There are various problems with Myers-Briggs typing, and with other forms of personality analysis, particularly when applied in a workplace setting:
Process Problems
Confidentiality & Privacy
In some group settings, personality testing is offered as an activity, in which results would be shared with the group. Immediately, there are serious concerns about confidentiality and respect for privacy. In some ways, a "personality test" is analogous to a medical test. Most of us would be uncomfortable volunteering to share our medical test results (such as a chest x-ray or urinalysis) with a group of co-workers, employers, or strangers.
Peer Pressure
While this type of activity is often respectfully introduced as voluntary, there is a strong peer pressure element. A person who would choose not to share or participate would immediately stand out as an outsider. It would not be irrational to wonder if a non-participant might be doing so because of some sort of personal difficulty, uncooperativeness, or "personality problem." Such an unhealthy dynamic is due to the activity itself, more than the dynamics of any individuals in the group.
Labeling
Also, in terms of process, the Myers-Briggs gently applies a sort of "diagnosis." The 4-letter code each person gets is actually a label. The label may be an interesting thing to reflect upon, but nevertheless it is a label, obtained from answering a small number of simple questions.
Some of the Myers-Briggs types, such as INFJ, typically occur in only 1-2% of test cohorts. If test results are shared as a group activity, with a group of 30-50 people, and with the group then dividing up according to type, then the INFJ group could be a group of just one person! This would literally "single people out," and give rise to an unnecessary experience of isolation or exclusion, rather than celebrating group togetherness or harmony.
The principle of giving people labels based on a few minutes of superficial assessment is troubling, in terms of process. It is exactly this sort of pattern that I encourage people to step away from as a practice, particularly with regard to mental health care policy. As a general principle, I like to be very reserved about diagnostic labels even after spending many hours (or even years) knowing someone; I think it is quite an egregious practice to pull out a label after a 10-minute questionnaire!
When people are given labels of any sort, particularly if such labels are sanctioned by some sort of official test, "clinical wisdom," or group approval, then it can push people into conforming to these labels. It is a form of typecasting. If you are told that you are something, you are more likely to believe that as part of your sense of self. If a teacher you trust tells you that are a talented mathematician, but not a very good writer, and shows you a questionnaire result which "proves" this, then you may be less likely to pursue an interest in writing! Of course, such feedback could be given in a way which is honest but not discouraging (e.g. "you could be a truly great writer -- but I think you need to put an extra hour of work into your writing every day this year!")
In the Myers-Briggs, the labels are generally benign, each of which laden with various positive affirmations, but the process is troubling, as it is an example of a sensitive, personal identity issue being pushed through forces of labeling and peer pressure.
Of course, the Myers-Briggs can be used simply as a tool of playful engagement in a group, almost like a "party game." It can very reasonably help a group acknowledge, respect, and admire the diversity of personal and interactional styles. And the labels can be taken with a grain of salt. None of the labels are framed in a pejorative way at all, they are all framed as style variations, which all carry different strengths or gifts that are to be respected in a group.
Yet, these process issues must be considered with great seriousness.
Validity
Personality traits of any sort occur in a continuum in the population. Generally, traits such as extroversion are normally distributed. Most people in any group will be close to the population mean on any trait. Yet the Myers-Briggs gives people a categorical label (for example "E" or "I") for these traits which are actually on a continuum. It is a literal example of "dichotomous thinking," which is considered a "cognitive distortion" in CBT theory. It would be like describing people as "tall" or "short" in a height category. Or like describing the climate of a place as "hot" or "cold." Furthermore, because of this dichotomous labeling, there is a huge statistical "fuzziness" caused by the majority of people having traits near the mean. People in the 48th percentile for extroversion would be considered an "I" while people in the 52nd percentile would be considered an "E." In reality these two people would have differences in this trait which are not statistically significant, yet they are given labels which are dichotomously opposite, in just the same way as a person with extreme introversion would be considered different from someone with extreme extroversion.
Reliability
Furthermore, because of this statistical "fuzziness," there will be limited test-retest reliability for the Myers-Briggs, particularly for those people whose scores are closest to the mean. For those people who were farther from the mean, they would not have needed to fill out a questionnaire to tell if they were introverts or extroverts! They could have just told you verbally, without any questionnaire at all!
Carl Jung
The Myers-Briggs system was based on some of Carl Jung's theories. But one of the wise themes in Carl Jung's thinking was about acknowledging contrary elements, or duality, in personality, both in the course of development through life, but even cross-sectionally in a given moment. This allows for being an "extrovert" but also having elements of "introversion" at the same time, or in different situations! Personality phenomena can be understood as much more situationally dependent than we might think. There may be elements of extroversion that can occur in particular contexts, in a particular culture, while the same person may show much more introversion in other contexts, or in another cultural milieu.
I think this basic wisdom is often missed with testing of this type. I encourage people to reflect upon all elements of their personality, and to consider how some kind of "opposing" trait is always present, and deserving of nurturance, alongside any "dominant" trait.
Money & Marketing
Test makers are earning money through administering these questionnaires, publishing books about them, leading seminars, etc. According to some estimates, it is a 500 million dollar per year industry, growing by about 10% per year. On their very "slick" website, one can enroll to take the MBTI for a fee of $49.95. The website is introduced by a compelling, poetic quotation by Carl Jung:
It is quite masterful salesmanship to convince people that doing a brief questionnaire, then getting a very questionable 4-letter code based on the results, is actually a form of "looking into your own heart, to allow you to awake from a dream..."Your visions will become clear only when you can look into your own heart. Who looks outside, dreams; who looks inside, awakes.Carl Jung
It is important to acknowledge the marketing agenda at hand when participating in this kind of "corporate" psychology. Just as in other areas of salesmanship and marketing, the claims made by those administering or grading the tests may be designed to please the audience sufficiently, so that people may continue buying the product. This is not necessarily unethical, it is simply commerce and free enterprise...but such economic and marketing dynamics should be left to the marketplace, and not imposed uncritically upon a work setting, particularly given the peer pressure dynamics mentioned above.
For a related discussion of this, see my other post "The Business of Psychological Questionnaires": http://garthkroeker.blogspot.ca/2015/11/the-business-of-psychological.html
Arguably, a free test could be created with similar motives, to be used in a group setting for the purposes of reflecting upon group dynamics and personality styles. Actually, no tests need be done at all! If diversity of personality styles, and respect for different approaches or attitudes is the theme of a group seminar, then this could simply be discussed together, rather than any form of testing be involved!
References
Here are some good references to look at, addressing some of these points:
http://www.indiana.edu/~jobtalk/Articles/develop/mbti.pdf
http://fortune.com/2013/05/15/have-we-all-been-duped-by-the-myers-briggs-test/
https://www.theguardian.com/science/brain-flapping/2013/mar/19/myers-briggs-test-unscientific
http://www.businessinsider.com/myers-briggs-personality-test-is-misleading-2014-6
http://signallake.com/innovation/personality.pdf