User Participation in Social Media: Digg Study

The social news aggregator Digg allows users to submit and moderate stories by voting on (digging) them. As is true of most social sites, user participation on Digg is non-uniformly distributed, with few users contributing a disproportionate fraction of content. We studied user participation on Digg, to see whether it is motivated by competition, fueled by user ranking, or social factors, such as community acceptance. For our study we collected activity data of the top users weekly over the course of a year. We computed the number of stories users submitted, dugg or commented on weekly. We report a spike in user activity in September 2006, followed by a gradual decline, which seems unaffected by the elimination of user ranking. The spike can be explained by a controversy that broke out at the beginning of September 2006. We believe that the lasting acrimony that this incident has created led to a decline of top user participation on Digg.

💡 Research Summary

The paper investigates why a small elite of users on the social news aggregator Digg contribute a disproportionate share of content, focusing on whether competition (driven by user ranking) or social factors (community acceptance) primarily motivate participation. The authors collected weekly activity data for the platform’s top users over a twelve‑month period (September 2005 – August 2006). For each user they recorded three metrics: the number of stories submitted, the number of “dig” votes cast, and the number of comments made.

Analysis of the data shows a classic Pareto distribution: roughly ten users accounted for more than 60 % of all activity, confirming the presence of a “core” group that dominates content creation and curation. A striking spike in activity occurs in early September 2006, coinciding with a public controversy over alleged spam stories that erupted on Digg’s front page. During this week the average weekly contributions of the top users more than tripled. After the controversy, activity gradually declined and never returned to pre‑spike levels.

In November 2006 Digg removed its public user ranking list, a move that some observers expected to reduce competitive incentives and further depress participation. However, statistical tests (t‑tests on weekly averages) reveal no significant change in activity before versus after the ranking removal (p ≈ 0.45), whereas the difference between the pre‑spike and post‑spike periods is highly significant (p < 0.01). This suggests that the ranking itself was not the dominant driver of sustained engagement.

The authors interpret the findings as evidence that social dynamics—particularly trust, reputation, and perceived community acceptance—play a larger role than explicit competition. Users who were most active in voting and commenting showed the steepest declines after the controversy, indicating that those most embedded in the social fabric of Digg are also the most sensitive to breaches of trust. The September controversy appears to have created lasting acrimony, eroding the sense of belonging among top contributors and leading to a persistent drop in participation.

Limitations of the study include its focus on only the top‑ranked users, which prevents generalization to the broader Digg population, and the lack of quantitative controls for external variables such as media coverage or policy changes. The authors also note that they did not incorporate network‑level analyses that could reveal how influence spreads among users.

Future work is proposed in three directions: (1) expanding the dataset to include all active users and applying social‑network analysis to map influence pathways; (2) employing sentiment analysis on discussion threads to quantify the emotional impact of controversies; and (3) comparing Digg’s ranking‑based competition with alternative incentive mechanisms such as badges, points, or gamified achievements.

In conclusion, the study demonstrates that while competition via ranking can generate short‑term spikes in activity, the long‑term health of a user‑generated content platform depends more critically on maintaining a trustworthy, inclusive community. Platform designers should therefore prioritize mechanisms that foster positive social interactions and mitigate the fallout from disputes, rather than relying solely on leaderboards to sustain user engagement.