Clemens Stachl, Timo Koch, and Peter Romero published an article in Computers in Human Behavior, titled “Age and gender in language, emoji, and emoticon usage in instant messages”

What can you find out about a person solely through a text message? Apparently, a lot! The authors of this study analyzed instant text messages and were able to predict users’ age and gender above baseline-levels. Text is one of the most prevalent types of digital data that people create as they go about their lives. They identified indicators such as using informal language, functional words like 1st person singular pronouns, or even the usage of emoji and/or emoticons to profile an individual. This research builds on previous studies that analyzed texts of social media posts and most of the findings were congruent. This paper raises awareness on the controversial possibilities in appropriating user information through instant messages. It sheds a critical light on the shift of big tech companies like Facebook towards loosening up their privacy protection policies of their messaging services.

The research has been already featured in @el_pais.