Psychometric properties of the Avoidance and Fusion Questionnaire for Youth: A psychological measure of psychological inflexibility in youth

Highlights

• A study to further validate the Avoidance and Fusion Questionnaire for Youth.

• The measure is supported psychometrically.

• The shorter AFQ-Y8 worked as well. Hence we recommend the shorter version.

• The measure has research and clinical utility.

Abstract

Background

Consistent with the theory underlying Acceptance and Commitment Therapy (ACT), a construct termed psychological inflexibility has been suggested. There are few validated measures of psychological inflexibility for children and adolescents. One such validated instrument is the Avoidance and Fusion Questionnaire for Youth (AFQ-Y). The results from a single study have not so far been replicated for youth.

Objective

The aim of this study is to test psychometric characteristics for Swedish versions of the AFQ-Y17 and AFQ-Y8. The participants were recruited from an on-going study within institutional care for young people (12–20 years) with psychosocial problems.

Findings

In this study of a sample of 159 Swedish adolescents (15–20 years of age) with psychosocial problems, the full 17-item scale was used, and analyses of the 8 items that constitute the short version (AFQ-Y8) were conducted later. The short version showed better psychometric properties than the full 17-item scale. Confirmatory factor analysis (CFA) was conducted to examine factor solutions. The better fit for AFQ-Y8 was indicated by three of the most common measures of model fit: the comparative fit index (CFI) value exceeded .95 (the recommended cut-off value), the root mean squared error of approximation (RMSEA) was below .08 (the recommended cut-off value), and the value of the standardized root mean squared residual (SRMR) was below the recommended .05 for the AFQ-Y8 scale, which the AFQ-Y17 scale failed to attain.

Conclusions

This study supports the reliability, convergent validity and generalizability of both AFQ-Y17 and the shorter version AFQ-Y8. With regards to validity, both versions related in a theoretically consistent way with other psychological constructs. AFQ-Y8 was well represented by a single factor structure, while AFQ-Y17 showed a less good fit to a single factor structure. Overall, the AFQ-Y17 and AFQ-Y8 may be valuable clinical tools in reflecting changes in psychological inflexibility among adolescents. However, since the shorter version, AFQ-Y8, had psychometric properties that were at least as robust as the full 17-item scale, the shorter version is recommended for use among adolescents.

Introduction

There is a new generation of behaviour and cognitive therapies that focuses on processes such as acceptance, mindfulness, attention and values (Hayes, Barnes-Holmes, & Wilson, 2012). According to Hayes, Boyd, and Sewell (2011), these approaches differ from more traditional cognitive behaviour therapies (CBTs) on philosophical, theoretical, and clinical grounds and are referred to as “contextual behavioural therapies” or “contextual CBTs” (Hayes, Villatte, Levin, & Hildebrandt, 2011). Acceptance and Commitment Therapy (ACT) is one of those therapies and has a broad and growing body of evidence for its effectiveness and efficacy. Meta-analytic studies and systematic reviews have shown that ACT is effective with medium to large effect sizes when compared to wait list control groups or treatment as usual comparison groups, across a very wide range of problem areas (A-Tjak et al., 2015; Hayes et al., 2006, Ost, 2014, Powers et al., 2009). ACT has also been applied across a variety of child and adolescent populations, even if there are yet few published studies that examine their apparent efficacy (Halliburton & Cooper, 2015).

Among the more notable differences between ACT and more traditional CBT models is the emphasis in ACT on “effective living” (as clients define it) and not on symptom alleviation per se; thus, the goal of ACT is to promote living meaningful lives, even if doing so causes stress and self-doubt. To this end, the goal of ACT is to increase psychological flexibility, which is defined as the ability to contact the present moment, as it is and without psychological defence, and based upon what the situation affords, to change or persist in behaviour in accordance with one's values (Hayes & Strosahl, 2004). When persisting in behaviour in accordance with one's values, people can experience psychological events like stress, sadness, pain and anxiety, even if they are making progress towards living a more meaningful life. Hence, there is a need for measures that capture this construct of psychological flexibility, across different populations and situations.

The Acceptance and Action Questionnaire (AAQ-II) is a widely used measure of psychological inflexibility for adults. The AAQ-II has adequate psychometric characteristics (Bond et al., 2011), and it has been adapted for specific populations, such as smokers (Gifford et al., 2004), pain patients (Vowles, McCracken, McLeod, & Eccleston, 2008) and those coping with epilepsy (Lundgren, Dahl, Yardi, & Melin, 2008). The AAQ-II has also been tailored for different use in specific situations, such as the work context (Bond, Lloyd, & Guenole, 2013).

However, there are few validated measures of psychological inflexibility for children and adolescents. To our knowledge, the only existing questionnaire with published psychometric properties, is the Avoidance and Fusion Questionnaire for Youth (AFQ-Y), (Greco et al., 2008aa, Greco et al., 2008bb), which Greco et al., 2008aa, Greco et al., 2008bb showed was a valid and reliable measure of psychological flexibility in this population.

There have been two published tests of the psychometric properties of the AFQ-Y for adults (Fergus et al., 2012, Schmalz and Murrell, 2010). Schmalz and Murrell (2010) that compared the AAQ-II with AFQ-Y. These examined the possibility that the AFQ-Y may serve as a better global measure of psychological inflexibility than the AAQ-II, at least amongst children and adolescents. Two reasons for suspecting this are: 1) that the AAQ-II, though psychometrically sound, does not measure psychological inflexibility in a more general way since it does not address avoidance of physical sensations, and 2) the AFQ-Y was developed as a measure of psychological inflexibility in children and adolescents, specifically. The authors maintain that the items were created to use what they considered to be less ACT-specific language that required less knowledge of ACT to understand the intended meanings. Schmalz and Murrell (2010) found that AFQ-Y and AAQ-II scores were significantly related, but they were not so closely related as to say they measured the same construct. They go on to purport reasons as to why the AFQ-Y is perhaps a more exacting and comprehensible measure of psychological inflexibility.

Fergus et al. (2012) also compared the AAQ-II with AFQ-Y. Three of their main findings were: 1) The AFQ-Y demand a lower reading level. The AFQ-Y (but not the AAQ-II) demonstrated a reading level at or below the recommended 5th or 6th grade reading level. 2) The AFQ-Y demonstrated adequate reliability and validity, and 3) The AFQ-Y showed incremental validity over the AAQ-II in predicting several psychological symptom domains among adults.

Taking the findings from Fergus et al., 2012, Schmalz and Murrell, 2010, the AFQ-Y appears to be a more appropriate and promising assessment tool for older adolescents relative to the AAQ-II, especially when reading levels are low. Despite these findings, there is a need to replicate and further analyze the psychometric characteristics of the AFQ-Y. Thus, the present study has four aims: 1) to further validate the AFQ-Y, 2) to test the reliability of findings in comparison to earlier research, 3) to test the generalizability of the measure in a non-English language used by people in a different country and culture (Sweden), 4) to test the validity and reliability of a shorter 8-item version of the AFQ-Y.