Myers-Briggs Type Indicator fairly accurate or or fairly spuedo?

[The Tao's Revenge]

So in one of of my classes has a section about Myers-Briggs Type Indicator. I came out as an INFP healer type. Reading about it seems to produce astounding insights into successes, strengths, and troubles, and weaknesses I've had.

However I've also seen horoscopes that do that too. So you can see my distrust. Confirmation bias and junk.

I was wondering how accurate the test is considered to be from reputable socio-psycological sources?

[don't ask]

Nonsense - astrology equivalent

[susan]

It's a useful heuristic with high face validity and somewhat lower construct validity. However, it can be useful in relationship counseling and workplace problems. The 16PGF or Big Five Inventory are more robust as they are built on factor analysis rather than deriving from a non-empirical theory.

FWIW, I find that most people I've used it with do not find it horoscope-like, in that they do not identify with the majority of the types, generally only waffling between a maximum of 3 related types, if at all--e.g., between INFP and INFJ for a person who does not score far into the P end of the P-J axis.

[DrDeth]

It's not entirely bogus. Here's a quick and dirty test for bogusity:

Read off a few profiles of freinds or co-workers who are farily well aquainted. Generally, both bogus profiles and correct profiles will have everyone nodding their head and agreeing. Example- Bob is a Capricorn or an ISTJ, read the long version generalized profile of one.

After a couple, slip in one that's the total opposite- If a ISTJ, read a ENFP instead, pretending it's the right one. Or a Leo for Pisces (or whatever). For MB, dudes will not agree. For Astrology or other Bogus theories, dudes will still nod their heads and agree.

That's not to say MB is accurate. Often when tested, many people will have a score which is fairly close and can test either way. However, the profile of a ISTJ is fairly close to the one for a ISTP. Not to mention, there are more than 16 types of personalities in the world.

It's vague, and inaccurate, but it's not entirely bogus.

[DrDeth]

Originally Posted by DrDeth

For Astrology or other Bogus theories, dudes will still nod their heads and agree.

.

Note that this is called the Forer effect.

[Auntbeast]

I got pressured into taking this from a wacked out superior. Whatever it was I test out as, she told me my previous field (internet provider) was a job I was supremely unsuited for and that that sort of job would make me incredibly unfulfilled. She also indicated that I was profoundly suited for real estate and that I would pretty much live a life of rainbows, fluffy kittens and golden retriever puppies.

To this day, I consider my job as an internet provider the greatest, most appropriate, fulfilling job, most intellectually challenging job I ever had, I was in it for 7 years before the bottom fell out. I lasted about one year in real estate and hated pretty much ever minute of it.

Now, I don't know if it was the test itself, the answers I gave, or the woo-woo of the wacko that gave it to me, but for me, it was as far off base as could be. It didn't seem to be able to handle an outgoing, technically proficient, avid reader, extrovert very well. Like most stuff, the bigger the box, the more it can hold.

[meanoldman]

I am of the opinion that it is mostly useless. It merely tells me back what I already think of myself. (Who would have thought I am an introvert simply because I enjoy spending time by myself. Wow, what a reveal. Next, it'll tell me I think I am clever and inciteful)

It would be much better if say 4 to 6 co-workers filled it out anonymously about another person. Might actually reveal that person to be an arse or something.

[Nametag]

The MB isn't a job screening method. Using it as one will occasionally work, as a stopped clock is right twice a day.

The categories in the MB seem real enough; they may or may not be as basic as the authors think. The real question is whether the questions are capable of sorting people into those categories. I find them clumsy and obtuse, requiring people to understand themselves about as well before taking the test as one might expect to afterward.

[alterego]

Four of the MBTI dimensions correlate with four of the Big 5. The MBTI does not measure neuroticism. The MBTI has good test-retest reliability as well.

The creation of the test is a bit hand-wavy. It is intended to put you into bins - either you are introverted or extraverted, whereas the Big 5 put you along a continuum which is much more realistic.

The MBTI is good for what it's supposed to be used for - helping someone find a job that suits them, for example. Psychologists and social psychologists prefer the Big 5.

I am certified to administer the MBTI and I am also a scientist. I wouldn't use it in a study, but I do hand out official self-scoring versions to my friends. In my opinion the personality descriptions in the MBTI manual are quite remarkable and some of the best personality descriptions you can find anywhere.

[alterego]

The Wikipedia article for the MBTI bears my bark. At this point in its editing history it was written almost entirely by me:

http://en.wikipedia.org/w/index.php?...oldid=22775380

Since then it looks like its been substantially improved, though I don't have time to read it.

[Harriet the Spry]

I don't think it's fair to characterize it as "pseudo." It is extensively studied in the academic literature on psychology and management. Measuring anything about people is more complicated than sticking a pH strip in a vial of acid. That said, here is a citation of one journal article on the subject: Myers-Briggs Type Indicator Score Reliability Across: Studies a Meta-Analytic Reliability Generalization Study, Educational and Psychological Measurement 2002; 62; 590, Robert M. Capraro and Mary Margaret Capraro.

Here is the abstract:

The Myers-Briggs Type Indicator (MBTI)was submitted to a descriptive reliability generalization (RG) analysis to characterize the variability of measurement error in MBTI scores across administrations. In general, the MBTI and its scales yielded scores with strong internal consistency and test-retest reliability estimates, although variation was observed.

While the study in general supports the MBTI, I will include the paragraph that summarizes the criticisms, for balance:

"Although the MBTI has been reported by Murray (1990) to be “the most widely used personality instrument for nonpsychiatric populations” (p. 1187), there have been controversies regarding the indicator’s measurement characteristics. Pittenger (1993) observed “that there is insufficient evidence to support the tenants[*] and claims about the utility of the test” (p. 467). A contrary view to Pittenger was expressed by Hammer (1996). Other researchers (Comrey, 1983; McCrae & Costa, 1989) postulated that the MBTI did not adequately represent the Jungian theory on which it was presumably based. The forced-choice response format and false assumptions that all people can be divided into groups have also been criticized (Girelli & Stake, 1993; Vacha-Haase & Thompson, 1999). Another criticism concerns gender
weighting. Specifically, different weights are applied for men and women on the “Thinking-Feeling” scale based on socialization effects (Myers & McCaulley, 1985), leading to difficulty in comparing men andwomen on this scale (Vacha-Haase & Thompson, 1999)."

* I assume this error was in Pittenger, since Capraro and Capraro had it in quotes, but they didn't [sic] it.

I agree with susan that there are newer approaches that have better psychometric properties (e.g. Big 5). That doesn't necessarily mean those measures are currently well suited to replace MBTI for applied use.

I hope the OP will base his impression more on the input he gets that reflects some knowledge of academic psychology and less on the impressions of people whose sole knowledge of the field is having taken the test. This goes double for anyone who had the test used for any type of employment selection process, since the test explicitly should not be used for that purpose. That indicates to me that whoever administered their test probably misrepresented/ misinterpreted it.

People can be studied. It's not easy.

[alterego]

Originally Posted by Harriet the Spry

This goes double for anyone who had the test used for any type of employment selection process, since the test explicitly should not be used for that purpose.

I suppose you meant that it shouldn't be used as it is used in China/Japan - to select employees that have traits that you believe are desirable for your company.

However, it can and should be used for people who want to learn more about their personality so that they might pick a better field. Indeed, that's what it was made for originally.

[The Tao's Revenge]

Hmm well I can confirm I didn't find very good matches in some other types. The jest of what you're both saying saying to be it's accurate with in an error range, but it does have some accuracy.

The course book listed just 4 personality types that didn't use all 4 metrics. Those being Analyzer (iNT), Supporter (iNF), Director (SJ), and Creator (SP). I got curious because the book test said I was a strong creator but I'm not really that assertive and my behavior generally tends to the Supporter description. I can't seem to find any online documentation of those 4 types, and the metric choices seem random. Even just using two there should be 8.

Tried an online test and INFP seemed a lot closer match. I'm supposed to write an introspective paper about my learning style (visual, auditory, or kinestetic), brain type (left verses right), and temperament profile which I just now as I write this figured out is an add on to the MBTI. What the book calls analyzer, supporter, director, creator maps to rational, idealist, guardian, and artisan in the Keirsey Temperament Sorter. Atleast they use the same personality type metrics, but they give different descriptions for the personality type. In the book for example NF maps to supporter, but in the KTS NF maps to idealist. Oy vay.

Anyway I'm prolly boring ya'll. DrDeth, Susan Your explanations gave me some incite that'll help. Thanks

[The Tao's Revenge]

Originally Posted by alterego

The Wikipedia article for the MBTI bears my bark. At this point in its editing history it was written almost entirely by me:

http://en.wikipedia.org/w/index.php?...oldid=22775380

Since then it looks like its been substantially improved, though I don't have time to read it.

Wow so many posts all of a sudden while I was typing that.

I just want to say that article has done a pretty good job telling me stuff the book lacks and thank you very much for your contributions to it.

[mks57]

Years ago, my software engineering group took the test. The interesting part was how skewed we were from the general population. I was an INTP. My boss was an INTP, as were several others in the group.

[ToeJam]

Hey, another INFP! Welcome to the club.

Though yeah, I do fluctuate between INFP/INFJ. And sometimes I get a T.
But that's the thing, there's variance in the results, but as was said earlier, it tends to be only between a few select choices. I will always be an I, i know that much. And many of the others don't apply to me at all. So I tend to go between only a few of them, which is better than astrology (where i just like to pick one at random and assume that's mine for the day).

[The Tao's Revenge]

Originally Posted by Ro0sh

Hey, another INFP! Welcome to the club.

Yay the club now has 2. That's a 100% increase in one day.

Though yeah, I do fluctuate between INFP/INFJ. And sometimes I get a T.
But that's the thing, there's variance in the results, but as was said earlier, it tends to be only between a few select choices. I will always be an I, i know that much. And many of the others don't apply to me at all. So I tend to go between only a few of them, which is better than astrology (where i just like to pick one at random and assume that's mine for the day).

On the T days do you pitty da foo'? Well the book descriptions of P/J seemed to be a toss up for me too. We'll see.

[Half Man Half Wit]

Originally Posted by meanoldman

I am of the opinion that it is mostly useless. It merely tells me back what I already think of myself. (Who would have thought I am an introvert simply because I enjoy spending time by myself. Wow, what a reveal. Next, it'll tell me I think I am clever and inciteful)

That's the biggy, I think, plus some measure of wanting to conform once you're presented with a nice and clear-cut label to fit yourself into. That it does better than astrology shouldn't be any surprise even if there's nothing very meaningful behind it, since everybody does have some measure of self-knowledge, and that's what the test really assess, I think. I have of course not performed the experiment, but it wouldn't surprise me much that if you asked persons to directly grade themselves on the scales used by the test -- e.g. introversion vs. extraversion, thinking vs. feeling etc. -- there wouldn't be great statistical differences to having them actually take the test. It's however not my area of expertise, so I could be horribly wrong about that.

[Nava]

Originally Posted by meanoldman

I am of the opinion that it is mostly useless. It merely tells me back what I already think of myself.

Originally Posted by Nametag

The categories in the MB seem real enough; they may or may not be as basic as the authors think. The real question is whether the questions are capable of sorting people into those categories. I find them clumsy and obtuse, requiring people to understand themselves about as well before taking the test as one might expect to afterward.

I've never taken the MB as part of a selection or orientation process, just for kicks and giggles, but those two points are true of about every "personality test" I've had to take either in school or as part of selection processes. Some are so evident I can get any score I want: if only I knew what is it the HR-type person has decided is the perfect profile, I'd be able to get any job in the world (not any, but... those where the test is the deal breaker, absolutely).

[KCB615]

We were discussing the MB tests at work a week or two ago, as it seems a few of us get along very well, and other supervisory/management type folks seem to fit a different clique. I'm a strong INTJ (Kiersey's "Mastermind"), my Captain was an ENFJ ("Teacher"), one of my firefighters is an INFJ (Counselor), and the Chief is an INTP ("Architect"). We get along swimmingly, have mutual goals and ideals, and generally have a good, productive time at work.

One of my other firefighters is an ESFJ ("Guardian") and the Deputy is an ESTJ ("Supervisor"). The both of them greatly complicate my life, I'll leave it at that.

The MBTI may not be true science, but it's definately more than astrology is. It managed to lend some outside views of the dynamics of a group more so than any other instrument I've seen could.

[DrDeth]

Originally Posted by alterego

The Wikipedia article for the MBTI bears my bark. At this point in its editing history it was written almost entirely by me:

http://en.wikipedia.org/w/index.php?...oldid=22775380

Since then it looks like its been substantially improved, though I don't have time to read it.

I disagree with one line in that wiki:

It has also been argued that the terminology of the MBTI is so vague and complicated that it allows any kind of behavior to fit any personality type, resulting in the Forer effect, where an individual gives a high rating to a positive description that supposedly applies specifically to himself.

As noted, I had doubts, so my tester allowed me to test for the Forer effect, and I found that although two close personality types would get that sort of answer, complete opposites would not. Thus, it appears that within noted limitations, MB does not fail a "Forer test".

[alterego]

Originally Posted by DrDeth

I disagree with one line in that wiki:

It has also been argued that the terminology of the MBTI is so vague and complicated that it allows any kind of behavior to fit any personality type, resulting in the Forer effect, where an individual gives a high rating to a positive description that supposedly applies specifically to himself.

As noted, I had doubts, so my tester allowed me to test for the Forer effect, and I found that although two close personality types would get that sort of answer, complete opposites would not. Thus, it appears that within noted limitations, MB does not fail a "Forer test".

[IMHO] The descriptions in the MBTI manual are so good that you don't really need to take the test - you could just read them all and pick the one that describes you best. This is in stark contrast to the descriptions found online which are much lower accuracy and quality.

[Chefguy]

This test is administered in the Foreign Service orientation class in order to allow the HR people to justify their salaries. An interesting demonstration of the "validity" was done after the test was scored and everyone received their letters. The administrator set up a flip chart at the front of the class and had all of the "I"s come up to the front and write down the characteristics of introverts. Then she had the "E"s do the same.

The writing exercise itself was bogus, of course. The purpose of doing it was to show the physical behaviors of the people in each group. If you've never seen 15 people try to hide behind a flip chart, you've missed quite a sight. The "I" group was mostly true to the idea of an introvert, standing behind others so as not to be noticed, standing behind the chart, looking everywhere but at the rest of the class, lots of nervous behaviors. The "E" group was just the opposite.

[jjimm]

I work for a business psychology company that uses the MBTI (amongst other psychometric instruments). Prior to joining, I thought the MBTI was total hooey, being based on a third party's liberal interpretation of attributes that Jung pulled out of his ass. Which it is. With character and personality descriptions that could fit anyone, and as much validity as a horoscope.

However. I took the test and retook a different one, with the same outcome. I did the Step I analysis and the Step II, which also gave the same result. Then I participated in a team breakdown with my colleagues, where a psychologist who had never met us before, looked at our types, then stood up and explained specific things about potential pitfalls of interaction between group members with different types - before any of us had even contributed our opinion.

Things like one coworker is likely to get intimidated when required to think creatively without a decent preparation time (completely true and she almost cried with relief when this was exposed). Things like I (an extreme ENTP outlier) am ferociously creative, able to be extraordinarily productive on a tight deadline, and am immensely adaptable - but my boss is likely to be very disappointed if she ever asks me to come up with a detailed schedule, or to work to a predefined plan (also completely true).

And she told us much more about each other, all of which was absolutely spot-on. The entire process was really useful to help us understand each other: since that third-party revelation of our working preferences, we've all got along much better and a great deal of tension in our team has been dissipated.

So I conclude that it's way more impressive than horoscopes, and much more perceptive than "cold reading" tricks could possibly achieve. It might only be partially correct, and it might be correct for the wrong reasons, but there's definitely something in it.

I am also convinced that the Forer effect doesn't apply here; furthermore, the tests are so designed that it is extremely difficult to fool them, even by people who know them inside out: they contain "checksum" loading throughout.

Examples of the tests being used unethically (prescribing careers and pedefining personality types for jobs) are merely companies' misuse of the products, which are not designed for such applications.

[alterego]

Originally Posted by Chefguy

This test is administered in the Foreign Service orientation class in order to allow the HR people to justify their salaries.

Without knowing all the details I can't make a definitive statement. However, if this statement is true - that salaries were determined on the basis of personality - it would be an unethical use of the MBTI and CPP would revoke the practitioners license. I don't know if it's also illegal.

[jjimm]

Another anecote: our company has the 4 x 4 MBTI character type posters (pdf) on the wall, per department, with people's names written in the type box that they have received from the tests. The marketing department's types are all clustered in the bottom-right, the accounts department in the top-left, and the IT department in the top-right. It's really striking to see.

[WarmNPrickly]

My wife had to take this thing. It seems to be borderline between telling you what you already know and the free stress tests offered at the mall. It probably helps justify paying the HR department too.

[alterego]

Originally Posted by WarmNPrickly

My wife had to take this thing. It seems to be borderline between telling you what you already know and the free stress tests offered at the mall. It probably helps justify paying the HR department too.

That's a wonderful response for GQ.

[The Tao's Revenge]

Originally Posted by WarmNPrickly

My wife had to take this thing. It seems to be borderline between telling you what you already know and the free stress tests offered at the mall. It probably helps justify paying the HR department too.

Well honestly actually it did teach me something new. Generally if I make someone unhappy I'll think it's cause I did something wrong. Sometimes people are just jerks though, but I won't see that right away and think "oh crap". Reading the description where it pointed that made me consider how much I do that and aware I need to get better at not taking it so personally. Continue teaching myself to do a mentl check when that happens to see if I really screwed up or they're just being asshats.


Also Chefguy. As a hardcore introvert; yea standing in front of a bunch of people looking at you... Oh my goodness D:, but even introverts can learn to use extrovert social skills, well maybe not as efficiently, but they can. I've observed and modeled extroverts and can fake it pretty good.

[Madgolf]

If you ever meet a MBTI-tested Inspector type, you will giggle with delight at the genius of that one particular category of the 16. (Unless you *are* an Inspector, in which case giggling is quite rare except in perhaps instances of mating, etc)

The other 15 types flit in and out of their pigeonholes as they go through the normal stations of life. Before I had kids I was a solid INTP and after the kids I was a solid INTJ. Now that I am burnt-out and jaded, I may very well be an INFJ.

Inspector types remain Inspector types and usually solidify further as time goes on.

MBTI is accurate at the moment only.

[Reply]

Another INFP here; I've taken the test multiple times in my life from multiple sources, including at least one "professional" one, and always got the same result. It's useful as a way to stereotype people into one of 16 categories based on what you know about them or what they think about themselves... that's gotta be a heckuvalot more useful than stereotyping them based on mostly random birth dates. "Hi, I'm introverted and tend to be emotional, and while I tend to trust my gut feeling most of the time, I also like to take a wait-and-see approach when it comes to people or events" says a hell lot more than "Hi, I was born in mid-June".

No personality test completely encapsulates someone's individuality -- hell, even 20 years of marriage won't necessarily reveal that -- but it's certainly a step up from shallow smalltalk and feigned socially-appropriate behaviors.

If employers or others want to use tests like that as qualifications, well, that's their choice. The good part is it lets them weed out potential undesirables much more easily, but they lose out on the odd introvert who might have an uncanny knack at gauging customer needs, etc. Best Buy, if I remember correctly, used a similar test that, frankly, only rewarded cheaters. They ended up with ruthless people who'd distort the truth or outright lie to get what they want, but maybe that's what they wanted in salesmen anyway.

On the other hand, as a way of getting to know people "at a glance", it can be very helpful. For example, if I'm not mistaken, the eHarmony dating service is based on very similar principles -- match people according to their self-identified personalities and the personalities they claim to be attracted to.

That's my completely non-scientific but very interested opinion. The GQ-esque answers have already been given, anyway

[Chefguy]

Originally Posted by alterego

Without knowing all the details I can't make a definitive statement. However, if this statement is true - that salaries were determined on the basis of personality - it would be an unethical use of the MBTI and CPP would revoke the practitioners license. I don't know if it's also illegal.

Ouch. I don't think anybody has ever so completely misread something I've written. It was a joke: the HR people have to do something to justify their own salaries.

[gallows fodder]

Originally Posted by Nametag

The MB isn't a job screening method. Using it as one will occasionally work, as a stopped clock is right twice a day.

The categories in the MB seem real enough; they may or may not be as basic as the authors think. The real question is whether the questions are capable of sorting people into those categories. I find them clumsy and obtuse, requiring people to understand themselves about as well before taking the test as one might expect to afterward.

I took a psychometrics class in college and we discussed the MBTI (though this was way back in 1999, so my memory of our discussion isn't the greatest) . One of the problems with it is precisely the way it sorts people into the categories -- it goes by simple majority. That is, if there are 20 questions relating to Introversion/Extroversion, and you answer 11 questions in favor of Extroversion, you are labelled an E. This implies that there is a clear cut-off between E and I, and that there's a significant difference (I don't mean that in terms of statistics) between someone who scored a 10 on that scale and gets labeled as an I and someone who scores an 11 and gets labeled as an E. Because the labels are so insensitive, it doesn't acknowledge that the difference between someone who scored a 10 and someone who scored and 11 on that scale is going to be much smaller than someone who scored a 1 and someone who scored a 20.

I had to take this test at work in the early 2000s and I was labelled an INXP because I fell right in the middle of the T-F scale. They put me with the Fs (maybe because I'm a woman?), and I found that I related with the Ts more often than with the Fs because we were looking at ourselves in the context of the work environment. Well, at work, where I was a sales assistant who worked with numbers and statistics all day, my emotional responses were irrelevant and so I became more of a T. At the same time, I wasn't as much of a T as the people sitting at the T table, who were Ts at work and at home.

Also, IIRC, many of the questions presented false dichotomies. I remember feeling frustrated during test taking because both answers to a question would apply equally to me.

[Elret]

I have found it incredibly useful in relationships. Over the years I (a very solid N) have frequently had difficulty communicating with most S types. being able to identify that this is where the communication breakdowns occur has helped me immensely in establishing strategies to deal with that.
It also helped me a lot growing up. As an F in a family of Ts I was often misunderstood. After my family became familiar with MBTI, they realized that I wasn't broken, as they had sort of thought I was, and we learned better ways to work together.
As a career selector, I don't know, but for relationships I really believe it has its strengths.

[Chronos]

Quoth Chefguy:

The writing exercise itself was bogus, of course. The purpose of doing it was to show the physical behaviors of the people in each group. If you've never seen 15 people try to hide behind a flip chart, you've missed quite a sight. The "I" group was mostly true to the idea of an introvert, standing behind others so as not to be noticed, standing behind the chart, looking everywhere but at the rest of the class, lots of nervous behaviors. The "E" group was just the opposite.

Interesting... I'm very introverted in one-on-one interactions, and every time I've taken the MBTI, it's shown me well on the I end of the scale, but I'm quite comfortable and competent at speaking to a group. So I guess I would have ended up being the guy all the other Is were hiding behind.

Quoth Reply:

On the other hand, as a way of getting to know people "at a glance", it can be very helpful. For example, if I'm not mistaken, the eHarmony dating service is based on very similar principles -- match people according to their self-identified personalities and the personalities they claim to be attracted to.

So they claim, but when I took the eHarmony personality evaluation, the results were both more vague and less accurate than the typical horoscope. The only things they got right were that I don't drink or smoke.

[Reply]

Originally Posted by Chronos

Quoth Reply:So they claim, but when I took the eHarmony personality evaluation, the results were both more vague and less accurate than the typical horoscope. The only things they got right were that I don't drink or smoke.

eHarmony wasn't great for me either. I had slightly better luck with okcupid once I answered about 500 questions. After that they just started getting trivial/stupid. Neither one, unfortunately, was as spot-on as the MBTI. You figure they'd just use a readily-established, time-tested system, but nooo.... they just had to invent their own. Makes me wonder... are there dating services based on the MBTI system?

[jjimm]

gallows fodder, two explanations to your valid complaints.

First, the MBTI questionnaire you're talking about is the Step I, which is something of a blunt instrument and bears all the issues you mention. However, the Step II is much more nuanced, and gets round most of your very valid criticisms. (Whether the Step II is totally valid, or is instead a hand-waving attempt to explain away the very discrepancies you mention in the overly simplistic model is open to debate.)

Secondly, the false dichotomies are indeed that, by design: firstly, in additon to the answer you give, they measure which of the false dichotomies (e.g. "I am sensitive / I am intelligent") you deem more important to you; secondly, the questionnaire actually asks questions on the same subjects multiple times in different ways. This validates the results internally, and is to avoid test manipulation by people 'in the know'.

[alterego]

Originally Posted by Chefguy

Ouch. I don't think anybody has ever so completely misread something I've written. It was a joke: the HR people have to do something to justify their own salaries.

Okay I admit it went over my head. However, there are a lot of companies in foreign countries that only hire employees with a particular MBTI type.

[alterego]

Originally Posted by gallows fodder

I took a psychometrics class in college and we discussed the MBTI (though this was way back in 1999, so my memory of our discussion isn't the greatest) . One of the problems with it is precisely the way it sorts people into the categories -- it goes by simple majority. That is, if there are 20 questions relating to Introversion/Extroversion, and you answer 11 questions in favor of Extroversion, you are labelled an E. This implies that there is a clear cut-off between E and I, and that there's a significant difference (I don't mean that in terms of statistics) between someone who scored a 10 on that scale and gets labeled as an I and someone who scores an 11 and gets labeled as an E. Because the labels are so insensitive, it doesn't acknowledge that the difference between someone who scored a 10 and someone who scored and 11 on that scale is going to be much smaller than someone who scored a 1 and someone who scored a 20.

This isn't true. The questions are designed and tested to ONLY provide information about that cutoff line. Questions that are shown to provide information about other regions of the scale are thrown out. Although the report puts you on what appears to be a scale it is completely invalid to interpret that scale as meaning more or less extroverted. All of the statistical information is compressed into determining which side of the cutoff the person is on. Thus, there is a truly significant difference beween that one point for I/E. The test is designed around that point. This is the big difference between the MBTI and the Big 5.

[Chefguy]

Originally Posted by alterego

Okay I admit it went over my head. However, there are a lot of companies in foreign countries that only hire employees with a particular MBTI type.

Wow. That does seem unethical.

[jjimm]

Originally Posted by alterego

This isn't true. The questions are designed and tested to ONLY provide information about that cutoff line. ... This is the big difference between the MBTI and the Big 5.

...if you ignore the MBTI Step II analysis.

[alterego]

Originally Posted by jjimm

...if you ignore the MBTI Step II analysis.

If you want to explain Step II I'm not getting in your way, but his answer is equally false for Step II.

[Lamia]

Originally Posted by Chefguy

This test is administered in the Foreign Service orientation class in order to allow the HR people to justify their salaries. An interesting demonstration of the "validity" was done after the test was scored and everyone received their letters. The administrator set up a flip chart at the front of the class and had all of the "I"s come up to the front and write down the characteristics of introverts. Then she had the "E"s do the same.

The writing exercise itself was bogus, of course. The purpose of doing it was to show the physical behaviors of the people in each group. If you've never seen 15 people try to hide behind a flip chart, you've missed quite a sight. The "I" group was mostly true to the idea of an introvert, standing behind others so as not to be noticed, standing behind the chart, looking everywhere but at the rest of the class, lots of nervous behaviors. The "E" group was just the opposite.

In one of my classes in grad school we all took some short version of the MBTI. The purpose of this exercise was mostly to learn about something we might encounter when out in the professional world, not to evaluate our own personalities, but our instructor did ask us how many people came out "E". Out of 40-odd people in the class, there were only 4 "E"s.

Did I mention that this was a graduate program in library science?

Apparently our "I"/"E" split was pretty typical of library students. When I mentioned this to other people in the program who were not in that same class, they always asked "Who were the four extroverts?" No one was surprised when I named them. It was pretty clear that they really were much more extroverted than the rest of us. Very few of my classmates were the hiding behind a chart type, and most did not seem introverted or withdrawn at all when around people they knew, but the four "E"s definitely stood out from the rest of us. One of the four was even nicknamed "the social director" because she was usually the one who planned parties and nights out for our social circle.

[ianzin]

A given score on a Myers-Briggs test tells you a score on a Myers-Briggs test. That is all. No more, no less. To claim that this score corresponds to anything at all in the real world is a matter of faith. If you want to believe it, it's true. Otherwise, it isn't... or... putting it another way, there's no good reason or good evidence to back up the supposed correspondence.

A lot of people want 'psychometric' tests such as MB to yield helpful information about people (such as misguided fools running HR departments who want some help recruiting the 'right' people). And lots of other people are quite prepared to cater to this market and to charge money either devising tests or administering them ("hey, if there's money being offered for any old rope that sounds right, we may as well grab some of it"). And so the money sloshes around the system, with both buyers and suppliers untroubled by the fact that the 'science' involved is exactly on a par with astrology and forms just as good a basis (i.e. none whatsoever) for understanding personality types or making recruitment decisions (which, as some have already pointed out, MB wasn't designed to be used for anyway).

One friend of mine used to run one of the largest organisations supplying and administering psychometric tests here in the UK. I got to know him some years after he had pulled out of that particular market. He cheerfully admitted that he had no idea whether psychometric testing was an accurate guide to anything at all.

Another friend of mine currently works for one of the largest and most reputable companies that deals in psychometric testing related to the workplace, and even lectures on how to use and administer such tests. I once invited him, in a perfectly friendly way, to send me any evidence he could find, in the academic literature, to verify that the use of such testing produced results any better than chance (e.g. better staff retention levels measured over a significant period of time, or better performance, or productivity, or job 'fit' and satisfaction, measured by some pertinent criterion). He was intrigued by the challenge and said he would indeed look for such evidence. He even had the funding and resources to allocate some of the searching to a junior colleague. One year later, I am still waiting. None found so far.

Interested readers may like to look at a book called 'How to succeed at psychometric tests' by David Cohen. Cohen is a qualified psychologist and this book points out that, in essence, a score on a psychometric test only tells you how good you are at passing that psychometric test. The fact that employers oftne regard such a score as 'meaningful' in terms or recruitment or promotion is just absurd, but such is life.

Some of the readers of these boards know that I have some expertise in the field of what is known as 'cold reading'. I have yet to see any so-called 'results' derived from MB or any other psychometric testing that can, in terms of accuracy, relevance or utility, be successfully differentiated by any objective test from complete nonsense I make up according to basic cold reading principles.

Just for the record, there's more to cold reading than the Forer effect, but the Forer effect is certainly one part of the mix.

The accurate answer is that psychometric testing (MB or any other form) is not a measure of anything and certainly not a science... it's a market. Some people want 'personality analysis magic' and others are prepared to put on white coats and pretend they can provide it... for a price.

[DrDeth]

Originally Posted by ianzin

Interested readers may like to look at a book called 'How to succeed at psychometric tests' by David Cohen. Cohen is a qualified psychologist and this book points out that, in essence, a score on a psychometric test only tells you how good you are at passing that psychometric test. .

There's no "passing" of a MBTI test, no type is "better" than any others.

[WarmNPrickly]

Originally Posted by DrDeth

There's no "passing" of a MBTI test, no type is "better" than any others.

If employers aren't using these tests to make decisions about what to do with their employees, then why are they making their employees take them?

[Musicat]

What's a "spuedo"? A speedo in another language?

Anyway...

I would like to recommend a recent book on this subject, The Cult of Personality, by Annie Murphy Paul, Simon & Schuster, 2004. It is subtitled "How personality tests are leading us to miseduate our children, mismanage our companies, and misunderstand ourselves."

She covers the range of mainstream concepts like MMP to pop concepts like MBTI, with history and application info for all. It's a good read.

[DrDeth]

Originally Posted by WarmNPrickly

If employers aren't using these tests to make decisions about what to do with their employees, then why are they making their employees take them?

Well, it was a teambuilding exercise when I took them.

[ralph124c]

i'd say MBI is pseudoscience. no reliable testing to prove its classifications.
Has anyone ever challenged its use in court?

[Chefguy]

Originally Posted by Musicat

What's a "spuedo"? A speedo in another language?

It's Spanish rap for "Yes, I can."