5 November 2015 at 16:53 #2309
“Fragile Online Relationship: A First Look at Unfollow Dynamics in Twitter”
This paper investigates the reasons behind a user’s decision to unfollow someone on the online social networking service, Twitter. The Authors try to understand why, by analysing both quantitative and qualitative data that they have collected.
The Authors Kwak, Chun and Moon focus on Korean-speaking Twitter users specifically and attempt to explore two main questions for research. The first asking what are the characteristics of unfollow behaviour and the second, simply put is why? To achieve the first, daily snapshots of follow relationships and tweets were taken of 1.2 million users over 51 days and the data analysed. The second involved interviews with 22 users in order to find what lay behind each of their decisions to unfollow someone. There were certain criteria that the Twitter accounts had to meet to be deemed Korean. They were:
1. User had written at least one tweet in Korean.
2. User’s biographical information was written in Korean.
3. User’s location was written in Korean.
4. User’s screen name was in Korean.
In 2011, when the paper was written, Twitter’s popularity was on the rise and it was established enough to be the subject of analysis. The unfollow dynamic narrows that area of investigation with the potential to offer interesting sociological insight. Although they do not explain deeper, the Authors acknowledge that cultural differences may play a factor in the data gathered. They note that reciprocity (as defined by them) among their Korean pool of users was almost three times that of the global Twitter network. An interesting point of note and an area worthy of further investigation.
The Authors look at a number of different issues that they suggest can affect whether a user remains following another user on Twitter. One of the main reasons put forward by the Authors for someone not choosing to unfollow another user is the reciprocity of a relationship. Such as whether a user’s passiveness in a ‘Twitter relationship’ is an indication of inevitable unfollowing. Another is timing, or rather the boredom of a relationship. Does the increased length of a relationship within Twitter increase the chances of one member unfollowing another? Followees’ informativeness is another factor in the unfollow dynamic, with the data catalogued accordingly depending on whether a tweet was retweeted or marked as favourite.
After the data was collected, the Authors held their interviews. These interviews allowed the Authors to ask questions they felt would add to the data further. The questions asked of the participants were as follows:
1. Why did they decide to unfollow?
2. If they thought that the unfollow was aware that they had been unfollowed.
3. Whether the participant would have behaved differently if unfollow actions were made known.
4. If the participant had ever ended a relationship with another online social network.
5. If corporate accounts were ever followed.
6. Asking the participants to choose ten users that they would never unfollow.
From the answers given by the participants, it seems that the main reason for unfollowing is ‘burst tweets’. These are tweets sent frequently in short spaces of time. The reason the participants disliked burst tweets was that they pushed tweets from other followees down the participants timeline causing them to have to sift through unwanted content to get the tweets they most wanted. A possible solution suggested by the Authors was to have Twitter include the ability to collapse and expand multiple tweets from one account. Reddit has a similar function for its comments. It has been requested but at the moment of writing, it has not come about. (Twitter Developer Forums, 2012)
The data was gathered by ‘crawling’ Twitter’s users and their actions. The paper does not explain what crawling is or what it involves so to the uninitiated, a lot of the data may seem shrouded in mystery or lack merit. However, the fact that this paper is written by members of the Computer Science department in KAIST, Korea, one would assume that readers are involved in the field and would be familiar with certain terminology and techniques. Likewise the use of various mathematical coefficients. To those more interested in the social aspect as opposed to the techniques used, the inclusion of interviews may offer more value to the research, rightly or wrongly.
The article does not explain if the closure of an account affects the data gathered or if it shows as a ‘mass unfollow’ action. They do make note of a significant increase in the number of followees over the course of the research. Although they do not give an explanation, my own theory is that during the 51 days of analysing accounts throughout the summer of 2010, the Korean football team were beaten out of the World Cup and the North Korean Government created an official Twitter account. These events alone could easily explain the spike in following behaviour stemming from a deep interest by Korean users both home and abroad. (Harlan, 2010)
Unfollowing someone in Twitter is severing the ties of a relationship. It’s probably seen as a negative act as opposed to a positive one but is it really? A positive move for the follower surely? To put an end to something that isn’t working for both parties? Neither the data analysis or interviews asked the question that after they unfollowed someone, if users attempted to substitute that relationship for a new one that could offer what the unfollowed had originally implied or had been assumed by the followee. Does unfollowing satiate the need for the information originally desired completely, or is it just the source of that information that is replaced?
The idea of the article was intriguing and for the most part, was very interesting. However, I feel that more interview subjects would have added further weight to the research. Although it was an article leaning neither one way or the other, all the data collected felt heavy, to this reader at least. I’m sure another would soak it up and add it to their own research. But my interest lay with the interview respondents putting a voice (or tweet) behind their decisions to unfollow.
Harlan, C. (2010, August). Retrieved October 2015, from Washington Post: http://www.washingtonpost.com/wp-dyn/content/article/2010/08/20/AR2010082005741.html
Twitter Developer Forums. (2012, June). Retrieved November 2015, from Twitter Community: https://twittercommunity.com/t/expanding-and-collapsing-tweets-on-timeline/2874
You must be logged in to reply to this topic.