Techno Telepathy: How Innocuous Data Points Can Predict our Deepest Secrets
Behavioural Science algorithms can predict your secret prejudices, insecurities and even tastes in porn
Using data on the types of books, films, music and television that people like, they were able to predict things that we usually try to keep to ourselves – prejudices, insecurities, and tastes in pornography. The study also investigated the underlying role of personality. The findings highlight the massive power of behavioural science and predictive analytics to read people’s minds, the opportunity and the ethical imperative.
As an illustration, the researchers could predict whether a respondent watched Hentai (sexually explicit Japanese manga or anime) to an 84% accuracy, simply by asking for media preferences. People who watched Hentai were more likely to enjoy animated TV, foreign language films, and science books (among other things).
Other findings include:
People who enjoy philosophy books are 21% more likely to be prejudiced against Brexiteers
People who enjoy action movies are 18% more likely to be insecure about going bald
People who enjoy reality TV shows are 11% more likely to watch VR/3D pornography
The study was led by Patrick Fagan, Capuchin’s Chief Scientific Officer and former Lead Psychologist at Cambridge Analytica. Fagan said, “Every time we use technology, we leave an invisible data footprint in the websites we visit, products we buy, and content we watch. This data might seem bland but it actually gives a lot away about our hidden psychology – our fears, hates, and desires. This gateway into the subconscious is a very powerful technique for understanding and influencing people.”
The research rests on the psychological principle of ‘thin slices’. Outward behavioural cues, like the music you listen to on Spotify, are the result of underlying personality traits, which can be used to infer behaviours in other areas of life. Somebody with liberal political views is more likely to listen to jazz, for example, because being open to new things predicts both. We found that intimate behaviours were linked to personality and attitudes – for example:
Watching BDSM was linked to higher levels of agreeableness
Watching cuckolding was linked to being a Conservative
Watching gangb**** and sub/dom porn was linked to wanting to remain in the EU
Watching live cams, casting couch, and celebrity porn was linked to being an extravert
Overall, the findings support that data can be used to predict the most intimate things, can be used by companies to understand and address their audiences much more effectively. However, in light of the massive pools of data owned by some companies, this is a sensitive area. As the ongoing Facebook / Apple contretemps and the exodus of users to more private platforms like Telegram show – the need to use data ethically has never been more urgent.
“Data science and behavioural science are powerful but morally neutral tools. Just like splitting the atom can be used to power cities or destroy them, predictive analytics can be a force for ill or good. The data must be used ethically - by being transparent, giving people the ability to opt out, and listening sincerely to users’ concerns,” said Fagan.
The research was conducted online with a nationally representative sample of 2,000 British adults, who completed a self-report survey on their media preferences, personality, and ‘secrets’. All respondents provided informed consent and participation was entirely voluntary and anonymous. People who gave inconsistent, incorrect and/or socially desirable responses were excluded from the analysis.
About Capuchin Behavioural Science
Capuchin is a behavioural science-based research and consulting agency based in London, UK. They help clients understand the psychology of audiences, make brand propositions more resonant, and design engaging communications and effective nudges.
Learn more about how we work:
Get in touch and find out how we can help you.