Psychologists are becoming increasingly good at predicting someone’s personality from their online behavior. In a 2015 study, researchers at the University of Cambridge and Stanford University developed a model that can predict personality traits using little more than a Facebook-like record. With 10 likes, the algorithm was just as accurate as a coworker. It was as accurate as a friend with 70 likes. It was more reliable than a partner of 300 likes. This idea is further pushed by new studies published in the Proceedings of the National Academy of Sciences. A team of information-savvy psychologists noticed that three of the Main Five dimensions of personality — extroversion, conscientiousness, and openness to experience — could be reliably derived from people’s smartphone use. The other two dimensions of the big five— pleasantness and emotional stability — remain concealed from view, at least for now. The Big Five model, also known as the OCEAN model, was first built in the 1980s with the help of the Government. It is the most commonly used and well-established psychological framework for organizing personality traits.
Data were collected on users’ contact and social behavior, music consumption, app use, accessibility, overall phone usage, daytime, and nighttime activity levels.
Some data streams have proven more beneficial than others. For example, in predicting the “love of order and sense of duty” component of the conscientiousness trait, a particularly useful data stream was the average battery level of the phone when it was not on the charging lifeline cable.
The reliability of these predictions is close to that observed for predictions based on digital footprints on social media sites, “the researchers wrote.
“[This] shows the possibility of collecting information on the private characteristics of individuals from behavioral trends passively obtained from their smartphones.”
The Princeton team has acknowledged the clear ethical issues.
“The present work serves as a harbinger of both the advantages and the dangers posed by the widespread use of smartphone-based behavioral data,” they wrote.
It might sound weird at first, but a team of researchers in Australia has come up with a way to predict your personality traits using just the sensor on your screen.
Well, that and the records of your call and messaging operation. Also, the method functions better with certain traits than others. But it’s fascinating to see how we can make links across such seemingly unrelated stuff.
There’s a range of previous research into how various aspects of your mobile apps and social networking use — such as your message language, how you style your Facebook profile, or how much regular exercise you do — can be used to determine your characteristics. “Activity like how fast or how far we’re going, or when we pick up our phones at night, always follows trends and these trends tell a lot about our personality style,” said one of the group, computer scientist Flora Salim from RMIT University in Australia.
This is the first time researchers have combined this sensor reading with conventional phone activity; to confirm such research, we will need to see further experiments performed similarly, ideally with larger sample sizes.