Mills, Richard (2009) Social news web-sites with democratic interfaces : analysis of one month’s voting from Reddit.com. Masters thesis, UNSPECIFIED.
R_Mills_Mres_Applied_Social_Statistics_2009_Dissertation.pdf - Accepted Version
Available under License Creative Commons Attribution.
Download (5MB)
Abstract
This study aims to develop an understanding of the activity taking place on “Social News” websites which use quantitative democratic interfaces. The data analysed were 3,446,522 votes from one such site (Reddit.com) for a single month. Data were analysed in terms of Users, Links, and Sub-Reddits. Exploratory analyses revealed that exponential distributions dominate many facets of activity on the site – conforming to the Power-Law observed by previous research. Users, Links and Sub-Reddits were Partitioned around Medoids to determine if there were different “types” of each. Clusters fitted to Users data suggest that Users tend to take on different roles in the community by prioritising a certain kind of activity (i.e. voting or submitting). A concept of “community involvement” was found useful in describing the different types of User on the site. Furthermore, Row-Column association models suggested that the users who were the most active and involved were the most likely to submit popular content. Latent trajectory analysis was also employed to look for patterns to the temporal distribution of votes on Links. The potential for quantitative democratic interfaces to facilitate communication between large groups of people is discussed; and some proposals for how these systems might be further studied are put forward.