(Repost from 2016)

1) —“The Myers-Briggs rests on wholly unproven theories”—
Well, it rests on observation of demonstrated motivations. So does all of psychology, and all of sociology, both of which are demonstrably pseudoscience created as pseudosciences by Boaz, Marx, Lenin and Trotsky, Freud, Cantor, Adorno’s Crew, and Mises, as an alternative to Darwin, Spencer, and the Marginalists in Economics. In fact, it appears that almost everything written by each of these authors is a fabrication of wishful thinking correspondent with reality. Right now we are in the process of overthrowing keynesianism because of its externalities. Hayek suggested that the twentieth century would be remembered as a new era of mysticism (which we call pseudoscience today). He was right.

But all that said, the MBTI rests on a subset of observed preferences in behavior. These preferences exist, and are demonstrated in the work place.

2) —“The Myers-Briggs provides inconsistent, inaccurate results”—

So does a Big5 of 30-100 questions.

A 20 question IQ test is however, pretty predictive. What does this mean? It is easier to measure intelligence, harder to measure neuroticism(big5), and harder yet to measure work behavior.

The results are inaccurate because (a) there are too few questions, (b) most people don’t fit into an exact block but around the edges of one (c) the ‘dimensions’ being tested are difficult to test – and most importantly to test ‘positively’ (meaning without asking the survey taker to be too self critical.)

The problem is that for a test of this nature to produce accurate results it must consist of something on the order of 600 questions, about one sixth of which detect lies, or uncertainties. MB is ‘good enough’ that over time one can take the simple test, evolve greater undrestanding of one’s self, and ‘narrow down’ one’s score.

On the other hand the Big 5 judges these properties:

a) Openness (inventive/curious vs. consistent/cautious)
b) Conscientiousness (efficient/organized vs. easy-going/careless)
c) Extraversion (outgoing/energetic vs. solitary/reserved)
d) Agreeableness (friendly/compassionate vs. analytical/detached)
e) Neuroticism (sensitive/nervous vs. secure/confident)

These are DIAGNOSTIC categories that DO correspond loosely to what we understand may be brain functions.

It should be fairly obvious to people that these spectrum can easily be mapped to the MBTI (See Attached table). And this table will tell you all that you need to know:

i) MBTI Does not test for neuroticism – which we can consider good or bad. I consider it good because there is no way to spin it ‘good’ in all cases. But I believe this is one reason for variation between the two procedures.

ii) There is very high correlation between:
Extroversion-Introversion /Extroversion (.7)
Sensing-INtuiting/Openness, (.7)

….and less but still significant correlation between

Thinking(criticizing)-Feeling(empathizing)/Agreeableness (.4)
Judging-Perceiving/Conscientiousness. (.5)

As I understand it, the difference between Big5 and MBTI models is that TF and JP are heavily influenced by Neuroticism(insecurity vs confidence), and this is not accounted for in the brevity of the MBTI test.

Ironically the MBTI axis of Judging(organizing) – Perceiving(Iterative) probably MORE predictive and useful than the Conscientiousness measure, since I am fairly sure the Big 5 model is incorrectly diagnosing what is an important part of our division of cognition. I always pair myself with and INTJ. Why? I will absolutely figure it out, no matter what it is. The INTJ will absolutely positively get it done, no matter what, and I won’t. This method of thinking is not directly visible in the Big5

So the truth is that GIVEN THE CORRELATIONS and given that we are testing for very subtle differences, it is EXTREMELY hard to claim that the MBTI fails without saying the Big5 also fails.

Except that the MBTI teaches you to understand how to work with people in a division of perception, cognition, knowledge and labor, and the big5 teaches you what is WRONG with people in some strange freudian utopia where there is an ideal type of person. And it is this fundamental totalitarian error of Freudianism that is buried in the Big5: the ideal type: one-ness. Universalism. Equality. Ideal. Whereas that was not the hierarchical division of labor that was central to the western tradition and central to Neitzsche’s work.

Realistically it is the difference between the consumer model that is good enough for everyday work, and the professional model that requires precise measurement in order to perform medical operations.

What I dislike about the Big5 is it’s hypothesis of a perfect (Feminist) individual. MBTI doesn’t do that. It just tells you how people are, and assumes you can tell the differnece between the secure and insecure becuaes they don’t wanna tell people using a consumer product that mostly they are insecure. When actually, using something like MBTI long enough will reduce a LOT of your insecurities.

iii) The Dichotomy Model proposed by Jung is false. We have at least five if not six or seven major axis of personality that affect our behavior – which I won’t get into right now. But what does that mean? We’ll find out in a minute…

BUT! This simplistic error of dichotomy helps us understand why personality testing is difficult, and why the simplified version of MBTI is ‘pretty good’.

Humans really are terrible comparing more than a two dimensional representation of anything. We evolved to compare one thing with another. But most of our intellectual advancement has been the product of learning how to compare increasingly complex things.

So if we can graph two functions on a plane we can visualize them. If we can take slow motion video of a horse running we can analyze what it’s really doing rather than guess – something which stumped artists for all of history until the era of photography.

Statistics is rife with aggregates that falsely inform us. Left and right are insufficient models for analysis of politics. two dimensions are insufficient to capture all but four simple axis. Three dimensions can create a better nolan chart. It takes three dimensions and some work to create a class diagram.

For those with rudimentary understanding of economics as a study of equilibria, supply demand charts are hard enough. but what about multiple supply demand charts? We have to create models at that point using software, because we cannot visualize the results.

For those who are involved in Austrian economics, look at the difference between Hayekian triangles: how he worked to create a model of intertemporal production cycles.

This is the problem when we talk about five or more dimensions of personality: we cannot represent them simply.

Each personality trait represents a spectrum – a line with different variables, at each end of which are points of failure. And modeling multiple dimensions how they appear as demonstrated behavior is pretty difficult.

So, lets imagine a bunch of tall tubes standing on end, arranged in a circle. We fill each with liquid measuring each of the 5+ personality traits. Now, even if marginal difference in behavior between the extremes is only say on a scale of ten on each one (and I think it’s more than that), that’s a lot of combinations of personality types available to us.

But we could however, instead of combinations state ratios (intersections), or basically a truth table (binaries). And this is what MBTI tries to do. Produce binaries where there might be many in between, just so that we get ‘close enough’ to start working with people.

The reason to do this is because the average human mind just cannot really manage to do more than that.

Now back to our ‘tubes’, lets take our circular stack of tubes and draw a horizontal plane through all of them in the middle. This is the way that Big5 looks at personality measurement.

But we can draw many planes at many angles, in order to treat some properties more or less importantly than the others. This is how MBTI looks at measurements: that each plane we draw, if we draw 16 of them, will produce an ideal type that we can use to understand others.

So in this sense, MBTI USES 16 IDEAL TYPES that you empathize with, AND BIG5 USES ONE IDEAL TYPE and a lot of properties that you have to rationalize.

Once you see this, and grasp that they are measuring 4 of the same properties, this makes sense.

MBTI is a mass market teaching tool. And it works.

As a ‘professional’ I use my own categories.

3) —“The Myers-Briggs uses false, limited binaries”—
This is a ‘feature’ not a bug. The reason MBTI is successful is that PEOPLE CAN USE IT, and you can take it over and over again and start to understand yourself and others.

4) —“The Myers-Briggs is largely disregarded by psychologists”—

So is IQ. So is Nature vs Nurture. And Freudian psychology was an non-empirical pseudoscience constructed by introspection and guesswork just like Jung’s – and arguably remains so outside of experimental psychology. It is cognitive science not psychology we follow today.

Unfortunately, I’ve used pretty much every model on the market, and while I DO use a more predictive model, which produces graphs of the four major personality traits, (blame avoidance being my favorite), MBTI fits the GOOD ENOUGH model for 90% of the world’s work force. And that’s why it’s good. ‘Cause 90% of the ordinary folk in the world can learn how to use it until something better comes along.

I prefer:
I) moral biases: feminine(left)/balanced(libertarian)/masculine(conservative),
II) altruistic-trusting/balanced/not-trusting-selfish,
III) extraversion/balanced/introversion,
IV) autistic-analytic/balanced/empathic-solipsistic,
V) rigid-organized(closing things off)/balanced/ intuitive(preserving options)-irresponsible,
VI) endurance-patience/balanced/frustration-impulsivity,
VII) paranoia-fearfulness/balanced/confidence-steadiness,
VIII) verbal IQ in .5 std deviations from 100. (scale of -5 to +5 because more or less is irrelevant.)

With those 8 measurements I am pretty sure we can lock down almost everything about a person.