ABSTRACT
Given the high importance of Web use in day-to-day life, understanding and predicting Web browsing behavior are important endeavors for academic as well as commercial research. We propose that routines – behaviors that are performed regularly (activation) and in similar ways (execution) – can be used to describe underlying structures in Web use and, thereby, improve the predictability of subsequent Web use behavior. Based on two large-scale Web-tracking data sets, we developed indicators for routine activation and execution, tested their reliability and stability, and examined their association with the predictability of six Web use behaviors: overall Web use, news use, shopping, Facebook, YouTube, and Google use. Our results show that routine measures derived from tracking data were only moderately reliable but very stable. Routine activation was consistently and substantially associated with higher predictability for all Web use behaviors, whereas routine execution was less consistently and not strongly linked to predictability.
Disclosure Statement
No potential conflict of interest was reported by the author(s).
Data Availability Statement
The data described in this article are openly available in the Open Science Framework at https://osf.io/6q9ac/.
Open Scholarship
This article has earned the Center for Open Science badge for Open Data. The data are openly accessible at https://osf.io/6q9ac/.
Supplementary Material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/15213269.2022.2121286
Notes
1. One possible drawback of the squared correlation is that, in theory, a high negative correlation between two months of Web use behavior could also result in a high prediction score. However, in practice, this did not happen: Although 20% of the raw serial correlations were negative, their average score was r = −.05.