ACT) and have been shown to be associated with a wide range of psychological disorders and quality-of-life problems. In recent years, research has documented that these processes maintain psychopathology across multiple clinical conditions. The Acceptance and Action Questionnaire (AAQ) has been the most widely used measure to assess these constructs, although the original version presented limitations in internal consistency and factor structure. The AAQ-II was recently developed to overcome these limitations, demonstrating superior psychometric properties to the first version while maintaining similar levels of external validity.

The objective of the present study was to analyze the functioning of the Spanish translation of the AAQ-II in six independent samples from recent Spanish studies. It was hypothesized that this Spanish version would present psychometric properties, factor structure, and validity data similar to the original scale.

Method

Participants

Six independent samples totaling 712 participants were analyzed. Sample 1 consisted of 51 patients (48% female; age range: 20-63 years, M= 35.9, SD= 10.9) with emotional and psychotic disorders, all receiving treatment at public mental health services in Almería province. Sample 2 included 253 undergraduate students from Universidad de Almería (62% female; age range: 17-64 years, M= 21.9, SD= 5.9). Sample 3 consisted of 122 teaching students (69% female; M= 21.4, SD= 4.4). Sample 4 included 132 participants recruited from Universidad de Almería students and their relatives (63% female; M= 33.56, SD= 12.88). Sample 5 included 119 undergraduate students (68% female; age range: 18-30 years) who volunteered in a study on transfer of functions. Sample 6 comprised 35 patients (57% female; M= 44 years, SD= 12.7) with emotional and/or psychotic disorders from a public mental health center in Granada.

Instrument(s) under study

The Acceptance and Action Questionnaire - II (AAQ-II; Bond et al., 2011) is a 7-item measure that assesses general experiential avoidance and psychological inflexibility. Each item is responded to on a 7-point Likert scale. The items reflect both unwillingness to experience unwanted emotions and thoughts (for example, "I am afraid of my feelings") and inability to be in the present moment and behave according to values when challenging psychological events are experienced (for example, "My painful experiences and memories make it difficult for me to live a life that I would value"). The Spanish translation was conducted by authors Ruiz and Luciano (2009) and subsequently revised and adjusted for the present study.

Other outcome measures

Multiple instruments were used to assess the validity of the AAQ-II. The SCL-90-R (Symptom Checklist-90 - Revised; Derogatis, 1994) was administered to Samples 1 and 2 to assess general psychological symptoms. The Beck Depression Inventory - Short Form (BDI-SF) and the Brief Metacognitions Questionnaire (MCQ-30) were administered to Sample 3. Sample 4 completed the Penn State Worry Questionnaire (PSWQ), the Anxiety Sensitivity Index (ASI), the General Self-Efficacy Scale (GSC), the Kentucky Inventory of Mindfulness Skills - Reduced (KIMS-R), and the Coping Strategies Inventory (CSI). Sample 6 completed the Short-Form Health Survey (SF-36).

Data analysis

Pearson product-moment correlations between each item and total score were computed, as well as internal consistency analysis via Cronbach's alpha for each sample and for the total dataset. Kaiser-Meyer-Olkin indices and Bartlett's sphericity test were computed for each sample to determine adequacy for factor analysis. Exploratory factor analyses were conducted via principal component analysis for each sample and for the total dataset. Mean scores between clinical and non-clinical samples were compared using Student's t-test, and Pearson correlations between AAQ-II scores and other instruments were calculated.

Results

Item analysis showed correlations between each item and total score ranging from .58 (item 6) to .75 (item 4). The overall Cronbach's alpha for the scale was .88 (range across samples: .75 for Sample 1 to .93 for Sample 6). All samples were adequate for factor analysis according to Kaiser-Meyer-Olkin indices (Samples 1-6: .75, .85, .73, .84, .78, .82 respectively) and Bartlett's sphericity test, which was significant in all cases.

Exploratory factor analysis revealed one main factor in all samples, although Samples 1 and 3 showed a second small factor according to Kaiser criterion. When factor analysis was conducted with the total dataset (Kaiser-Meyer-Olkin index: .87; Bartlett's chi-square: 2312.05, p < .0001), a clear one-factor solution emerged. This factor explained 57.33% of the variance in item scores, while the second and third factors explained 10.88% and 9.11% respectively. Item saturations with the main factor ranged from .69 to .84. Scores in clinical samples (1 and 6) were significantly higher (overall M= 32.64, SD= 9.12) than in non-clinical samples (overall M= 21.22, SD= 7.76; t= 12.69, p < .0001). Women obtained significantly higher scores than men in three samples (Sample 1: U= 218, p= .028; Sample 2: t= -3.57, p < .001; Sample 4: t= -2.33, p= .022).

Regarding external validity, AAQ-II scores correlated significantly with all psychological symptoms assessed by the SCL-90-R and Global Severity Index (range: r= .55 in Sample 2 to r= .58 in Sample 1). In Sample 3, AAQ-II scores correlated with depressive symptoms (r= .45) and negative metacognitive beliefs (r= .46 with uncontrollability, r= .42 with need to control thoughts). In Sample 4, correlations were found with worry (r= .56), anxiety sensitivity (r= .43), negative self-efficacy (r= -.40), adequate coping with problems (r= -.46), and KIMS acceptance skills (r= -.49 for acceptance without judgment, r= -.31 for act with awareness). AAQ-II scores also significantly correlated with the level of electric shock selected in Sample 5 (r= -.28), and with mental health (r= -.58) and physical health (r= -.44) indices of the SF-36 in Sample 6.

Discussion and conclusions

The data obtained demonstrated that this Spanish version of the AAQ-II possesses: (a) good internal consistency (overall alpha= .88), (b) construct validity (factor analysis yielded a one-factor solution), (c) discriminant validity (AAQ-II significantly discriminated between clinical and non-clinical samples), and (d) strong external validity with substantial correlations with a wide range of psychological symptoms, quality-of-life measures, and other psychological constructs. The data obtained were very similar to those reported in the original AAQ-II version by Bond et al. (2011).

A substantial difference was observed between the two clinical groups in internal consistency, likely due to differences in administration procedures. Participants in Sample 1 responded individually, while those in Sample 6 had items read aloud by a psychiatrist, which may have facilitated understanding and more accurate responses. The authors acknowledge several limitations: the proportion of women was significantly higher than men in most samples; some scales used for concurrent validity lacked formal validation; no information was obtained about treatment course in clinical samples; and although the study is a compilation of six samples from independent studies, this also constitutes a strength in demonstrating that the translation functions well across diverse contexts.

Relevance to measurement in ACT/CBS

This research contributes significantly to the availability of reliable instruments to measure experiential avoidance and psychological inflexibility in Spanish-speaking contexts. By confirming that the Spanish version of the AAQ-II maintains the psychometric properties of the original version, the study facilitates clinical research and therapeutic practice in ACT in Hispanic populations. The scale provides a valid and reliable measure for both research evaluation and monitoring change in the therapeutic process, being especially valuable for understanding how psychological inflexibility relates to psychopathology and quality of life.