Activity Engagement as Escape from Self: The Role of Self-Suppression and Self-Expansion

Abstract

A two-dimensional model on escape motives in activity engagement was developed in three studies. We suggested that motivation to escape the self through engagement in an activity partly derives from intentions to either prevent negative affect or to promote positive affect. A scale for measuring these intentional mindsets was developed consisting of the subcategories self-suppression and self-expansion. Results showed that the dimensions were differently related to psychological predictors, outcomes, and experiences in the activity engagement. We argue that the present research introduces escapism as a relevant theoretical and empirical concept applicable to several types of activity engagements.

Keywords:

• activity engagement
• escapism
• flow
• regulatory focus theory
• subjective well-being

Notes

The term self-expansion has previously been used in relation to interpersonal closeness. Aron, Aron, Tudor, & Nelson (1991) described the inclusion of other into one's self as self-expansion (see also Blanchard, Perreault, & Vallerand, 1998). In the present work we define the term more broadly; that is, we apply the term to the description of how the individual's self becomes more complex through positive experiences derived from promotion focused activity engagement.

For more information on the reliability and validity of web-based surveys, see Gosling et al., 2004.

Partial correlations were also computed to investigate the validity of self-expansion and self-suppression controlled for variance shared with the alternative escapism dimension. The results from these analyses showed no substantial discrepancy from the bivariate correlations. Therefore, only zero-order correlations are presented in Table 3.

A bootstrapping procedure has recently been suggested to be the best method for testing mediation effects (Preacher & Hayes, 2008). However, in AMOS, bootstrapping is not possible to conduct using data with missing values. We did not perform imputation to account for missing data. We therefore used the Sobel method (Preacher & Hayes 2004) to test for mediation effects in Study 3.