Perfectionistic Cognitions: Stability, Variability, and Changes Over Time

ABSTRACT

The construct of perfectionistic cognitions is defined as a state-like construct resulting from a perfectionistic self-schema and activated by specific situational demands. Only a few studies have investigated whether and how perfectionistic cognitions change across different situations and whether they reflect stable between-person differences or also within-person variations over time. We conducted 2 studies to investigate the variability and stability of 3 dimensions of perfectionistic cognitions while situational demands changed (Study 1) and on a daily level during a highly demanding period of time (Study 2). The results of both studies revealed that stable between-person differences accounted for the largest proportion of variance in the dimensions of perfectionistic cognitions and that these differences were validly associated with between-person differences in affect. The frequency of perfectionistic cognitions increased during students' first semester at university, and these average within-person changes were different for the 3 dimensions of perfectionistic cognitions (Study 1). In addition, there were between-person differences in the within-person changes that were validly associated with concurrent changes in closely related constructs (unpleasant mood and tense arousal). Within-person variations in perfectionistic cognitions were also validly associated with variations in unpleasant mood and tense arousal from day to day (Study 2).

Acknowledgments

We would like to thank Dr. Dorota Reis, Dr. Charlotte Ottenstein, and Dr. Simona Maltese for valuable comments on an earlier version of this article. We would like to thank Julia Auer for her help in data collection.

Notes

¹ The complete correlation table including correlations between the affect variables and correlations with depression can be retrieved from https://osf.io/ya9ey/.

² Referring to the three-dimensional structure of the MPCI (each dimension being unidimensional), we distributed items to the first (three items) or the second indicator (two items) so that the items with comparable factor loadings were balanced across the parcels (item-to-construct balance; e.g., Little, Cunningham, Shahar, & Widaman, 2002).

³ A more detailed table of the consistency, occasion-specificity, indicator-specificity, and reliability coefficients for the LST model as well as latent correlation and regression analyses supporting the construct validity of the latent trait factors of PS, CM, and PP in their associations with DPS, DPC, dispositional affect (T1), affect in response to the exam (T4), and depression (T1) can be retrieved from https://osf.io/ya9ey/.

⁴ Analogous to the perfectionistic cognitions LC model, the latent state measurement models for unpleasant mood and tense arousal were based on two item parcels each (validated parallel test halves; Steyer, Schwenkmezger, et al., 1997) and strong factorial invariance.

⁵ An extended version of Table 4 including the reference models of DPS and DPC as the only Level 2 predictor variables is available at https://osf.io/ya9ey/.