Gaudreau and Thompson (2010) provided a 2 × 2 model of perfectionism that focuses on the interactive effect of two dimensions of perfectionism – personal standards perfectionism (PSP) and evaluative concerns perfectionism (ECP) – and distinguishes four sub-types of perfectionism. They evidenced an interactive effect of PSP and ECP on general negative effect, i.e., the effect of ECP is reduced when PSP is high. They also hypothesized a similar effect on a measure of psychopathology. We respond to Gaudreau and Thompson (2010) by testing this interactive effect on depressive symptoms. Analyses of data from a student sample (N = 338) failed to evidence an interactive effect of PSP and ECP on depressive symptoms. ECP was positively associated with depressive symptoms while a mild negative correlation between PSP and depression was observed. Although our results do not fully support Gaudreau and Thompson’s 2 × 2 model, they suggest that some of the sub-types they proposed are relevant for predicting depressive symptoms.
Over the past decades, perfectionism has been predominantly conceptualized as a multidimensional personality trait (e.g., Hewitt, Flett, Besser, Sherry, & McGee, 2003, but see also Shafran, Cooper, & Fairburn, 2002). From the different facets identified across studies, two core dimensions emerge (Frost, Heimberg, Holt, Mattia, & Neubauer, 1993). On the one hand, the positive strivings dimension (also called, perfectionistic strivings, Stoeber & Otto, 2006; Personal standards perfectionism, Gaudreau & Thompson, 2010) is viewed as a self-oriented motivation to set and strive for high standards (Frost et al., 1993 and Gaudreau and Thompson, 2010) and on the other hand, the maladaptive evaluation concerns dimension (also called, perfectionistic concerns, Stoeber & Otto, 2006, evaluative concerns perfectionism, Gaudreau & Thompson, 2010) is a struggle for perfection, perceived as imposed by others, accompanied by self-criticism and doubts about one’s capacity to achieve (Frost et al., 1993 and Gaudreau and Thompson, 2010). A wealth of studies examined the positive and negative outcomes of each dimension according to a dimensional or a group-based approach (for a review, see Stoeber & Otto, 2006). Stoeber and Otto (2006) proposed a conceptual framework that combined both approaches (See Fig. 1, left panel). They represent the two dimensions of perfectionism (perfectionistic strivings and perfectionistic concerns) as two orthogonal dimensions that can be combined to distinguish three groups of perfectionists: healthy perfectionists (high in perfectionistic strivings and low in perfectionistic concerns), unhealthy perfectionists (high in both dimensions) and non-perfectionist (low in perfectionistic strivings). In spite of its heuristic value, this framework entails some limitations. Firstly, the two dimensions are not perfectly independent and their inter-correlations can be high (e.g., r=.45, Bieling, Israeli, & Antony, 2004; r=.58, Gaudreau & Thompson, 2010). Second, the non-perfectionist group gathers under the same label two subgroups of individuals who are both low on perfectionistic strivings but differ on the perfectionist concerns dimension. Yet, this latter dimension is widely associated with negative outcomes. Therefore, such a distinction can lead to different conclusions in a group-based approach that peremptorily considered both profiles as control. The first limitation had been addressed by statistically controlling for the other dimensions when examining the correlation between one dimension and a negative or positive outcome. If this strategy is statistically elegant, it does not take into account the potential interactive effects of both dimensions on positive as well as negative outcomes.
