Multivariate analysis approach to relationships between perfectionism and obsessive compulsive symptoms

Keskin S., Boysan M., GOEKTAS I.

TURKIYE KLINIKLERI TIP BILIMLERI DERGISI, vol.28, no.3, pp.319-326, 2008 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 28 Issue: 3
  • Publication Date: 2008
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.319-326
  • Van Yüzüncü Yıl University Affiliated: Yes


Objective: In general, scientific studies consist of multivariate measurements. Both correlation and regression analyses are the most common methods used to determine the relationships between variables. However, in some cases it may be necessary to examine the relationship between two variable sets, called X and Y variable sets. For this purpose, canonical correlation and multivariate regression analyses, which are multivariate methods, may be preferable. In this study, canonical correlation and multivariate regression analyses were used to determine the relationship between two variable sets, perfectionism and obsessive-compulsive symptoms. Material and Methods: The X variable set comprised self-oriented perfectionism, socially prescribed perfectionism, and others-oriented perfectionism. Impulsiveness, washing, checking, rumination, and precision formed the Y variable set as dependent variables (obsessive-compulsive symptoms). A total of 219 subjects were included in the analysis. Results: According to the canonical correlation analysis, the first two canonical correlations (0.406 and 0.283) were significant. However, the results of the regression analysis indicated that the Determination coefficients for dependent variables except for "Precision" ranged from 9.1% to 11.6% and were significant. Conclusion: Using multivariate and multiple regression analyses concurrently may be useful. In addition, multivariate regression and canonical correlation analyses together may provide the opportunity to evaluate the relationships between variable sets in detail.