Finding a Data-Driven Definition of Binge-Watching

Binge-watching, the act of watching large amounts of television at a time, has become a popular phenomenon internationally; however, it has yet to be sufficiently defined. In order to define binge-watching, data was collected on specific watching instances from 216 undergraduate students at a large research university. Hierarchical and k-means cluster analyses were conducted in Phase I to empirically determine how binge-watching should be defined. In Phase II, that definition was tested by correlating the number of instances of binge-watching in a one-week period, collected by seven days of daily diary logs, with several theoretically related measures including body mass index, dissociative tendencies, psychological distress, compulsion to watch, boredom proneness, and escapism through watching. The data-driven definition was found to be that eight hours or more of continuous watching was binge-watching, while anything less than that was not. In Phase II, the frequency of binge-watching through the seven-day period was calculated based on that new definition. The frequency of binge-watching was positively correlated with body mass index and dissociative tendencies with statistical significance at the alpha = .05 level. Compulsion to watch was not statistically significant; however, there was a positive correlation. These findings indicate that the proposed data-driven definition has concurrent validity. Psychological distress, boredom proneness, and escapism through watching were not statistically significantly related to binge-watching frequency, nor did the effect sizes indicate a correlation may exist. Potential reasons for these results are discussed. The definition found in this study will be helpful to other researchers as research into binge-watching continues to grow.