That’s a good point, reading Carl Jung is in my todo list. I generated an personality embedding from data and only mapped it to MBTI and the Big 5 out of personal curiosity rather than using it for work. The main surprise I got from that work was just how genetic personalities are.
It would make sense that the E/I axis would be clearer if early childhood experiences were used. That would make me a clear extrovert. I think the questions should be updated for that context or at least have some additional framing around the test.
> The main surprise I got from that work was just how genetic personalities are.
This idea was surprising to me too, but when I thought about it I realised it shouldn't be:
- We know that different people frequently react differently to the same stimuli (and that this is true even of very young children).
- Basically everything else in the human body shows genetic variation. It would be weird if the brain didn't.
- There is a strong evolutionary rationale for a population having a variety of complimentary congnitive styles.
I think the MBTI is most interesting is it taken to be only the genetic component of personality (providing us with AFAIK the only theory of the structure of that aspect of personality), while simultaneously leaning into the idea that personality is modified by nurture and environment (but that these changes are not described by one's MBTI type).
> That’s a good point, reading Carl Jung is in my todo list. I generated an personality embedding from data and only mapped it to MBTI and the Big 5 out of personal curiosity rather than using it for work. The main surprise I got from that work was just how genetic personalities are.
Could you possibly expand on the details of "generated an personality embedding from data"?
The data was a huge amount of complete url history and quite a lot of feature engineering goes into developing the features. Mix of Gaussian processes, random forests, GDBT, and a lot of hand coded stuff using regexes and state machines. To get the embedding the features were fed into a multi layered RBM auto encoder. Despite quick initial success of the embedding I was unable to evangelize deep learning to the org and left out of frustration. Andrew Ng, before he was famous, visited us to evangelize deep learning and my colleagues just berated him with their views about how much better Bayesian is, it was embarrassing.
It would make sense that the E/I axis would be clearer if early childhood experiences were used. That would make me a clear extrovert. I think the questions should be updated for that context or at least have some additional framing around the test.