City Research Online - CORE (PDF)

Gillanders, D.T., Bolderston, H., Bond, F.W., Dempster, M., Flaxman, P.E., Campbell, L., Kerr, S., Tansey, L., Noel, P., Ferenbach, C., Masley, S., Roach, L., Lloyd, J., May, L., Clarke, S. & Remington, B. (2014). The Development and Initial Validation of the Cognitive Fusion Questionnaire. Behavior Therapy, 45(1), doi: 10.1016/j.beth.2013.09.001 City Research Online Original citation: Gillanders, D.T., Bolderston, H., Bond, F.W., Dempster, M., Flaxman, P.E., Campbell, L., Kerr, S., Tansey, L., Noel, P., Ferenbach, C., Masley, S., Roach, L., Lloyd, J., May, L., Clarke, S. & Remington, B. (2014). The Development and Initial Validation of the Cognitive Fusion Questionnaire. Behavior Therapy, 45(1), doi: 10.1016/j.beth.2013.09.001 Permanent City Research Online URL: http://openaccess.city.ac.uk/8562/ Copyright & reuse City University London has developed City Research Online so that its users may access the research outputs of City University London's staff. Copyright © and Moral Rights for this paper are retained by the individual author(s) and/ or other copyright holders. All material in City Research Online is checked for eligibility for copyright before being made available in the live archive. URLs from City Research Online may be freely distributed and linked to from other web pages. Versions of research The version in City Research Online may differ from the final published version. Users are advised to check the Permanent City Research Online URL above for the status of the paper. Enquiries If you have any enquiries about any aspect of City Research Online, or if you wish to make contact with the author(s) of this paper, please email the team at [email protected] brought to you by CORE View metadata, citation and similar papers at core.ac.uk provided by City Research Online  1 The development and initial validation of The Cognitive Fusion Questionnaire David T Gillandersa*, Helen Bolderstonb, Frank W. Bondc, Maria Dempsterd, Paul E. Flaxmane, Lindsey Campbelld, Sian Kerrd, Louise Tanseyf, Penelope Noelg, Clive Ferenbachh, Samantha Masleyd, Louise Roacha, Joda Lloydc, Lauraine Mayc, Susan Clarkei, Bob Remingtonb *Corresponding author aUniversity of Edinburgh, School of Health in Social Science, Teviot Place, Edinburgh, United Kingdom, EH8 9AG, [email protected], Tel: +44(0)131651 3946 b School of Psychology, University of Southampton, United Kingdom, SO17 1BJ. c Institute of Management Studies, Goldsmiths, University of London, New Cross, United Kingdom, SE14 6NW. d Clinical Psychology, Royal Cornhill Hospital, Aberdeen, United Kingdom, AB25 2ZH. e Dept. of Psychology, City University London, United Kingdom, EC1R 0JD. f Orchard Clinic, Royal Edinburgh Hospital, Edinburgh, United Kingdom, EH10 5HF. g Clinical Psychology, 7 Dudhope Terrace, Dundee, United Kingdom, DD3 6HG. h Sir George Sharp Unit, Cameron Hospital, Leven, Fife, United Kingdom, KY8 5RR i The School of Health and Social Care, Bournemouth University, Bournemouth, United Kingdom, BH1 3LT. Acknowledgements: We are very grateful to George Deans, Morag Taylor and Sam Aitcheson, NHS Grampian; Kevin Power, NHS Tayside; Mark Ramm, Sarah Gillanders and Louise McNeil, NHS Lothian; Alan Harper, Kathryn Quinn, NHS Fife; David Whitty, Dorset Health Care University NHS Foundation Trust; Nick Maguire, University of Southampton, and Ken Laidlaw, University of Edinburgh. The ACT workplace intervention study was supported by a grant from the Economic and Social Research Council (ESRC grant no: RES- 061-0232) awarded to Paul E. Flaxman.  2 Disclosure Statement There are no conflicts of interest declared Abstract Acceptance and Commitment Therapy (ACT) emphasizes the relationship a person has with their thoughts and beliefs as potentially more relevant than belief content in predicting the emotional and behavioral consequences of cognition. In ACT, ‘defusion’ interventions aim to ‘unhook’ thoughts from actions and to create psychological distance between a person and their thoughts, beliefs, memories and self-stories. A number of similar concepts have been described in the psychology literature (e.g. decentering, metacognition, mentalization and mindfulness) suggesting converging evidence that how we relate to mental events may be of critical importance. Whilst there are some good measures of these related processes, none of them provides an adequate operationalization of cognitive fusion. Despite the centrality of cognitive fusion in the ACT model, there is as yet no agreed measure of cognitive fusion. This paper presents the construction and development of a brief, self-report measure of cognitive fusion: The Cognitive Fusion Questionnaire (CFQ). The results of a series of studies involving over 1800 people across diverse samples show good preliminary evidence of the CFQ’s factor structure, reliability, temporal stability, validity, discriminant validity, and sensitivity to treatment effects. The potential uses of the CFQ in research and clinical practice are outlined. Key words: Cognitive-Behavior Therapy, Acceptance and Commitment Therapy, Measurement, Questionnaires, Cognitive Fusion, Mindfulness Introduction Acceptance and Commitment Therapy (ACT) is a form of cognitive behavioral therapy (CBT) that emphasizes distancing from thoughts, rather than changing thought content as a key ingredient in therapy. The ACT model has been thoroughly described  3 elsewhere, interested readers are directed to Hayes, Strosahl and Wilson (2011). Briefly, the ACT model describes 6 overlapping processes (willingness, defusion, contact with the present moment, clarity of values, committed actions and a flexible repertoire of perspective taking skills known as ‘self as context’) that lead to psychological flexibility. In the ACT model, distancing from thoughts is known as ‘cognitive defusion’ and its counter process is ‘cognitive fusion’: the tendency for behavior to be overly regulated and influenced by cognition. When ‘fused,’ a person acts on thoughts as though they are literally true, cognitive events come to dominate behavior and experience over other sources of behavioral regulation, and he or she becomes less sensitive to direct consequences. An example of this is when a person with social anxiety is dominated by self-focused attention, negative evaluations of the self and such cognitive events have a strong tendency to regulate overt behaviors (such as entering social situations). In contexts of fusion, thoughts are taken literally; the anxious person simply is socially inept, rather than seeing these self-evaluations as mental events. Fusion refers to the relationship a person has with his or her own cognitive events, on a continuum from fused (dominated by, entangled, believed, taken literally) to defused (experienced as mental events and not necessarily needing to be acted upon). In ACT, the purpose of defusion is to afford greater choice of behavior, such that the socially anxious person could choose to enter the feared situation, even when they are fearful and their mind is predicting the worst. Cognitive defusion overlaps with but is distinct from other processes of the ACT model. Willingness describes a behavioral stance that is open to unpleasant private events and a letting go of attempts to control or avoid their form, frequency or intensity. Contact with the present moment refers to flexible awareness of experience in the here and now, encompassing sensation, emotion, cognition, and kinesthetic awareness. Self-as-context refers to a set of perspective shifting skills in which the self is experienced as hierarchically  4 organized in relation to self-content (e.g., what one believes about oneself). It relies on the more basic process of cognitive defusion, and involves a shift from experiencing self-content and ‘the self’ in a relationship of equivalence, to the self as ‘containing’ self-content, be it positive or negative. Experimental and clinical work in ACT has often operationalized fusion using the proxy construct of ‘believability of thoughts’. Such an operationalization suffers from focusing merely on the content of one’s thought and so represents only a narrow aspect of fusion. A broader behavioral operationalization includes dominance of cognitive events in a person’s experience, inability to view cognitive events from a different perspective, reacting emotionally to thoughts, behavior being highly regulated by cognitive events, attempts to control thinking, over-analysis of situations, evaluating and judging thought content, as well as aspects of literality and believability. Cognitive defusion is similar to a number of existing concepts within psychological therapies, namely: decentering (Fresco, Moore, van Dulmen & colleagues, 2007; Safran & Seagal, 1990), metacognitive awareness (Teasdale et al., 2002; Wells, 2008), mindfulness (Bishop et al., 2004), and mentalization (Bouchard et al., 2008; Fonagy & Target, 1997) It is also distinct from thought-action fusion (TAF: Shafran, Thordarson & Rachman, 1996), despite the similarity in name. Decentering or ‘metacognitive awareness’ refers to the capacity to take a detached view of one’s thoughts and emotions (Fresco et al., 2007; Teasdale et al., 2002; Wells, 2008; 2009). Decentering has long been part of cognitive therapy, though initially it was as a precursor to cognitive modification (Hollon & Beck, 1979). Recent analyses of Mindfulness Based Cognitive Therapy (MBCT: Segal, Williams & Teasdale, 2002) suggest decentering may be a more active ingredient in preventing depressive relapse via the disruption of depressogenic cognitive patterns (Teasdale et al., 2002). Fresco, Segal, Buis, and Kennedy  5 (2007, p. 448) define decentering as “…the capacity to take a present-focused, non- judgmental stance in regards to thoughts and feelings and to accept them” (Our italics). Fresco et al. also note (p. 236) that an important “facet” of decentering is self-compassion, and, indeed, there are items representing this facet in Fresco et al.’s measure of decentering, the Experiences Questionnaire (EQ). The ACT model of psychological flexibility separates out the processes of defusion and acceptance, whereas decentering collapses them. In addition, from an ACT perspective, self-compassion is not a quality of defusion, rather, it emerges from four of the psychological flexibility processes described above: defusion, acceptance, self-as-context, and present-moment awareness. Thus, decentering (or metacognitive awareness) is similar to defusion, though fusion is more narrowly defined and more behaviorally operationalized, in that the function of defusion is greater choice of action, rather than the disruption of cognitive patterns. Despite the similarity in name, Thought Action Fusion (TAF: Shafran et al., 1996) describes biased metacognitive belief content (e.g. thinking of something bad happening will make it more likely to happen) rather than the shift in perspective that is suggested by other terms such as decentering, metacognitive awareness, or fusion / defusion. Arising from the psychodynamic or developmental tradition in psychology is the notion of reflective function or mentalization (Fonagy & Target, 1997; 2002), which is the capacity of a developing child to understand the mental states of itself and others. Whilst awareness of one’s own mental states may overlap with the notion of cognitive fusion / defusion, mentalization is a broader construct related to affect regulation and attachment (see Gumley, 2010 for an overview). Finally, mindfulness (e.g. Mindfulness Based Stress Reduction, [MBSR] Kabat-Zinn, 1990) shares a degree of overlap with cognitive defusion. Indeed, in the ACT model, present moment awareness is often instigated as a precursor to cognitive defusion. An example is the  6 ‘classic’ ACT defusion exercise ‘Leaves on the Stream’ (Hayes et al., 2011, pp. 255 - 258.) This begins with a focus on the present moment (akin to mindfulness of the breath and of the body, sounds etc.) before using imagery (of thoughts as leaves floating on a stream) that deliberately creates a shift in the relation between cognitive events and the self. From an ACT perspective, however, mindfulness is an emergent quality of the four psychological flexibility processes (noted above) that also produce self-compassion. Thus, defusion forms a key component of both decentering and mindfulness, but in order to produce the psychological qualities of those latter two constructs, ACT maintains that additional processes need to occur. In addition, mindfulness is often seen either as a method of stress reduction (as in MBSR) or to facilitate decentering (as in MBCT) or as an end in itself, rather than in the service of potentiating valued, goal directed action, which is the explicit function of defusion in ACT. In summary, there are a number of constructs that overlap with cognitive fusion, suggesting that a number of therapy approaches have considered aspects of how we relate to our own mental experience to be important in understanding mental disorder and behavioral function. Whilst there is overlap between these concepts and fusion, the construct of fusion is a more narrowly defined process that contributes to other psychological qualities such as decentering and mindfulness. In addition, central to the construct is that the purpose of stepping back from cognitive events is to facilitate taking action that is consistent with one’s values, rather than to disrupt negative thinking styles, change metacognitive beliefs or reduce stress. There are some existing scales that have been developed to measure cognitive fusion and defusion. The Believability of Anxious Feelings and Thoughts Scale (BAFT: Herzberg et al., 2012) measures the believability of thoughts in anxiety disorders and so, as noted above,  7 measures a relatively narrow conceptualization of fusion. In addition, it is content specific to anxiety disorders, giving it less relevance as a generic measure of cognitive fusion. The Avoidance and Fusion Questionnaire for Youth (AFQ-Y: Greco, Lambert & Baer, 2008) measures avoidance and fusion in children and adolescents and appears to have adequate psychometric properties in adult populations (Fergus et al., 2012). The scale, though, assesses several ACT processes (notably fusion and avoidance) and so cannot be regarded as a distinct measure of cognitive fusion. An early ACT study (Zettle & Hayes, 1987) operationalized defusion by incorporating a ‘believability’ scale into the existing ‘frequency’ scale of the Automatic Thoughts Questionnaire (ATQ: Hollon & Kendal, 1980). This measure has adequate psychometric properties (Zettle, Rains & Hayes, 2011). It is however, a further example of how cognitive fusion has been operationalized along the relatively narrow dimension of believability. Furthermore, its items reflect cognitive content in depression, limiting its use as a generic measure of cognitive fusion. Finally, a very recent addition to the available measures of cognitive defusion is the Drexel Defusion Scale (DDS: Forman et al., 2012). It shows a theoretically coherent factor structure, good psychometric properties and a coherent pattern of convergent and divergent validity, in a clinical and non-clinical sample. The DDS has four potential limitations, however. Firstly, the scale provides an extended instruction set that describes what is meant by the term defusion. This could be problematic because it potentially equates the act of fused responding with the understanding of the concept of defusion. Secondly, the scale presents 10 brief vignettes in which respondents rate how likely they imagine they would be to defuse in various situations, rather than to report on their actual experiences of defusing in their own life. Thirdly, the authors recognize that in providing an extended instruction set, the scale itself may prime defused responding. Finally, the DDS assesses fusion with bodily  8 sensations, emotions and cravings as well as thoughts around specific situations (e.g. social anxiety, loss / sadness and anger) meaning the items are content specific and, thus, lack generality. In summary, some scales have been developed to measure cognitive fusion but they tend to be content specific, operationalized in a narrow way, may confound measurement with understanding and contain priming effects. The present research therefore aimed to develop a brief, self-report measure of cognitive fusion of sufficient quality and flexibility in terms of item content and psychometric properties to facilitate it’s use in a variety of settings (clinical, community, laboratory), with many different populations. The studies reported here describe the construction and development of the Cognitive Fusion Questionnaire (CFQ), the examination of its factor structure in a non-clinical sample, followed by item reduction (Study 1). Study 2 further examines the factor structure of the CFQ across a wider range of samples, using confirmatory factor analysis (CFA). Study 3 examines the concurrent, divergent and incremental validity of the CFQ via relationships with a wide range of other measures of psychological processes and functioning. Study 4 reports internal and test-retest reliability and the ability of the CFQ to distinguish between different samples of participants. Study 5 reports an evaluation of the CFQ’s sensitivity to treatment and mediation of treatment outcome in an intervention designed to reduce cognitive fusion. Method Item Development Items were generated by DG, HB, MD and FB, based on expert knowledge and practice of ACT and Relational Frame Theory (RFT: Hayes, Barnes-Holmes, & Roche, 2001). In addition, FB has developed two previous ACT-based measures (Bond et al., 2011; Bond, Lloyd & Guenole, 2013). Consistent with Hayes, Luoma, Bond, Masuda and Lillis  9 (2006), fusion was considered to be a uni-dimensional construct, representing a continuum from fusion to defusion, with a wide variety of behavioral manifestations. Each item was worded to express a specific instance of fusion or defusion and to be as concrete and behaviorally operationalized as possible. Hence the complete set of items addressed a broad functional definition of fusion made up of the many forms that fusion can take, including: believability of thoughts (two items: e.g. I believe the thoughts that pop into my head); taking thoughts literally (four items: e.g. My thoughts are facts); reacting emotionally to thoughts (three items: e.g. My thoughts cause me distress or emotional pain); behavior being governed by thoughts (five items: e.g. I get so caught up in my thoughts I’m unable to do the things I most want to do); trying to control thoughts (four items: e.g. I need to control the thoughts that come into my head); over analyzing situations (four items: e.g. I overanalyze situations to the point where it is unhelpful to me); evaluating thought content (four items: e.g. I make judgments about whether my thoughts are good or bad); dominance of cognition in a person’s experience (six items: e.g. Its such a struggle to let go of upsetting thoughts, even when I know letting go would be helpful); perspective taking (six items: I find it easy to view my thoughts from a different perspective); and detached awareness of thoughts (six items: e.g. My thoughts just come and go and I’m not too attached to them). These later items are examples of defusion responses and are reverse scored. A total of 44 items were generated. ACT experts (committee members of the British Association for Behavioural & Cognitive Psychotherapy ACT Special Interest Group) were asked to comment on and rate how well these items represented cognitive fusion and defusion. Nine committee members provided feedback on item clarity and suggested adaptations to some items. A number of items were reworded based on this feedback. Items that received a modal rating of ‘Moderately’ or ‘Highly’ representative from the consensus panel were retained in the first  10 draft of the CFQ. This first draft containing 42 items (26 fusion, 16 defusion) was examined in study 1 (see Table 1 for the full list of items). Samples Several samples were recruited to test different aspects of the CFQ. In some cases samples were recruited for other projects and where samples shared similar characteristics (e.g. healthy community dwelling adults) these samples were combined opportunistically. These samples are described below. Sample 1: A young adult sample was recruited (predominantly students and associates of the chief investigators [MD & DG]). The sample comprised 592 participants: 174 were male (29.4%) and 418 were female (70.6%); 381 participants (64.4%) were aged between 17- 24 years; 173 (29.2%) were aged between 25-34 years; 21 (3.5%) were aged between 35-44 years; 8 (1.4%) were aged between 45-55years; and, 9 (1.5%) were over 55 years old. Sample 2: Sample 2 was another convenience sample, collected for 3 different studies, whose data were opportunistically pooled. These were all community dwelling healthy adults and included a sub sample of UK Prison Service Officers recruited for another study by LM and JL. Sample 2 comprised 447 adults. There were 247 women (55.4%) and 200 men (44.6%). Their ages ranged from 18 to 77, with a mean of 38 years (SD = 13). The age, gender ratio and CFQ score were not significantly different in any of the three subsamples that make up sample 2 (Mean Age: 34.8 – 39.6 years, SD: 11 – 13.5 years, F (3,444) =2.28, p=. 08, ns; Gender Ratio [% female]: 48% - 58%, λ = 1.06, p = .29; Mean CFQ score: 18.92 – 20.17, SD: 7.72 – 9.82, F (2,445)=.710, p=.49, ns). One of the subsamples was recruited for a study specifically to examine the test-retest reliability of the CFQ (n=82, 58% female, mean age: 41, SD: 13.8) and is equivalent to the other participants in sample 2 in terms of age, gender and CFQ score. In addition some participants completed paper and pencil measures and others (n = 113) completed measures online. There was no difference  11 between these methods of administration on the final CFQ score (Online mean: 20.16 [SD: 7.72], Paper mean: 19.71 [SD: 8.03], t(255) = .485, p = .63) Sample 3: Participants were volunteers for an ACT based worksite psychological skills training / stress management program as part of another study led by PF. They were recruited from two large public sector organizations in the UK: a governmental organization and a healthcare organization. This sample included 242 people, 188 (78%) of whom were female. Age ranged between 20 and 69 years old, with a mean age of 41 years (SD: 10 years). Sample 4: Participants in sample 4 were recruited from National Health Service Mental Health Services across the UK, as part of other studies by HB, SK, SM, PN and DG. They represent a broad sample of different mental health difficulties. 129 (60%) were recruited from specialist settings, such as services for people with eating disorders, community mental health teams and a personality disorders service, whereas 86 (40%) were recruited from primary care psychological services. All participants were referred by treating clinicians (psychologists, nurses, psychiatrists) who confirmed the presence of psychological disorder. Participants were excluded on the basis of primary problem relating to alcohol or substance misuse, and conditions such as significant learning disability, head injury or dementia. Sample 4 comprised of 215 individuals, aged 17 to 68, with a mean age of 40 years (SD: 13); 136 of these participants (63.3%) were female. Despite the diverse sources of recruitment for this sample, the subsamples did not differ in terms of age or CFQ total score (Specialist mean age: 40.35 years, SD: 12.89, Primary Care mean age: 38.58, SD: 12.69, t (213) =.933, p = .35, ns; Specialist CFQ: 34.57, SD: 8.31, Primary care CFQ: 33.91, SD: 7.70, t (213) =.59, p = .55, ns), suggesting equivalence across these diverse samples. Types of psychological disorder represented in this sample included major depression, anxiety disorders, post traumatic stress disorder, eating disorders, interpersonal problems, low self-  12 esteem, complicated grief reactions, personality disorders, recovery from psychotic experiences including bipolar disorder and schizophrenia. These diagnoses were assessed by clinician diagnosis rather than structured clinical interview. Sample 4 was considered a broad sample of the range of mental health problems presenting in United Kingdom NHS mental health services and was designed in this way because of the hypothesized transdiagnostic relevance of cognitive fusion. Sample 5: Sample 5 comprised 133 people (72% female) recruited for a study of adjustment processes in people with multiple sclerosis by CF and DG. The sample ranged in age from 21 to 75, with a mean age of 49 years (SD =11). Participants were recruited from specialist NHS neurological services. Diagnosis of MS was confirmed by the referring physician. Individuals who were deemed by the referring clinician to be too cognitively impaired to provide informed consent or respond to the study questionnaires were not recruited. Sample 6: Sample 6 was recruited for a separate study of the relationship between cognitive fusion, rumination, metacognitive beliefs and depression by SK and DG. The sample comprises of people with current major depressive disorder (MDD): (n=26, 58% female, mean age: 42.35 years, SD: 12.8,), people recovered from MDD (n= 21, 57% female, mean age: 41.67 years, SD: 10.95), and people who have no history of depressive episode (n=27, 59% female, mean age: 44.11 years, SD: 11.92). Current and lifetime depression status was confirmed by Structured Clinical Interview for DSM [SCID: First, Spitzer, Gibbon, & Williams, 1996]. The three groups were equivalent in terms of age and gender and differed significantly in terms of current mood, as measured by the Centre for Epidemiological Studies Depression Scale [CESD: Radloff, 1977]. Participants for currently depressed and recovered groups were recruited through General Practitioner (GP) services and Primary Care Psychological Services. Participants for the never depressed group were  13 recruited through community groups and services. Sample 6 was used to test the incremental validity of the CFQ in predicting depression, compared to established predictors such as rumination and metacognitive beliefs. Sample 7: Sample 7 consisted of 219 caregivers of people with dementia, recruited for a separate study into caregiver distress by LR and DG. 144 (66%) of the caregivers were female, the mean age of sample 7 was 68.6 years (SD: 11.5 years), with a range of 31 to 95 years. 87% of the sample were caring for a spouse with dementia, the remainder were adult children of the care recipient. 88% were living with the person with dementia. Regardless of sample, all participants gave informed consent and all studies followed codes of conduct for research with human participants as detailed by the British Psychological Society (BPS, 2009) and the universities and NHS Trusts that were sponsoring the research. All studies received approval from the relevant university ethics committees (University of Edinburgh, Goldsmiths, University of London, City University, or Southampton University). In addition, Local NHS Research Ethics Committees approved all research that sampled participants from NHS settings and equivalent ethical procedures were followed in other settings. Data screening and treatment of missing data In all samples data were screened at the individual item level to confirm that each item had responses covering the full range of the response format, and that responses were normally distributed. Missing data analysis showed low levels of missing data. Where individuals missed less than 3 items on the CFQ, these missing items were prorated based upon their scores for the other CFQ items. Where an individual had 3 or more items missing on the CFQ, they were excluded from further analysis. Across all samples, only 19 cases were excluded (less than 1% of the total number of participants). A total of 29 individuals had up to 2 items prorated (.2% of the total data). For other measures used in this set of  14 studies, the default list wise deletion option in SPSS was chosen as the most pragmatic way of dealing with missing data. In list wise deletion, cases with missing data are simply removed from that analysis, reducing sample size. Samples were analyzed separately for some studies and combined for other studies, these are described for each of the studies and sample size is given for each analysis. Study 1: Exploratory Factor Analysis and Item Reduction in a Non-Clinical Sample Participants Participants were as described above for Sample 1. Measures Participants completed a paper and pencil form of the CFQ, containing 42 items. They answered on a seven point Likert scale with the same instructions and response format as the final version of the CFQ (see appendix). Analysis The goal of this exploratory analysis was to identify one or more latent variables underlying the observed variables; as a result, we conducted a common factor analysis (CFA) (Floyd & Widaman, 1995) and determined the number of factors to extract through parallel analysis (Horn, 1965), which is a very accurate factor extraction procedure (Zwick & Velicer, 1986). We used an oblique rotation (Promax), as we expected that these factors would be elements of a higher order factor, cognitive fusion (Nunnally, 1978), and therefore should be significantly correlated. Results Prior to exploratory factor analysis (EFA), 11 items with an item-total correlation of less than .04 were omitted (Nunally & Bernstein, 1994). For the remaining 31 items, the Keiser-Meier-Olkin test of sampling adequacy (KMO) was .933, which indicates a good degree of non-unique covariance amongst the set of items (Kaiser, 1974). A significant