Code has been found in Facebook’s Android app which could potentially allow likes about a post to be hidden
An engineer has uncovered what she is calling a Facebook experiment, which could potentially allow the firm to hide the amount of likes a post receives.
The engineer in question, Jane Manchun Wong, made a blog posting about the “prototype” feature that has been found hidden in code for the Facebook Android app.
Last year Facebook was forced to deny media reports that it had surreptitiously collected call and text data from Android phones.
According to Wong, since it was discovered that Instagram had a prototype it is testing in a number of countries to hide the public like counts of photos, Facebook has followed suit and begin working to hide the public like counts as well.
“I observed that Facebook has recently begun prototyping this hidden like/reaction count feature in their Android app by reverse-engineering the app and playing with the code underneath,” wrote Wong.
“Currently, with this unreleased feature, the like/reaction count is hidden from anyone other than the creator of the post, just like how it works on Instagram,” wrote Wong. “The list of people who liked/reacted will still be accessible, but the amount will be hidden.”
“By hiding the like/reaction counts from anyone other than the post creator, users might feel less anxious about the perceived popularity of their content,” wrote Wong. “Studies have shown that social media use may influence mental health, including leading to depression and anxiety.”
In 2013, researchers from the University of Cambridge claimed that much could be deduced from looking at Facebook Likes, including sexual orientation, race and age.
Likes are publicly open by default, so privacy advocates are concerned about how much Facebook information other people and advertisers can access outside of what users allow.
Cambridge’s Psychometrics Centre, in collaboration with Microsoft Research Cambridge, had analysed a batch of over 58,000 US Facebook users who had volunteered to take part.
Between 65 and 73 percent of the time, the researchers were able to predict relationship status and substance abuse from Facebook Likes.
Their models, which fed Likes into algorithms that were corroborated with information from profiles and personality tests, were accurate 95 percent of the time in distinguishing African-American users from Caucasian Americans.