Emotional Contagion Through Social Networks

Social networks are by no means a new phenomena, and track back to the origins of our species. Language and communication has been of utmost importance in evolving us into what we are today. Previously, most of our social interaction was physical, but in our time the trend is towards increasing online social interaction. In this essay we look at a study that measures some impacts of this trend.

In the paper Kramer et al. (2014), a team of researchers of Cornell University in collaboration with Facebook, researched the contagion of emotions through online social networks, namely Facebook. The experiments took place over the course of one week January 11-18, 2012.This was done on the basis that emotions can be spread in physical social settings, and thus examining how emotions that spread merely via written text with the removal of non-verbal cues (e.g. smiling, crying), could lead to some interesting results.There were 689,003 unknowing participants. In the study, the researchers altered what kind of posts that users would be exposed to in their news feeds.The participants were also selected randomly, resulting in 155,000 participants that posted at least one status during the week. Posts were determined to be either positive or negative if they contained one or more words, defined by the “Linguistic Inquiry and Word Count”  (LIWC) software, which correlates with self-reported and physiological measures of well-being. During the span of one week, two groups of users were analyzed. One of the groups were exposed to posts containing more positive words, and less negative words. The other group of users would be exposed to the opposite, namely posts containing less positive and more negative words.

It was then analyzed what kind of emotions the subjects would express after being exposed to a certain type of posts (i.e. negative or positive).

A part of the groundwork for the research were three other studies about social emotional contagion. One study had suggested that there were possibilities of emotional contagion in networks, but not proved the results for massive social networks. Another study researched whether non-verbal cues were necessary to produce emotional contagion or if verbal cues could be enough to make a difference, but failed to prove it. The last study came to the conclusion that when people were exposed to the happiness of other people, the reaction could actually be negative, and produce depressive thoughts. This suggests that people may feel more alone and negative when seeing that other people around them express happiness. Compared to this study, the research of Kramer concluded with the opposite results.

The results showed that when users were exposed to posts with less negative words, the individual user is less likely to post anything containing negative words themselves. Likewise, when exposed to posts with less positive words, their own expressions would contain less positive words. With this, one can conclude that the emotion of other users in your social network may affect your own emotions and the mood of your posts on Facebook. The research also tracked how many posts each user posted, giving an interesting result for the study.

“Approximately 3 million Facebook posts were analyzed, containing over a 122 million words, 4 million which were positive (3.6%) and 1.8 million negative (1.6%).” – Kramer, 2014

The numbers showed that when users were exposed to posts with few positive and negative words, in other words less emotional or expressive posts, the users themselves produced less posts than the users who were exposed to a news feed that were manipulated for positive or negative affection.

The research of Kramer also got some other results. First of all, it proved the opposite of the other research, showing that users do not necessarily become affected in a negative way when others express themselves positively. One of the issues with the study was that they were not sure if word only was enough to affect the emotions of the users, based on one of the studies in the groundwork. Instead, the research of Kramer showed that it is possible, and thereby that social media like Facebook may produce emotional contagion even if the communication happens online and not in-person.

Mean number of positive (Upper) and negative (Lower) emotion words (percent) generated people, by condition. Bars represent standard errors.
Mean number of positive (Upper) and negative (Lower) emotion words (percent) generated people, by condition. Bars represent standard errors.

Social networks is by no means a new phenomena, and tracks back to the origins of our species. Language and communication has been of utmost importance in evolving us into what we are today. Social networks in today’s society can probably be measured in several categories, but in order to simplify, we can consider the physical social networks on the one hand, and the non-physical ones on the other hand. The physical social networks are strictly bound to interactions between people in the physical realm, while the non-physical ones can take many shapes. The largest non-physical social network is Facebook, which has over 1.35 billion active users(Facebook, 2014). If we compare the number of test subjects in the Facebook Emotion Contagion Experiment and the total amount of Facebook users, the difference is vast. Should Facebook upscale the experiment, and exchange emotions with appropriate advertisement, we would probably see some differences, but nothing major at first.

In conclusion, the results suggest that even though the contagion effect is small per individual, it is still statistically significant, and on an even larger scale could perhaps an even more significant impact. It also raises the question of the power a social network has over its users, with the ability of manipulating what posts they see, in order to provoke a certain emotion. This may raise a possibility of the users being exposed to a certain type of content decided by advertisers, in order to provoke an emotion that makes the user more potential of buying the product.


Collins, N. (2014), How ‘Contagion’ Became Contagious. Pacific Standard Magazine, Downloaded from: <http://www.psmag.com/navigation/health-and-behavior/contagion-became-contagious-88732/>, Accessed: 07.11.2014


Kramer, A(2014), Experimental evidence of massive-scale emotional contagion through social networks, Downloaded from: <http://www.pnas.org/content/111/24/8788.full>, Accessed: 20.11.2014

Facebook, 2014,Company Info, Downloaded from: <https://newsroom.fb.com/company-info/>, accessed 29.11.2014.