Perfectionism is “a multidimensional personality disposition characterized by striving for flawlessness and setting exceedingly high standards of performance” (Stoeber, 2018, p. 3). Perfectionism has been interesting topic since the end of the 1950s. This concept has been tried to be measured by many researchers. Burns (1980) developed the first Perfectionism Scale. Three years later Garner, Olmstead and Polivy (1983) addressed this concept as a part of Eating Disorder Inventory. Frost, Marten, Lahart and Rosenblate (1990) defined perfectionism that depended on multidimensional perspective and they developed Multidimensional Perfectionism Scale which had six dimensions. Hewitt and Flett (1991) developed another perfectionism scale with three dimensions. These scales had similar and different aspects. Following an extensive literature review, Smith, Saklofske, Stoeber and Sherry (2016) developed a new perfectionism scale (The Big Three Perfectionism Scale) which has ten facets under three global factors. Therefore, The Big Three Perfectionism Scale (BTP) was proposed to have second order structure. The present study addressed the validation study of this scale in Turkish sample.

The purpose of this study was to validate factor structure of 45-items The Big Three Perfectionism Scale, originally developed by Smith, Saklofske, Stoeber, & Sherry (2016), in Turkish sample. Then the measurement invariance of the scale was evaluated for gender groups. In the first step, the three dimensions, second order, BTP Scale measurement model was examined for whole group. In the second step, measurement invariance for gender groups was examined. Cronbach alphas were calculated to get evidence for reliability.

About the Big Three Perfectionism Scale

BTP is a self-report scale with 45 items that measures perfectionism and originally developed by Smith, Saklofske, Stoeber and Sherry (2016). The BTP consists of three global factors:

Self-critical perfectionism: This factor has four facets: concern over mistakes (five items, e.g., when I make a mistake, I feel like a failure), doubts about actions (five items, e.g., I feel uncertain about most of my actions), self-criticism (four items, e.g., I judge myself harshly when I don’t do something right), and socially prescribed perfectionism (four items, e.g., People are disappointed in me whenever I don’t do something perfectly)

Rigid perfectionism: This factor has two facets: self-oriented perfectionism (five items, e.g., It is important to me to be perfect in everything I attempt), self-worth contingencies (five items, e.g., I could never respect myself if I stopped trying to achieve perfection).

Narcissistic perfectionism: This factor has four facets: other-oriented perfectionism (five items, e.g., It is important to me that other people to things perfectly), hypercriticism (four items, e.g., I am highly critical of other people’s imperfections), entitlement (four items, e.g., I expect other people to bend the rules for me), grandiosity (four items, e.g., Other people acknowledge my superior ability).

Method Participants Teacher Certificate Program students in Turkey constituted the sample of the study. BTP scale was applied to 609 pre-service teachers (449 female, 160 male). Their ages ranged from 18 to 40 (M=22.16, S=3.32). The data was collected during spring semester of the 2016-2017 academic year. All participants completed both demographic questionnaire (e.g., age, gender) and the 45-items BTP Scale. Data Analysis Firstly, the original form of the tool was translated from English into Turkish by five experts by following adaptation guidelines. Test and item analyses were performed to evaluate items. The reliability was calculated by the Cronbach-alpha internal consistency coefficient for all sub-scales. Confirmatory factor analysis was conducted for the three factors BTP measurement model using weighted least squares means and variance adjusted (WLSMV) estimation method with Mplus 7.2 (Muthén, & Muthén, 2013). Model fit was evaluated using χ², χ²/sd, RMSEA (root mean square error approximation), CFI (comparative fit index), NFI (normed fit index), and NNFI (nonnormed fit index) (Brown, & Cudeck, 1993, Kline, 2005, Hu, & Bentler, 1998). The measurement invariance of the BTP measurement model was examined across gender groups. Multi-group confirmatory factor analysis technique was used for this analysis. In current research, three levels of measurement invariance was examined: configural, metric and scalar invariance (Milfont, & Fischer, 2010, Vandenberg, & Lance, 2000). In the first level, BTP measurement model was tested whether the same factor structure across the gender. In the second level, metric invariance that factor loadings of BTP Scale were constrained to be equal across the gender was tested. In the third level, scalar invariance was examined constraining item intercepts to be equal across the subgroup in addition to factor loadings (Milfont, & Fischer, 2010; Vandenberg, & Lance, 2000). Chi-square test is sensitive to large samples and even small differences in covariance matrices could show significant chi-square value (Tabachnick, & Fidell, 2001). Therefore, χ² as a model fit index and ∆χ2 as a measurement invariance detection criteria were not considered (Cheung, & Rensvold, 2009). Measurement invariance was carried out by comparing ∆CFI and ∆RMSEA values with cutoff criteria (∆CFI≤.01; ∆RMSEA≤.015) suggested by Chen (2007) and Cheung and Rensvold (2009).

Results In current study, Cronbach alpha coefficients for self-critical, rigid and narcissistic perfectionism subscales were found as .89, .90, and .91, respectively. When Cronbach alpha coefficients were calculated for ten facets, Cronbach alpha coefficients ranged from .71 to .87 for these subscales. According to the present study results, reliability coefficients of the subscales were satisfactory. These evidences showed that BTP Scale was a reliable measurement tool to assess perfectionism for Turkish sample. First, the BTP measurement model was examined for the whole group. The results of confirmatory factor analysis showed that fit indexes had been meet expected criteria (RMSEA≤.08), CFI (CFI≥.90), NFI (NFI≥.90), and NNFI (NNFI≥.90). Therefore, factor structure with three global factors of BTP Scale proposed by Smith, Saklofske, Stoeber & Sherry (2016) was confirmed in Turkish sample for the whole group. Then, measurement invariance across females and males was examined. The fit of configural model presented a good fit. Following configural model, metric invariance was examined by constraining factor loadings to be equal across the gender. Compared with configural model as a baseline model, the fit values of metric model supported the invariance. Scalar invariance was investigated by constraining item thresholds to be equal across females and males in addition to factor loadings. Scalar invariance did not result in a loss of fit. As a general, configural, metric and scalar invariance models resulted in adequate fit indices. This research has some limitations. In the study, all analysis was conducted on convenience sample of 609 students. To increase the generalizability of research findings, BTP Scale could be carried out in larger samples. In current research, measurement invariance was tested according to gender. Cross-cultural comparisons with original data are planned for future studies.

References Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In: K. A. Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 136-162). Beverly Hills, CA: Sage. Bums, D. D. (1980). The perfectionist's script for self-defeat. Psychology Today, 14, 34-51. Chen, F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modelling, 14, 464-504. Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling, 9, 233–255. Frost, R. O., Marten, R A., Lahart, C., & Rosenblate, R. (1990). The dimensions of perfectionism. Cognitive Therapy and Research, 14, 449-468. Garner, D. M., Olmstead, M. P., & Polivy, J. (1983). Development and validation of a multidimensional eating disorder inventory for anorexia nervosa and bulimia. International Journal of Eating Disorders, 2, 15-34. Hewitt, R L., & Flett, G. L. (1991). Perfectionism in the self and social contexts: Conceptualization, assessment, and association with psychopathology. Journal of Personality and Social Psychology, 60, 456-470. Hu, L. T., & Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity to underparameterized model misspecification. Psychological Methods, 3, 424–453. Kline, R. B. (2005). Principles and practice of structural equation modeling. New York: Guilford Publications, Inc. Milfont, T. L., & Fischer, R. (2010). Testing measurement invariance across groups: Aplications in cross-cultural research. International Journal of Psychological Research, 3(1), 111-121. Muthén, L., & Muthén, B. O. (2013). Mplus user’s guide. Los Angeles, CA: Author Smith, M. M., Saklofske, D. H., Stoeber, J., & Sherry, S. B. (2016). The big three perfectionism scale: A new measure of perfectionism. Journal of Psychoeducational Assessment, 34, 670–687. Stoeber, J. (2018). The psychology of perfectionism: An introduction. In J. Stoeber (Ed.),The psychology of perfectionism: Theory, research, applications (pp. 3-16). London: Routledge.Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics. Boston: Allyn and Bacon. Vandenberg, R. J., & Lance, C.E. (2000). A Review and Synthesis of The measurement invariance literature: Suggestions, practices and recommendations for organizational research. Organizational Research Methods, 3(1), 4-70.