Big Brother is watching you and you’re telling him everything he needs to know (if you’re on Facebook). Here’s why. A computer analysis of your ‘likes’ predicts the results of your completing a 100 item personality questionnaire, better than those whom you’ve friended on Facebook. [ Proc.Natl. Acad. Sci. vol. 112 pp. 1036 – 1040 ’15 ] Has the gory details.
We do know that people lie when completing such things and the MMPI (Minnesota Multiphase Personality Inventory) has a scale for lying. Apparently everyone steals from mommy’s purse at some point, and your lie score on the MMPI goes up if you say you never did.
The study used a mere 86,220 volunteers who completed the 100-item International Personality Item Pool (IPIP) Five-Factor Model of personality questionnaire, measuring traits of openness, conscientiousness, extroversion, agreeableness, and neuroticism. The sample used in this study was obtained from the myPersonality project. myPersonality was a popular Facebook application that offered to its users psychometric tests and feedback on their scores. The data was anonymized and is in the public domain. How normal such an individual can be I leave up to you.
Human personality judgments were obtained from the participants’ Facebook friends, who were asked to describe a given participant using 10 of the 100 items of the IPIP personality measure. E.g. the friends were filling out the 10 items as they thought the subject would (or as they saw the subject).
So it’s the same questionnaire. The paper pitted a computer algorithm based on your Likes to predict your IPIP responses against those of your so-called Facebook friends who presumably know much more about you than just your Facebook Likes. The algorithm won. It didn’t win by much. Computer-based judgments (r = 0.56) correlate more strongly with participants’ self-ratings than average human judgements did (r = 0.49). Surprisingly, neither did terribly well, but then we all know that our judgement of ourselves is usually rather different than others. It’s why city people often tell you what they’re ‘really like’, while Montanans don’t. They know that there are so few people around that they’ll see you again. Your long term behavior will tell them everything they need to know.
Update 31 Jan ’15 — I told the people I play piano trios with about the paper. The cellist (a retired Actuary) had an excellent explanation of why the algorithm was more accurate than the friends individually. See if you can think of the reason.
She notes that the 3 of us interact with each other individually, e.g. we act differently for each of our friends, exposing just the parts of our personalities we choose. They aren’t the same for everyone. Obvious, now that she’s thought of it (did you?)
As usual the Poets have said it better
And would some Power the small gift give us
To see ourselves as others see us!
It would from many a blunder free us,
And foolish notion:
Robert Burns (1786)