pattern is essentially something formal and non-intellectual

By controlling these other features – such as the follower count, verified accounts and the use of multimedia in tweets – the authors set out to assess whether the use of sentiment terms affected the probability of retweets. To quantify the effect of negativity and positivity they employed three Spanish sentiment lexicons. Negative sentiment, at least when it comes to the subject of politics, increased the chances of retweets, the authors found, but acknowledged that other features did play a role in predicting tweet virality.

{ The Guardian | Continue Reading }