MIT researchers built a theoretical model to study how news spreads on a Twitter-like social network and found that when a network is highly connected or when the views of its members are sharply polarized, false news will spread wider than news that is seen as more credible.