Does the Myers-Briggs personality assessment really tell you anything? | Hacker News Reader

Does the Myers-Briggs personality assessment really tell you anything?

straightdope.com

25 pointsrglovejoy27 comments

It's a model, which brings to mind my favorite quote on models:

  All models are wrong; some models are useful.

I.e. the question is not 'is it right', but 'is it useful?'. Personally, I think the MB model is somewhat useful - e.g. identifying some of the 'watch-out' points of various personality types.

Does the same argument excuse astrological signs? You're a Libra? Libras need to carefully balance their desire to be accepted by their peers with their conviction that they are in fact in the right. (Technically speaking that's just total horsepuckey I plucked out of the air, but put a lab coat on me and I think I could get 80% of the population to agree that it applied to them. Score one for science!)

I've been sharing this with my pregnant friends - http://xkcd.com/699/

Unfortunately astrological signs didn't give me any 'opposites' I can compare myself with. Having read through the descriptions of various character types associated with various astrological signs, I cannot say I identify with one any more than another.

MB-types give me a set of sliding scales. I show up as strongly I and strongly N, which allows me to compare myself with ES types - I recognise very little of myself in ES types, and a lot more of myself in IN types, which means I'm biased towards the assumption that there is at least some basis to it.

I can also make predictions about what types my friends are, and (before telling them my guess) ask them what type they are. Currently these predictions seem to be matching up pretty well, although a little fuzzy on the P/J axis.

I don't know much astrology so I'm not able to address that issue, but decades ago I studied handwriting analysis. I realized there were "opposites", but they were always described in a positive manner. For example, long endings on a script 'e' were associated with generous people. The opposite, a short 'e' ending, would be associated with a thrifty individual -- as opposed to the more blunt cheapskate.

Maybe that's just wordsmithing but it seems reasonable to characterize attributes in a positive way.

Like so many similar things, from divination to Dungeons and Dragons, the Myers-Briggs is a game designed to provide a structure on which to hang conversations and creative thought. [1] It gives you some ground rules, some traditional starting points, some classic teams to root for (go ENTJ, go!) and an assortment of useful pivots and props that can be artfully used by the participants to steer their conversation as it progresses.

As such, it's quite useful, though I would no more try to defend it as a scientific theory than I would try to defend chess as an accurate representation of medieval battle tactics. [2]

---

[1] You know what would be fun, or at least funny? A Dungeons-and-Dragons version of the Meyers/Briggs. In fact, I think the AD&D alignment system may be more than halfway there already.

[2] Which is not to say that medieval tacticians, to say nothing of modern tacticians, didn't play a lot of games as part of their practice.

> You know what would be fun, or at least funny? A Dungeons-and-Dragons version of the Meyers/Briggs. In fact, I think the AD&D alignment system may be more than halfway there already.

A few months ago, we actually had a "personality types and how to get along with them" sort of training at work (not Meyers/Briggs, but something similar). I thought it seemed kind of bogus, so when it was time to fill out the feedback forms afterwards I suggested that a follow-up "alignments in the workplace" training would be at least as useful.

Describing Myers-Briggs as witchcraft is gratuitously unfair to witches, but like many other forms of pseudo-science which make someone with few demonstrable skills sound like they earned their graduate degree, it is surprisingly durable.

"... Does the Myers-Briggs personality assessment really tell you anything? ..."

Yes, MB is about "self validation" and can be dangerous when used to pigeon hole individuals into occupations by third parties who rely on the results ~ http://www.skepdic.com/myersb.html

Human brain is an unstoppable pattern-matching machine and it will invent patterns when there were none if that's what it takes to satisfy the urge. That's why astrology, personality types etc are so popular - it satisfies both narcissism and pattern-craving at the same time.

>75% CS majors test intp. Thatsa lot of stars aligning.

And don't forget about a poll that appeared here a while back.

http://news.ycombinator.com/item?id=946249

As I said there, "Apparently, 76% of Hacker News is an NT type of one type or another."

Yes I forgot the urge to conform to a group.

Yes. It is very likely that once you have gone through what appears to be a laborious procedure to generate a description, you will accept whatever description comes out of that procedure.

http://en.wikipedia.org/wiki/Forer_effect

http://www.skepdic.com/forer.html

At times you are extroverted, affable, sociable, while at other times you are introverted, wary, reserved. Some of your aspirations tend to be pretty unrealistic. Security is one of your major goals in life

Can apply to almost anyone reading it:)

Extremely good multi-part series on understanding personality temperaments/indicators/etc.:

Part 1: Temperament for Dummies http://www.erictb.info/temperament1.html

Part 2: MBTI and the 16 Types and Cognitive Functions http://www.erictb.info/temperament2.html

Part 3: APS and Other Systems (Enneagram, Horney, Type A, etc) http://www.erictb.info/temperament3.html

Part 4: APS and Dynamic Type: adding moderate scales to EISeNFelT (from 16 to 81 types) http://www.erictb.info/dynamictype.html

From what I gather each person has potential to display ALL personality types. The question is to what degree and in what contexts we prefer them.

I sure like having some kind of framework (or several) to help me analyze people. Learning to identify traits using a few simplifying tools (MBTI, color code, enneagram, love languages, etc.) helps me quickly figure out how best to get along with someone so we can communicate and get stuff done (in a work environment especially), understand better how someone prefers to interact with ideas (casual conversation and debate), understand people's underlying motivations a bit (particularly important with family and relationships), and in general just "get" people.

-A softened ENTJ

I found it to be surprisingly accurate for myself. I never took the official quiz but I usually come out as either an INTP or INTJ. It seems that last slot is a close call for me. After reading over both types, INTP is the most accurate. I think it's up to most people to find where they best "fit" and that will be more difficult for each "category" for lack of a better word if they aren't definitively one way or the other.

So knowing that, now what? Well, I've tried using it to understand the general process of how others think and better understand how I myself think. I'm prone to introspection and I do find that many of the attributes of the INTP match my own personality. It gives me a handle on what sorts of things I should watch out for and helps me to recognize why I tend to do better in some areas. I think like any tool it's success will depend largely on how it's used.

Well said. There was a poll here a while ago, something like 75% of the people that responded on HN had NT in commmon - it can obviously help identify technical people. 90% of CS majors at my college were intj.

From a sky high view, it's been very good to help see how other people think - especially emotionally vs logically and organized vs disorganized. And every other ENTP I've met had damn similar thinking and habits as myself.

it can obviously help identify technical people

I would rather have my son, an aspiring programmer, be identified by his coding skills.

EDIT TO REPLY TO BELOW:

My son got into programming by programming. In that sense, I am NOT technical (as my HN profile should make sufficiently clear).

I'm a lot more learned in psychology than in computer science, and that is precisely why I distrust the Myers-Briggs, because it has never been properly validated.

Coming from a technical family (as I assume you are technical) it's probably easier to get him in to some of those things. For younger people who perhaps don't have the same level of guidance from an adult it can be useful to identify areas or professions that they may not have considered or even been aware of.

I remember when I first started programming - I tried it out because it seemed interesting. I realized something was different when I was completing week long assignments in 20 minutes and watched everyone else really struggle.

Sometimes people need whatever help they can get in finding some direction. If they aren't fortunate to have a person or environment to guide them something like a website with information about the MBTI can be a rough substitute.

Related: "Caring for Your Introvert" http://www.theatlantic.com/doc/200303/rauch

I question the validity of any assessment that is presented to assessors (the paying customers) as a verified and reliable instrument, but then presented to the assessed as a "starting point", and who should be instructed "...that the respondent is the expert; the only person who can verify which type fits best."[1]

It may test something to some degree of accuracy, but when the makers of the test turn it into a game of picking which of 16 all-positive self-descriptions they like the most, I have to wonder what the point is beyond being a well-marketed horoscope.

[1] http://www.myersbriggs.org/myers-and-briggs-foundation/ethic...

i think one of the main confusions is that people think it is a statistically-constructed personality test like the ones that are designed and bred as such

To MBTI skeptics: I agree that the MBTI is a flawed instrument, but do you assert that personality testing is intrinsically flawed, that personalities don't have measurable patterns/traits/components? Why? Is it really that inconceivable that there are broad tendencies and trends in personality? Is everyone such a "unique snowflake" that there's no patterns or trends?

I personally think there's definitely some underlying dominant factors, but the fact is that determining what they are and measuring them is really hard, because it's really hard to do substantive psychology in a truly scientific manner. I think something like a principal components analysis should be possible in the theoretical ideal -- and binary trait based schemes are just a crude approximation to such a scheme. For example, scoring as "ENTP" wouldn't mean that you have zero introversion, or zero feelings -- just that you score closer to E than I, etc.

How can you go so easily from believing there is such a thing as personality traits to supporting a complex, copyrighted system that has no valid research basis and is actively sold to big business HR departments with the disclaimer that it is basically nothing more than a party game?

People with an interest in the science of personalities should dislike MBTI more than casual observers. It's a pop-culture perversion of the entire field.

Like I said, I think the MBTI is broken. I'm asking why people act as though the idea of personality testing is intrinsically flawed from the get-go and on the level with astrology and numerology.

I don't believe personality testing is intrinsically flawed so I can't speak for those that do but ...

I think the answer to that question could be that the first exposure to the field is MBTI which is blatantly nonsense and generally promoted by people who wouldn't know science if it bit them on the ass.

Reply on news.ycombinator.com