Emotional Contagion on Facebook
In recent years, the negative effects of social networking has become a wider topic for discussion. Commentators both inside and outside of tech are critical of social networking sites and how they might be altering our behaviour. From misinformation to emotional contagion, critics from a range of fields are voicing their concerns about social media.
Many of us continue using social media sites despite these warnings. We’re convinced that our emotional and psychological health are unaltered. Like tuning out advertising, we’re able to tune out the platform and be our authentic selves online. Whilst making a conscious effort to keep our emotions in check, can we be sure that we’re expressing ourselves online as we intend to?
A grand experiment
A 2014 study by Facebook’s Data Science team suggests that the platform itself can influence how we express ourselves. The team were trying to find out whether emotional states can be transferred from one user to another via the social network. By experimenting with the news feeds of over 689,000 users, the researchers were able to establish that textual content alone was enough to transfer mood from one user to another.
The experiment analysed words in status updates from friends of the users being observed. For one group of users, posts that featured negative words were omitted from the feed, essentially reducing the amount of negativity in their news feed. The user could still seek out the post by viewing the poster’s profile, but the content would not be found in the news feed. The reverse was applied to another group of users, where positive words, and therefore positivity, were reduced.
The posts by the users in these groups were then analysed and a consistent trend was observed. Over the following days, those users exposed to fewer positive posts, themselves created posts containing fewer positive words. The same was also true for users exposed to fewer negative posts, their subsequent posts were less negative.
The researchers pointed out that the difference was relatively small, but due to the shear numbers of users being researched and the volume of posts being created, this is still a significant difference. They highlight that in early 2013, the difference would have “corresponded to hundreds of thousands of emotion expressions in status updates per day.”
Isn’t it just empathy?
It can be unsettling to admit that the words and emotions of others can change how we express ourselves online. And it could be tempting to try to explain this away as human empathy among friends. However the research group took care to observe the users over a number of days. The intention was to avoid observing users reacting to shared news or a single event.
Whilst we may insist our online expressions are our own, this study proves that our mood can be altered by what we see. To some extent, we reflect what we see in our feeds, back on to the platform. Interestingly, the study found the effects were equal for both positive and negative emotions. The study also found there was no evidence of negativity bias. This despite the researchers’ expectations that negative content could yield a greater proportion of negative posts.
The emotional fallout
When the press learned of the research, the response was generally negative. There was an air of alarm that there was a team at Facebook experimenting on their users. In response to feedback on the study, Facebook said “We are taking a very hard look at this process to make more improvements.”
This is Facebook essentially vowing to rein in the reach of the Data Science team’s research. There is an argument for this as whilst Facebook can see the user’s emotional response altered in their Facebook posts, there’s no record of the mood or actions of those users offline.
Although the posts were determined as positive or negative by automated textual analysis (i.e. researchers were not manually determining what posts were hidden from users), those of us learning that Facebook could make such alterations to our news feed were unnerved by the experiment.
The study does state it was consistent with Facebook’s Data Use policy, which users agree to on creating an account and therefore consent to this type of research. In 2021, it’s arguably common knowledge that the news feed is an algorithm-driven filtered feed. At the time of the study, there may have been many users who either expected or believed their news feed to be a chronological list of all posts from their friends. In reality, this has never been the case since Facebook first launched the news feed feature in 2006.
For many years, Facebook has filtered the News Feed due to the shear volume of posts on the platform. This isn’t unique to Facebook, Instagram and Twitter are also well-known for their algorithmically curated feeds. Many users protest and content creators scramble to understand what metrics alter a post’s performance in the news feed. For the meantime, the complex workings of the news feed will remain a mystery.
Emotions in the real world
No matter what the response to the study, it does prove that emotional contagion on Facebook is occurring. It’s no stretch of the imagination to expect this is happening on all social networks. This could be particularly likely on text-driven social networks such as Twitter.
Although the effects are small, it’s clearly important for us to consider who we surround ourselves with online. Just like a common cold, we can ‘catch’ emotions from our friends. And whilst the study only observes our posts to Facebook, there’s no telling how our behaviour, mood and expressions are affected offline, in the real world.