The structure and reliability of the Clinical Perfectionism Questionnaire

https://doi.org/10.1016/j.paid.2012.02.003Get rights and content

Abstract

Pathological perfectionism is of increasing interest in clinical research, although the dimensionality of this construct is actively debated. Most studies refer to two underlying dimensions associated with evaluative concerns and personal standards, and multidimensional scales are used to capture these. The more recently proposed construct of ‘clinical perfectionism’ (CP), is argued as unidimensional, as is the Clinical Perfectionism Questionnaire (CPQ) arising from this. This study assesses the reliability and validity of the CPQ in a sample of young adults. Utilising a survey design, participants were 491 undergraduate students aged 18–30 years who completed a battery of psychometric measures, of whom 142 were retested after 4 months. After removal of two items, exploratory factor analysis and parallel analysis revealed two distinct factors broadly consistent with existing two-factor formulations of pathological perfectionism, but with modest internal consistency and test–retest reliability. Norms need to be established in order to ascertain meaningful cut-offs and to aid understanding about significant improvement in the different dimensions if the CPQ is to be used in future research. Further research also needs to consider the relative utility of the CPQ against already existing measures of pathological perfectionism in common use.

Highlights

► We report psychometric properties of the Clinical Perfectionism Questionnaire (CPQ). ► The modified 10-item CPQ has two dimensions. ► Dimensions are consistent with evaluative concerns and personal standards perfectionism. ► Future research should establish the merits of the CPQ against existing measures.

Introduction

The construct of perfectionism has long held interest as being both causal and maintaining factors of a variety of psychological conditions (see Shafran & Mansell, 2001 for a review). Arising from this, debates have arisen about the dimensionality of perfectionism particularly as these relate to certain disorders. Generally two higher order dimensions have been focused on: adaptive or ‘benign’ forms of perfectionism, and pathological or ‘problematic’ forms (Frost, Marten, Lahart, & Rosenblate, 1990). The former typically involves high self-imposed, personal standards (PS), while the latter involves self-critical evaluative concerns (EC) including excessive concern over mistakes and doubts about actions (see Dunkley, Blankstein, Masheb, & Grilo, 2006). Factor analytic studies have largely supported the importance of distinguishing between these two dimensions (e.g., Bieling et al., 2004, Dunkley et al., 2003, Hill et al., 2004) and measures reflecting these are in common use. In the main, researchers have used relevant subscales from the Frost et al. (1990) Multidimensional Perfectionism Scale (FMPS) or the Hewitt, Flett, Turnbull-Donovan, and Mikail (1991) Multidimensional Perfectionism Scale (HMPS). These measures are closely related (Frost, Heimberg, Holt, Mattia, & Neubauer, 1993).

Critical of this multidimensional approach, Shafran, Cooper, and Fairburn (2002) coined the term clinical perfectionism (CP) to describe “the overdependence of self-evaluation on the determined pursuit of personally demanding, self-imposed, standards in at least one highly salient domain, despite adverse consequences” (p. 778). Unlike the broader construct of perfectionism which may have some benefits such as positive striving (Bieling et al., 2004), the self-imposed standards in CP are dysfunctional, striving for these is continuous but results in multiple psychological consequences. CP is argued as the clearest conceptualisation of pathological perfectionism (Shafran et al., 2003, Shafran et al., 2002) and has been applied to a range of conditions particularly eating disorders (Rieger et al., 2010, Riley et al., 2007).

However the construct of CP is not without its detractors (e.g., Dunkley et al., 2006, Hewitt et al., 2003). In particular, Hewitt et al. (2003) argue that Shafran et al. (2002) overly emphasise the self-orientated aspects at the cost of wider relational and interpersonal dimensions that make up the multiple aspects which existing multidimensional measures are designed to capture. They go onto criticise CP as “a self-contained unidimensional model” (Hewitt et al., 2003, p. 1232) and if applied to treatment, risks bringing about temporary change only. Glover, Brown, Fairburn, and Shafran (2007) likewise consider CP as “largely speculative” (Glover et al., 2007, p. 86). Shafran et al. (2003) do not dispute that perfectionism (as opposed to CP) may have multiple dimensions but reemphasise that CP is a more “circumscribed clinical construct” (Shafran et al., 2003, p. 1218).

When first proposing the construct of CP, Shafran et al. (2002) considered that existing psychometric measures of perfectionism were too broad and therefore failed to adequately capture the elements of self-evaluation core to their theoretical model, and were problematic because of the inclusion of benign aspects of perfectionism. This prompted Fairburn and associates (see Riley et al., 2007) to propose a specific measure of CP, the 12-item Clinical Perfectionism Questionnaire (CPQ). The CPQ has since been used in a number of published studies (Shafran et al., 2004, Steele et al., 2011) and larger treatment trials including those targeting CP (Glover et al., 2007, Riley et al., 2007). Preliminary psychometric properties of the CPQ have been reported against an interview-based measure, the Clinical Perfectionism Examination (CPE) (Riley et al., 2007) which is itself awaiting the fuller publication of psychometric properties. These preliminary data showed that the CPQ had adequate convergent validity (r = .57), and that it could distinguish between clinical and non-clinical samples, although the data relating to these samples are unpublished (see Riley et al., 2007). Steele, O’Shea, Murdock, and Wade (2011) recently reported high internal consistency (α = 0.83) in a sample of 39 eating disordered women.

In light of the increasing use of the CPQ in clinical and non-clinical research, and the limited published psychometric qualities of this, the purpose of this study was to explore the psychometric properties of the CPQ in young adults, and to consider these findings in the context of debates about the dimensionality of CP. We hypothesis that the CPQ contains two perfectionism factors.

Section snippets

Design and participants

The research was completed as part of a wider range of studies investigating perfectionism and eating problems. Along with a second measure of perfectionism (see below), the CPQ was administered online at two time points 4 months apart using Survey Monkey (www.surveymonkey.com). Approval for this study was granted by the Victoria University Ethics Committee. All participants gave informed consent and anonymity was preserved. Testing at Time 1 took place in March 2009, with testing at Time 2 4 

Descriptives

First, descriptive statistics were calculated for all demographic and psychometric variables for the final sample (n = 491). Participants at Time 1 were aged between 18 and 30 years (M = 19.1, SD = 1.8), with significantly more women (66.2%) than men. In keeping with the ethnic makeup of the wider population in this setting, the majority (78%) of the sample identified as New Zealand European, with Maori being the next largest ethnicity (4.3%). Of these, n = 142 also participated at Time 2, and this

Discussion

This study extends existing limited published information regarding the psychometric properties of the CPQ. Notwithstanding the assumed original intent to develop a measure assessing a unidimensional construct, our findings suggest that amongst young adults a 10-item modified CPQ produces two robust scales that have moderate internal consistency, but more modest stability over a four month time period. The first dimension most closely mimics ‘personal standards’ in that it specifically involves

Conclusions

The shortened 10-item CPQ consists of distinct two dimensions with good internal consistency and modest temporal stability. Given the paucity of published data on the CPQ, more research is needed to replicate the factor structure in other age and community groups and there is a particular urgency to explore the CPQ in large clinical samples given that it is currently being utilised with such groups. Further research is also needed to establish clinically meaningful cut-off scores that are

Conflict of interest

The authors have no actual or potential conflicts of interest including any financial, personal or other relationships with other people or organizations within three (3) years of beginning the work submitted that could inappropriately influence (bias) their work.

References (24)

  • R. Shafran et al.

    Perfectionism and psychopathology: A review of research and treatment

    Clinical Psychology Review

    (2001)
  • Dickie, L., Wilson, M., McDowell, J., & Surgenor, L. J. (in press). What components of perfectionism predict drive for...
  • Cited by (47)

    • Psychometric properties of the Persian version of Clinical Perfectionism Questionnaire: Findings from a clinical and non-clinical sample in Iran

      2017, Personality and Individual Differences
      Citation Excerpt :

      Also, the results provide another evidence on structural validity of CPQ. Finally, the PS factor showed acceptable internal consistency, in line with previous research (Chang & Sanna, 2012; Dickie et al., 2012; Stoeber & Damian, 2014). However, internal consistency of EC factor was less than satisfactory in both samples.

    View all citing articles on Scopus
    View full text