Can we actually show that facebook influences elections?

More and more people get their news from social media sites nowadays. This fact combined with the suspicion that the facebook site has had a notable impact on the 2016 US elections and with the 2020 elections around the corner, make the linked news article quite interesting.

As anybody in research will tell you, in order to establish a causal effect (does facebook influence elections?), any experimental research design needs to have a control and experimental group, but with just about everybody using Facebook and/or Instagram, it is rather difficult to establish the real influence as there is no real “comparable” control group.

So, Facebook has taken it upon themselves to create a control group:

According to international media reports, the company has decided to ascertain the role of social media giants and their influence and therefore Facebook has decided to offer $120 to its users to deactivate their accounts at the end of September.

It will need to closely monitor any other behaviour participants have around news gathering (do they now use twitter, youtube, … instead?), but it is an interesting angle. I am curious about the results of this research project, but do wonder how much will be shared publicly.

Source (Republic World): Facebook To Pay US Users To Stop Using Its Services Until Elections

Mastodon