Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

In Defense of Myers Briggs (2020)

dynomight.net

Post reply on HN
Print view View original

61–70 of 352 posts

Re: In Defense of Myers Briggs (2020)

#61

Post by sunstone »

Xcelerate wrote: post #19

I've never really understood the harsh criticisms of MBTI. It's not super predictive, but not to be honest, nothing else in psychology is either; the whole field is just not anywhere near the level of physics or chemistry (not to blame the researchers; predicting humans is very hard). The big 5 and I think more recently an 18 dimensional breakdown (using PCA or ICA?) just aren't that much better than MBTI, despite ma…

The best that can be said of MBTI is that it has a slight correlation with one of the Big 5 dimensions. So it's not useless but it's close to useless. On the other hand it's expensive whereas the Big Five has been validated and it's free. Really MBTI is just a waste of time and money.

Re: In Defense of Myers Briggs (2020)

#62

Post by forgotmypw17 »

Summary: Myers-Briggs is a useful model, when a few adjustments are made.

1. Problems I've Discovered

1a. Myers-Briggs pigeon-holes people into one "personality type", and generalizes what they should think and do based on that. I think that's kind of silly and limiting.

1b. If you read the personality descriptions, it is favorable towards the "rare" types, describing them as brilliant geniuses, while the other types are portrayed as barely sentient worker drones. I think that's really condescending and divisive.

2. Adjustments I've Made

Nevertheless, I think the questions and the other text is a USEFUL tool for SELF-ASSESSMENT and also for UNDERSTANDING others, so I don't want to throw the baby out with the bathwater.

Here is how I make Myers-Briggs make sense and fit into my other mental models:

2a. First, instead of four scales, I consider it an EIGHT-SCALE system, with each trait being its own measure. For example, you can be high on the introversion scale (needs alone time to recharge) and at the same time also be high on the extraversion scale (connects with others readily, needs socializing to thrive.) I think the worldwide quarantine made this obvious to many an self-described introvert.

This applies to all the other scales as well. You can be someone with intense feelings and at the same time a thorough and dedicated thinker, who sits down and combs their mental models for hours. You can be very good at listening to your intuition while also being highly perceptive to your senses. And you can have firm general rules about how things "should be", especially as one gets older, while still remaining to new possibilities.

2b. The other adjustment I have made is that I consider each scale to be a measure of how much ABILITY a person has developed and PRACTICING that trait. In other words, beyond natural predisposition, each person can grow considerably in any of these traits, provided they practice and apply it regularly.

Conclusion: People are neither introverts, nor extraverts, but both. People are both thinkers and feelers, both judgers and perceivers, and both intuiters and sensers.

Re: In Defense of Myers Briggs (2020)

#63

Post by DiscountChocula »

Xcelerate wrote: post #19

I've never really understood the harsh criticisms of MBTI. It's not super predictive, but not to be honest, nothing else in psychology is either; the whole field is just not anywhere near the level of physics or chemistry (not to blame the researchers; predicting humans is very hard). The big 5 and I think more recently an 18 dimensional breakdown (using PCA or ICA?) just aren't that much better than MBTI, despite ma…

> There was once a poll of engineers at the tech company I work at, and I think the results were like 70% INTJ, 25% INTP, and 5% other. If MBTI lacked all predictive power, you would expect a distribution roughly equivalent to that of the general population.

You're assuming the types are proportionally represented in the general population, which we have no reason the believe even if the test works as advertised. For all we know urban environments are 70% INTJ.

Re: In Defense of Myers Briggs (2020)

#65

Post by jbay808 »

Xcelerate wrote: post #19

I've never really understood the harsh criticisms of MBTI. It's not super predictive, but not to be honest, nothing else in psychology is either; the whole field is just not anywhere near the level of physics or chemistry (not to blame the researchers; predicting humans is very hard). The big 5 and I think more recently an 18 dimensional breakdown (using PCA or ICA?) just aren't that much better than MBTI, despite ma…

In grad school I joined an extracurricular social society where you meet around a table with other grad students and discuss various topics selected in advance while eating pie and other goodies. Sort of like debate but without the structure or positions.

Anyway one time the topic was Myers Briggs and so we all filled out our profiles in advance so we'd have something to discuss.

It was a group of 20 people, and we discovered that 19 of us were INTJ. The other was INTP. That was when I decided the test has strong predictive power, for the same reason you did. Even if it's just able to predict a preference for discussing personality tests with a plate of cherry pie. INTJ has a background rate of about 2%.

Re: In Defense of Myers Briggs (2020)

#66

Post by jbay808 »

DiscountChocula wrote: post #63
Xcelerate wrote: post #19

I've never really understood the harsh criticisms of MBTI. It's not super predictive, but not to be honest, nothing else in psychology is either; the whole field is just not anywhere near the level of physics or chemistry (not to blame the researchers; predicting humans is very hard). The big 5 and I think more recently an 18 dimensional breakdown (using PCA or ICA?) just aren't that much better than MBTI, despite ma…

> There was once a poll of engineers at the tech company I work at, and I think the results were like 70% INTJ, 25% INTP, and 5% other. If MBTI lacked all predictive power, you would expect a distribution roughly equivalent to that of the general population. You're assuming the types are proportionally represented in the general population, which we have no reason the believe even if the test works as advertised. For…

> For all we know urban environments are 70% INTJ.

There are 16 categories, which means you'd expect about 6% in each if the test sorts evenly. It doesn't; in fact only between 2% and 3% of the population is INTJ. Unless one of the questions is literally "are you a tech worker", that's very conspicuous selectivity.

Re: In Defense of Myers Briggs (2020)

#67

Post by sosborn »

at_a_remove wrote: post #54
Quekid5 wrote: post #26

Earlier quoted context omitted.

I'm going to suggest a hypothetical. Let's ignore any personality test specifics just to generalize. Ok, so I have a candidate for a position, who shows +5% to "conscientiousness". What does that mean, practically, for me as an employer? It means exactly nothing, because that trait is so vague as to be ... meaningless. (That's just the start of the problems with these tests.) Even if it M-B (or whatever) is a repeata…

Imagine you have a job that is highly detail-oriented. Perhaps data must be perfect or a number of steps must be followed but checking the employee's work might be cumbersome. You have two applicants, roughly equal. One scores higher on conscientiousness ...

Do you believe that measurement is actually accurate?

Re: In Defense of Myers Briggs (2020)

#68

Post by sdenton4 »

jbay808 wrote: post #66
DiscountChocula wrote: post #63

Earlier quoted context omitted.

> There was once a poll of engineers at the tech company I work at, and I think the results were like 70% INTJ, 25% INTP, and 5% other. If MBTI lacked all predictive power, you would expect a distribution roughly equivalent to that of the general population. You're assuming the types are proportionally represented in the general population, which we have no reason the believe even if the test works as advertised. For…

> For all we know urban environments are 70% INTJ. There are 16 categories, which means you'd expect about 6% in each if the test sorts evenly. It doesn't; in fact only between 2% and 3% of the population is INTJ. Unless one of the questions is literally "are you a tech worker", that's very conspicuous selectivity.

On the other hand, a test with questions like 'are you a tech worker' might be a lot shorter and just as meaningful!

Re: In Defense of Myers Briggs (2020)

#69

Post by calt »

jbay808 wrote: post #66
DiscountChocula wrote: post #63

Earlier quoted context omitted.

> There was once a poll of engineers at the tech company I work at, and I think the results were like 70% INTJ, 25% INTP, and 5% other. If MBTI lacked all predictive power, you would expect a distribution roughly equivalent to that of the general population. You're assuming the types are proportionally represented in the general population, which we have no reason the believe even if the test works as advertised. For…

> For all we know urban environments are 70% INTJ. There are 16 categories, which means you'd expect about 6% in each if the test sorts evenly. It doesn't; in fact only between 2% and 3% of the population is INTJ. Unless one of the questions is literally "are you a tech worker", that's very conspicuous selectivity.

The test doesn't sort evenly. One of the most popular online test sites says in one of their descriptions - I forget which one - that "this is the rarest type with only x% of the population."

I had a female friend who was an INTJ, and told me that it was very rare among women. Being very, INTJ, she was referencing some rigorous statistics.

Re: In Defense of Myers Briggs (2020)

#70

Post by souprock »

sunstone wrote: post #61
Xcelerate wrote: post #19

I've never really understood the harsh criticisms of MBTI. It's not super predictive, but not to be honest, nothing else in psychology is either; the whole field is just not anywhere near the level of physics or chemistry (not to blame the researchers; predicting humans is very hard). The big 5 and I think more recently an 18 dimensional breakdown (using PCA or ICA?) just aren't that much better than MBTI, despite ma…

The best that can be said of MBTI is that it has a slight correlation with one of the Big 5 dimensions. So it's not useless but it's close to useless. On the other hand it's expensive whereas the Big Five has been validated and it's free. Really MBTI is just a waste of time and money.

Not that correlation with Big 5 should be a determining factor, but...

The chart in the article shows that E/I is strongly correlated with extraversion, S/N is strongly correlated with openness, T/F is moderately correlated with agreeableness (and weakly with conscientiousness), and J/P is moderately correlated with conscientiousness (and weakly with openness).

It's not a slight correlation, so if MBTI is useless then so is Big 5. Again though, choosing Big 5 as the standard is arbitrary and unfair. That automatically judges all other systems as inferior, without examining their merit.

Post reply on HN