In response to my earlier post about Myers-Briggs (where I suggested a more detailed notation for more nuanced communication about personality types), it was pointed out that there is some correlation between the four traits being measured, and this makes the system communicate less information on average than it otherwise would (The traditional notation would communicate 4 bits, my version would communicate ~9.2 if there was no correlation).
I do object to the characterization that it all measures “the same thing”, since none of the traits perfectly predicts the others, and all 16 of the traditional configurations have people they describe (though some are more common than others); but I do think it makes sense to try to disentangle things—if the I / E scale is correlated with the J / P scale, we can subtract some amount of J points for more introverted people, and add J points for the extroverts, so that an introvert needs to be more “judgmental” to be considered “J” than an equally judgmental extrovert, with the goal being that 50% of extroverts will be J, 50% P, and have a similar 50-50 split for introverts.
By adjusting for these correlations across all pairs, we can more finely detect and communicate the underlying traits that cause “Judgement” and “Perception” that aren’t just a result of a person being more extroverted (a disposition that rewards those who are best able to use their intuition) or introverted (which often leads to pursuits that require careful thinking distanced from our whims).
In response to my earlier post about Myers-Briggs (where I suggested a more detailed notation for more nuanced communication about personality types), it was pointed out that there is some correlation between the four traits being measured, and this makes the system communicate less information on average than it otherwise would (The traditional notation would communicate 4 bits, my version would communicate ~9.2 if there was no correlation).
I do object to the characterization that it all measures “the same thing”, since none of the traits perfectly predicts the others, and all 16 of the traditional configurations have people they describe (though some are more common than others); but I do think it makes sense to try to disentangle things—if the I / E scale is correlated with the J / P scale, we can subtract some amount of J points for more introverted people, and add J points for the extroverts, so that an introvert needs to be more “judgmental” to be considered “J” than an equally judgmental extrovert, with the goal being that 50% of extroverts will be J, 50% P, and have a similar 50-50 split for introverts.
By adjusting for these correlations across all pairs, we can more finely detect and communicate the underlying traits that cause “Judgement” and “Perception” that aren’t just a result of a person being more extroverted (a disposition that rewards those who are best able to use their intuition) or introverted (which often leads to pursuits that require careful thinking distanced from our whims).