TL;DR
While most research on media "bingeing" focuses on why we start watching, this study investigates the psychological mechanics of how we finally stop. Using real-time tracking of Netflix users, the researchers found that while enjoyment "pulls" us into the next episode, feelings of appreciation and "spoiled pleasure" (guilt) eventually provide the "push" needed to disengage, provided we have the cognitive self-control capacity to act on them.
What Problem Does This Paper Address?
In the era of traditional TV, "stopping cues" were built-in: a show ended, a commercial break started, or the nightly news came on. Modern streaming platforms like Netflix have systematically removed these cues through features like auto-play and personalized recommendations.
The authors argue that past communication research has been obsessed with the selection phase (why we hit play) but has largely ignored the disengagement phase (why we hit stop). By testing the AMUSE model (Appraisal of Media Use, Self-Control, and Entertainment), the study seeks to understand the "tug-of-war" between the hedonic pull of entertainment and the self-regulatory push of our long-term goals (like sleep or work).
Methods
The researchers employed an event-contingent experience sampling method (ESM) combined with digital trace data. This is a significant step up from traditional "recall" surveys where participants try to remember how they felt days ago.
- Tracking: Participants installed a Chrome browser extension called BWDAT that monitored their Netflix usage in real-time.
- In-situ Surveys: The extension triggered short surveys (averaging 33 seconds) at three critical moments:
- Pre-use: When the Netflix tab was opened (measuring baseline self-control).
- Within-use: Every time an episode finished (measuring enjoyment, appreciation, guilt, and goal conflict).
- Post-use: When the tab was closed (confirming disengagement).
- Analysis: They used Random Effects Within-Between (REWB) models to distinguish between how a person generally behaves (between-person) and how their feelings fluctuate from episode to episode (within-person).
Datasets
The study tracked 205 adult Netflix users (primarily university students in the Netherlands and Germany) over a two-week period.
- Observations: The final dataset included 1,519 pre-use surveys, 1,924 within-use surveys, and 1,022 post-use surveys.
- Behavioral Data: The average session lasted 1.71 episodes, though 63% of sessions were just a single episode.
Key Insights and Ideas
- Enjoyment vs. Appreciation: The study found a fascinating divergence. Enjoyment (feeling "fun" or "suspense") acted as a "pull" factor, making users less likely to stop. However, Appreciation (feeling the content was "meaningful" or "thought-provoking") actually acted as a "push" factor, increasing the likelihood of stopping. The authors suggest that appreciation might be more cognitively taxing, making users feel "full" or satisfied.
- The "Spoiled Pleasure" Effect: Guilt didn't always lead to an immediate stop. Instead, it worked indirectly by "spoiling" the fun. High levels of guilt were associated with lower enjoyment, which eventually weakened the "pull" to keep watching.
- Self-Control as a Filter: The study distinguished between self-control capacity (having the mental energy) and motivation (wanting to stop). Interestingly, people with higher self-control capacity were better at stopping even when they were enjoying a show. Conversely, those with high motivation to control themselves but low capacity often stayed "pulled in" by the enjoyment.
- The Threshold of Guilt: In the main analysis, guilt didn't directly predict stopping. However, in an exploratory analysis of longer sessions (2+ episodes), guilt did become a significant predictor of disengagement. This suggests that guilt needs to accumulate to a certain threshold before it overcomes the momentum of a binge-watch.
Key Results
- Enjoyment: Negatively related to disengagement ($b = -0.12$). The more you like it, the less you want to stop.
- Appreciation: Positively related to disengagement ($b = 0.11$). Meaningful content helps you find a "natural" stopping point.
- Guilt and Enjoyment: Guilt significantly lowered enjoyment ($b = -0.09$ within-person; $b = -0.17$ between-person), supporting the "spoiled pleasure" hypothesis.
- Self-Control Moderation: For individuals with high self-control capacity, the "pull" of enjoyment on staying was significantly weakened ($b = -0.34$ for the interaction).
Limitations and Future Work
- The "Observer Effect": The authors acknowledge that being prompted with surveys after every episode creates "friction" that doesn't exist in a normal Netflix binge. This might have made participants more self-aware and likely to stop earlier than usual.
- Browser Limitation: The tracking only worked on Chrome. If a participant switched to a smart TV or a mobile app, that data was lost.
- Sample Bias: The sample was 82% women and 90% students, which may not represent the self-control struggles of the general population.
- Future Directions: The authors suggest that future research should look at "addictive" design features (like the 5-second countdown to the next episode) to see how they specifically interact with these psychological appraisals to override self-control.
Summary updated