Information Cascades in Social Link Sharing
A study at the University of Southern California was performed to demonstrate the spread of information such as news links and photos through social media sites. The purpose of observing the spread of news through social websites is that people are sharing information with those whom they are friends with, which suggests that they have similar interests, and in turn would choose to spread the same information. Meaning, if someone were to see their friend posting a link to a news site, they would be more likely to post it, as the information would relate to them as well. The sites used in the study are Twitter and Digg, while both provide an outlet to share and comment on links, Digg is more specifically made for the sharing of news articles while Twitter is a more general social networking site. Both sites can provide a large amount of data and can also serve to demonstrate how news navigates through different structures of social networks. The researches took data by following a particular popular story and observing the progression that it made. Additionally, as both Twitter and Digg provide timestamps for the information posted, the researchers were also able to analyze how the timeline of information sharing can vary throughout different networks.
The data demonstrated that a story on Digg spreads faster, while one on Twitter spreads further, meaning it receives a greater number of posts. Researchers concluded that the faster spread on Digg is due to the site being an outlet for news, and the majority of users that use it will choose to share a link as soon as they see it, while the explanation for the further spread of information on Twitter is that there is a larger number of users who can spread links.
This study demonstrates an example of an information cascade and how social connections can affect the way that news spreads. An information cascade occurs when a person makes decisions or behaves like another, despite possibly disagreeing with their actions. Cascades are more likely when the information one receives is limiting, meaning they do not have all the knowledge needed to make a decision, and will act with the little that they have. This study exemplifies an information cascade, as people will be more likely to make decisions to agree with and share an article of news when they see their friends posting them as well. While they might disagree with the opinion being shared after independently exploring the subject more fully, they will most likely accept that which a friend or acquaintance with similar ideals believes as well. Although a person might not immediately share a link to a news article that they find online, observing information posted by a friend on a social networking site can influence them, which is where the role of information cascades in social networks can be observed.
http://www.aaai.org/ocs/index.php/ICWSM/ICWSM10/paper/viewFile/1509/1839