Science and Facebook – the good, the bad and the (very?) ugly
Facebook – usually a place for cat videos, holiday snaps and shared buzzfeed listicles. But sometimes, the social media platform appears to not only be conducive to procrastination, but also to science. Like the page itself, the experiments are somewhat controversial. Here are three scientific insights involving Facebook – along a spectrum of the good, the bad and the ugly.
During a hike in the province of Minas Gerais, Brazil, amateur botanist and orchid grower Reginaldo Vasconcelos took photos of a couple of plants. And, of course, posted them on Facebook. Three years later, he is the discoverer of a new plant species. And not just any, but of the second biggest carnivorous plant in South America, which has now been described in the journal Phytotaxa.
The “magnificent sundew”, Drosera magnifica, is the first plant species to be discovered on Facebook.
Experts realized that the photographed plant was a new species. It’s amazing that it was only discovered in 2013. Not only is the plant, as her name would suggest, “magnificent” and hard to overlook with a length of 1.5 meters. But the authors also describe the mountainous spot where it was discovered, the Pico Padre Angelo, as not particularly remote.
Drosera magnifica catches insects up to the size of fireflies win a tangle of 24 cm long, threadlike tentacles. Its glistering, sticky tentacles and leaves enclose the prey until it suffocates in the slime. Through the nutrients captured, the sundew can compensate for the lack of nutrients in the ground.
Although the magnificent sundew has only been discovered recently, it is already threatened. It is only found on a single mountain top, which is surrounded by coffee and eucalyptus plantations as well as cattle farms. Invasive plant species grow along the path to the summit, which is not protected by national park or reserve. Drosera magnifica was probably discovered just in time before its extinction.
A psychology experiment on Facebook, which manipulated the emotional content of the newsfeed, caused quite some turmoil. For nearly 700,000 users, Facebook filtered this overview of comments, videos, pictures and links by others in their social network without their consent, as they described in 2014 in the journal PNAS together with scientists of Cornell University and the University of California.
For one part of users, Facebook reduced the “positive emotional content”. These users themselves posted less positive content. For the other part, the amount of “negative emotional content” was reduced, they in turn posted less negative content. The study’s conclusion: our friends emotions can spread via “emotional contagion” through social networks, without personal contact or non-verbal cues.
A further conclusion for Facebook is likely to have been that manipulation without consent can cause shitstorms. No user knows whether they have been, or still are, a guinea pig for Facebook. That Facebook tests how content is presented in the newsfeed was probably clear to most users as this is one of the company’s ways to maximize its main product, ad clicks. The shitstorm’s real reason was probably that this experiment gave us users a clear insight into what Facebook can really achieve: to manipulate our emotions, deep inside our private sphere.
… and the ugly.
The, in my opinion, “ugliest” Facebook experiment also belongs to this category. During the 2010 midterm elections, scientists from the University of California tested together with Facebook whether it could get its users to vote (as described in a study in Nature).
In this experiment, all 61 million Americans of voting age who visited Facebook on the election day served as subjects.
They were divided into three categories. A group of about 60 million users received a message “Today is election day”. This contained a button “I Voted”, a link to information about the election and polling places, as well as photos of up to six friends who have already clicked “I voted”. A second group of about 600,000 users also received a message to go and vote, as well as information on polling places. They however did not know whether friends had already voted. A control group of also about 600,000 users did not receive any such message.
The scientists then analyzed the voting records to see who had actually cast a vote. Users in the first group voted slightly more often than those in the other groups, with around 60,000 additional cotes cast. The appeal to vote probably also spread via “emotional contagion”. Users in the first group also clicked “I voted” more often. This also caused their friends to go and vote, leading to probably an additional 280,000 votes cast. The scientists also speculate that their experiment lead to more additional votes cast, as they could only verify those votes where usernames correspond to actual names.
340,000 additional votes caused by Facebook don’t sound like much when faced with the entire US electorate of 96 million (in 2010). The study’s authors also assume that both republican and democrat voters reacted similarly to the call to vote. It’s not hard to imagine that Facebook could have the power to significantly influence the outcome of an election. It hardly requires complex algorithms to draw conclusions about our political views based on our posts, friends and groups. With an appeal to vote directed at only the supporters of one party, Zuckerberg and co could easily get additional votes cast. And George W. Bush won the 2000 presidential election with a margin of only 537 votes in Florida…