| Popis: |
Binging has emerged as a trending phenomenon among consumers of online streaming services and other media outlets that offer content consumption "on-demand." Current definitions of binging from content providers and market research firms were developed to describe television consumption and are usually based on the number of consecutive episodes of a single TV show that viewers watch in a single sitting. These definitions combine both the amount of content consumed and the category of the content, thus losing information on different ways that consumers might exhibit binge behaviors. In particular, to extend the literature, we define multiple dimensions of binging that account for the timing, volume, and type of content consumed. We quantify these dimensions using binging metrics that describe the extent to which an individual is engaging in binge consumption based on the interarrival times of consumption and breaks. We apply our binging metrics to two different datasets from contexts in which users not only consume content, but are also evaluated on their performance (i.e., online education and online video games). These applications allow us to examine how binging patterns may be predictive of downstream behaviors that are of importance to both consumers and content providers. In particular, we identify which binging types seem to help or hurt different performance outcomes. In addition, we examine how performance may be affected by counterfactual policy changes designed to manage binge behavior. |
