Goal-directed rumination and its antagonistic effects on problem solving: a two-week diary study

ABSTRACT

Background: The tendency to repetitively and intrusively think about a particular negative event, goal failure, or problem (i.e., goal-directed rumination) is generally associated with impairments in well-being, thus decreasing performance in solving this failure. However, rumination is also associated with higher levels of resources invested in problem solving, likely leading to an improvement in performance. Objectives: The current study thus examines the indirect effect of rumination via various mediators on subjective problem-solving performance in the everyday context. Design: Over a period of two weeks, 147 students completed a brief survey each evening (i.e., diary study). Methods: Data were analyzed by means of a multiple mediation model in the multilevel structural equation modeling (MSEM) framework. Results: The analyses revealed that perceived stress and negative mood negatively mediated the relationship between rumination and problem solving, while attention and effort positively mediated this relationship. Finally, both a negative direct and total effect of rumination on problem solving was observed. Conclusions: Conclusively, goal-directed rumination exerted a negative indirect effect on subjective problem solving via perceived stress and negative mood, whereas it positively affected problem solving via attention and effort. Possible limitations and implications are discussed.

KEYWORDS:

• Goal-directed rumination
• problem solving
• stress
• attention
• effort
• diary study

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. The models were estimated both at the within- and between-person level. For the sake of clarity, however, only the estimates at the within-person level are presented.

2. I also estimated a random coefficient (i.e., slope; Finch & Bolin, 2017) model. Due to the complexity of the model I decided to estimate a random coefficient model with manifest variables using the MLF estimator in Mplus. The means for the random slopes were all significant except for the path between tiredness and performance (equivalent to the original fixed slope model). The variances of the random slopes were not significant. In other words, there was no variation between individuals in the nature of the relationships.