Based on the results, the researchers say there are four categories of social media information that can predict 23 different "subscales" of personality. Voice and "network functions" (e.g. number of tweets, number of followers, number of replies, mentions, etc.) appeared to be the strongest predictors.
"We confirmed our hypothesis that the network and word statistics information had unique strengths in predicting interpersonal traits such as autism and mental health traits such as schizophrenia and anxiety," the study's research found. "We also found that all four types of social network service information are predicted to have intelligence."
"Our analysis has shown, for example, that people with high verbal intelligence tweet frequently and are more preferred," the authors write. "These people also showed stable timing in their responses."
Network information also predicted things like extraversion well, while timing was linked to things like intelligence and anti-social behavior.
Mental health predictors: "Emotional words had strong predictive power, with more negative and less positive words used to predict delusion, depression, and anxiety," they write.