Procrastination out of Habit? The Role of Impulsive Versus Reflective Media Selection in Procrastinatory Media Use

ABSTRACT

The pervasive access to media options seriously challenges users’ self-regulatory abilities. One example of deficient self-regulation in the context of media use is procrastination—impulsively ‘giving in’ to available media options despite goal conflicts with more important tasks. This study investigaes procrastinatory media use across 3 types of media (TV, computer, smartphone) from a dual-systems perspective, taking both person-level and situation-level predictors into account. Results from a 14-day long diary study (N = 347) suggest that procrastinatory media use is driven by automatic media selection, which is facilitated by strong media habits (person level) and low motivation for behavioral control (situation level). The results underline the value of a dual-systems perspective on media choices in our media-saturated environment.

Acknowledgments

We thank Frank Mangold and the reviewers for their very helpful comments on earlier versions of the article.

Supplementary material

Supplemental data for this article can be access on the publisher’s web site.

Notes

1. The original SRHI contains an item measuring self-concept (“That’s typically me”). Self-concept, however, is not an integral part of the habit concept and the respective item was excluded from the SRHI (Gardner, 2015; Lally et al., 2010).

2. Cognitive resources were assessed for each interval independently of media use. Therefore, its internal consistency was not tested for each media device separately.

3. Variances were computed for each media device separately, and the lowest and highest values are reported.

4. Due to the close semantic relation between automatic media selection and habit strength (see Method section), it may be suspected that their relation is biased due to similar item formulation. We therefore reran all models with an eight-item version of the habit strength scale excluding the similar items. Coefficients were weaker than in the model outlined here; however, all coefficients found to be significant remained significant. A table containing the respective results is available as supplemental material to this article.