The inherent unreliability of the Myers–Briggs Type Indicator. Its test-retest reliability is less than optimal. This is due in part to the false dichotomization of personality traits (a few of which have somewhat deficient construct validity) that would probably be roughly normally distributed if estimated as latent trait dimensions. Legitimately discrete personality types (i.e., not continous trait dimensions) are rare at most.

https://en.wikipedia.org/wiki/Myers-Briggs_Type_Indicator#Reliability

Of course it is possible. For at least three reasons. First, personality does change. Although major traits are relatively stable over time, that doesn't mean they are fixed and immutable, especially if we have new life experiences that challenge our former ways. Second, there may not be any such "thing" as a personality type. Rather, we are most likely looking at a number of dimensions along which people vary. The MBTI forces dichotomies on some of these continua. Third, the MBTI isn't an especially good personality test.

I do believe , personality will change at certain situation ! it also depend on your age .

Please refer to this new finding.

In general, personality is linked to situational factors, social cues, and time. For example, the personality my students show and report in classroom exercises is no doubt quite different from the personality they show and report on a Saturday night. William James classically identified this "self" differentiation by theorizing that people possess multiple identities based on social relations. So, it's not surprising to me that a person's personality can show change over time.

As Nick noted - the Myers-Briggs is not often used in scientific psychology for reliability concerns, but I've seen similar personality shifts emerge when using more reliable tests (e.g. the Five Factor Model / "Big 5"). One reason for the shift is that a person completes these tests within some context. The events of a person's day, his or her mood and emotional state, and what concerns weigh on a person's mind will all influence these tests. I say this in part out of anecdote - I find that if I take a personality test when tired, I tend to get far more introverted results than if I take a personality inventory when well-rested and energetic.

However, there's more to personality change than situational effects. For example, Allemand, Zimprich, and Martin (2008; "Long-term correlated change in personality traits in old age" - Journal = Psych. and Aging) identified several personality factors that show continued evolution when people reach more advanced age. Specifically, older adults tend to become more agreeable while having diminished neuroticism (emotional reactivity). This is just one of many studies illustrating that personality is more pliable than one might first think.

Finally, psychology theories and research informed by complexity science and dynamic systems frame people as being psychologically fluid. For example, people tend to seek coherence in life. That is, they tend to seek meaning in their goals, careers, family, and social relations and prefer that these many things all integrate in a coherent way. When faced with something that does not fit their world view, people tend to seek new meaning - and this involves major cognitive shifts. These cognitive shifts can lead to decoherence from which people build new understanding (see for example my article along with Steve Parkin and Robin Vallacher - "Destiny is in the details").

Similar processes likely carry over to personality - where people cultivate new understanding of who they are as they experience life change. See, for example, Vallacher, R. R., Nowak, A., Froelich, M., & Rockoff, M. (2002). The dynamics of self-evaluation. Journal = Personality and Social Psych Review, 6, 370-379.

If you consider personality nothing more than a system of behavioral habits, then with certain therapies (e.g., mindfulness, DBT, etc...), yes, those systems of habits can change. If you view personality as core values, morals, and more ingrained characteristics, then yes, but it is much more difficult to change. 

Dear Dr Jay, I do agreed with you , but due to the short fall of 5FM model , low of correlation in most of the finding , it become a 5FM Personality Paradox ! But , I do believe the external situation have the highest impact toward this type so call ed in-consistency . One of the possible solution was , he have to make use of some other cognitive model to fine tune our finding . Like Emotional -Cognitive Processing model and Mischel CAPS model . This is what I am currently doing....But there are too many situation arise during our day to day life , it is very challenging to take care of most of the situation studies. Then, a lot of resources is required ! 

Hi,

I've researched and taught the MBTI for 25 years, mostly from the perspective of the history of ideas and the social studies of science. I've also investigated it as a questionnaire, including its historical presumptions and development

None of the people commenting thus far have asked what instrument was completed online. If it was free, the best guess is the Humanmetrics questionnaire, which hasn't validity in the sense that it doesn't operate according to the presumptions of the underlying theory, C.G. Jung's typology.

Regarding the MBTI, the previous commentators seem unaware that the MBTI does not measure traits but indicates to a theory. Accordingly it's reliability is based on other criteria.

If the MBTI is the focus of the discussion, it has never claimed to definitively identified anyone's type. It is an indicator and depends on the proper setting for completion and an informed feedback session.

The MBTI also has significant reliability and validity, depending on what information you wish to seek out. My experience is that most commentators ion this aspect, do not trouble to do so.

Regards

Peter

I think Peter has a pretty good take on MBTI. From the inside, as I recall, those involved in the (lucrative) business of training people to administer the process by which someone is supposed to find the "best fit" description of their personality type are careful to remind you that the questionnaire is only one indicator, and that the decision on what and which of the 16 types you think you are is yours. The questionnaire provides some sort of statistical indication based on how you responded, but - frankly - if you think you fit the "S" and not the "N", who is anyone to disagree?

It's another question as to whether this rather limiting way of punctuating the rather more illusive concept of personality is worth worrying about. Probably better to keep an open mind and an open eye to living life to the full.

Thanks Chris,

A lot of this does go down to what goes on in a training room and the pressures of what has to happen in the day.

Giving feedback on MBTI results, as with most things, requires the right questions to be asked and observations made, as well as how it's delivered. You need to set things up so people have an idea of what it's about, for instance, and be aware that what you say there may not be listened to. Having the results beforehand and thinking through them was something I thought was important in that, and not everyone does that. I think you should also know what the items do, but that's not something that attracts people.

In my work, I found the MBTI to be about 85% accurate, which is higher than what most get from it. So that tells me you just don't hand it out and see what happens, unless you know the theory behind it and how it differs from trait methods and so on.

I do have to say that I've never experienced using the MBTI as a lucrative occupation, the opposite actually, and most people i know who use it (members of psychological type associations and people I personally qualified when I was able to do that) would be in a similar position.  But then I was someone who only used type and mostly taught others about it.

I wonder in that sense whether its use by people who use multiple methods in their consulting has an effect on the quality of its use, because it's more difficult to know about it in depth and the purpose of using it, in leadership, say, may be peripheral to what's being aimed at. I do know people who say they use it just to make a point about different people, and they're not really interested on whether the person has arrived at a sufficient understanding to accept a type code label.

Cheers

Peter

Thanks for the thoughtful response, Peter. The "lucrative" comment was a jibe at OPP, the company licensed to run training for MBTI in the UK - they operate a near monopoly. I didn't mean that trainers or educators who dispense the instrument with clients are making money off it.

How do you get to a figure of 85% accuracy? Do you mean that most people end up with an enduring sense that the type indicated is true for them? Or a more objective assessment?

I must admit that I enjoy talking about the underlying concepts but find the labels restrictive. Once diagnosed, everything in life then gets divided up, rather like a paint-by-numbers picture before it gets filled in.