The Myers Briggs Type Indicator (MBTi) has become the whipping-boy of personality inventories, in large part due to its wide popularity. Writers from Psychology Today to the BBC are taking swings at the metric. Observe that these authors are relegated to the blogging and opinion section of these notable publications. None of them bring calculable, contradictory evidence to bear against the internal consistency reliability estimates: split-half internal consistency reliability of continuous scores for numerous samples exceeding .75 for each scale in research. In short, the proof behind the pudding of the MBTI.
Washington Post writer Lillian Cunningham took a journalistic approach, mingling history, personal use stories, and quotes from Adam Grant, a loud critic of the inventory. Even then, Cunningham doesn’t attempt to disprove the internal consistency reliability estimates nor the test-retest reliability estimates of the MBTi.
Adam Grant’s own critique (www.PsychologyToday.com) of the MBTI demonstrates the careless criticisms pitted against the tool. To begin with, Grant insists on calling the inventory a test and then demands that a “test is only reliable if it produces the same results from different sources.” The MBTi is very clear: it is NOT a test. There are not right or wrong answers. Call it an assessment, an inventory, an instrument, or a metric. But it is not a test, any more than the Minnesota Multiphasic Personality Inventory (MMPI) is a test. These human metrics inventory to varying degrees the presence and or absence of specific traits. Nobody is raging against the MMPI or compares it to “palm reading,”-a straw man accusation-as Grant does the MBTi.
NOT ALL METRICS ARE THE SAME, AND THAT’S A GOOD THING.
The differences between the MBTi and the MMPI don’t stop there: the MBTi doesn’t presume to dictate specific jobs, vocations, or even future performance; moreover, the MBTi must be taken willingly. By contrast, organizations-many of which are top secret government agencies-employing the MMPI can and do require it. Both continue to be updated: the MBTi went through its last major revision in the past 8 years, with the implementation of Form Q and revision to the Step II facets. By comparison, the MMPI went through several significant revisions in the past 15 years.
The MBTi has room to grow, and must to remain at all helpful. But no more so than the MMPI-2 which is based on reliability and validity that is heavily weighted toward professional people who are more educated, and the potential inconsistency between a respondent’s basic profile as compared to the original MMPI.
WHAT IS THE MBTI: TWO VIEWS
The MBTi is an assessment which provides insight to developed preferences in four areas: one’s rich world (where one gets energy); how one takes in information (the Perceiving function); how one comes to closure or makes decisions (the Judging function); and which mental function one uses in the external world.
Grant calls the MBTi “the fad that won’t die.” The beauty and resilience of the MBTi is that it assumes the respondent best knows and understands him- or herself. Grant disagrees, citing research by Oh et al. Yes, these researchers “concluded that the validity of personality measures in predicting job performance is so low that the use of ‘self-report’ personality tests in selection contexts should be reconsidered.” Key words: job performance. These researchers are concerned with how personality tests predict job performance. The MBTi is decidedly not.
The MMPI assumes there are sociopathic tendencies that one is probably not aware of. The RightPath assumes there is a hardwiring within each of us about which we are unaware. All other inventories that I know of presume respondents can only really understand themselves via external assessment. The MBTi actually provides dignity to respondents and leaves them in the position to self-determination.
Because of this feature, it is also easy to manipulate. Where the MMPI-2 has 567 questions and takes more than an hour to complete, the MBTi Form Q has 143 questions and can take as little as 20 minutes. Anybody who has taken the MBTi more than once can pretty easily manipulate the inventory to get the results desired. Grant criticizes the metric because he self-tested as an INTJ at one time and an ESFP another time. Common factors that cause such mixed results include the failure or unwillingness of the respondent to participate with the intention of the metric or a deep lack of self-awareness, failure or inability to understand the metric, lack of self-awareness, or some combination of the three.
INCONSISTENCY MAY BE INTERNAL, OR THE FAULT OF USER ERROR
Grant’s appeal to science does not indicate preferences toward INTJ. He does that elsewhere. He states he has a preference for Introversion. His categorical patternization of information conveyed via simile and analogy is decidedly an iNtuitive preference. His logical, objective analysis of the metric loaded with challenge is remarkably a Thinking approach. And the unwavering declarative intonation of his writing presents a settled closure typical of a Judging type. Grant’s ability to derive an ESFP result is not, contrary to many INTJ’s way of thinking, an aspect of his relational, organizational, or informational dexterity; so much as it is a result of the manipulability of the metric. To say someone is an INTJ because he turns to science is like saying someone dislikes the metric because they are an ISTJ. It’s a cheap jab.
What Grant critiques as inconsistency can also be understood as the cognitive ability of humans to adjust based on new information. If I describe a possible outcome for a situation, and you encounter that situation, you have now been conditioned from ignorance to knowledge to choose alternative outcomes. Thus, the MBTI is descriptive of why and how one has responded or acted in the past, not prescriptively deterministic of future behavior. The whole field of human psychology is dependent upon some level of psychological elasticity regarding behavior. Those who hate feeling like the MBTi has put them in a box have likely had the instrument misapplied in its administration.
Grant further demonstrates his misunderstanding of the metric when discussing the T-F dichotomy. He writes, “In the MBTI, thinking and feeling are opposite poles of a continuum. In reality, they’re independent: we have evidence…that if you like ideas and data, you can also like people and emotions.” He uses research about the self-concern v. other-orientation model to substantiate his perspective. But again, this is false. An extravert may be extremely self-concerned while deriving great energy from being other-oriented. The MBTi does not promote this “bipolar continuum,” nor does it encourage a codependent undifferentiated state of relational enmeshment. This dichotomy does, however, help people understand in what ways and places they are renewed with energy.
The T-F dichotomy is about decision making. The assumption is that everybody makes decisions based either on objective analysis (Thinking) or core values (Feeling). Feeling does not mean emotional. Thinking doesn’t mean emotionless. People with a preference for Feeling can make decisions based on objective analysis when core values aren’t at stake, and those with a preference for Thinking sometimes make decisions based on the people involved. What is natural and what is learned are not the same thing.
Grant compares the MBTi to a physical exam that “ignores your torso and one of your arms.” Since Grant seems to appreciate analogy, I would compare the MBTI to a blood test. Blood can tell you about your cholesterol, insulin, and levels of white blood cells present. Nobody expects blood samples to tell you whether your arm is broken. What is tested and discovered is not inaccurate; and the results are only detrimental if one attempts to misapply conclusions universally. Just because a blood test isn’t a comprehensive exam doesn’t mean it isn’t important or relevant.
MODELS AND REPLICAS AND HOW THEY’RE DIFFERENT
The MBTi is a model. Every model is an incomplete representation of reality. A model of reality on the scale of reality is no longer a model, but a replica. The MBTi doesn’t attempt to replicate reality, but represent aspects of it in a modular form. Yes, the MBTi is limited. It doesn’t measure risk tolerance, the potential to lead, or reactivity level under duress. That doesn’t make it inaccurate; it just confirms that there are limits to the application of the results.
Whether or not the Big Five is a better metric is beside the point. At heart, Grant doesn’t seek to compare and contrast, as much as he seeks to critique. Only later and elsewhere does Grant provide a more substantial–but still unconvincing in my opinion–analysis of his objections, published (Sept 22) at the HuffingtonPost.com and self-published at LinkedIn.
Still, in the Psychology Today piece, Grant doesn’t seriously address the strengths and weakness of the MBTi, of which there are both. Instead, he stoops to bombastic hyperbole, comparing the MBTi to palm reading and horoscopes.
I have not demonstrated the benefits of the MBTi as a useful tool requiring minor introduction, ease of use and understand, and applicable to real-life relational and organizational structures. I did much of that here. Suffice it to say, popular assaults on the MBTi are little more than the application of the Japanese maxim to personality inventories: “The nail that sticks up gets hammered down.”