You can show that an animal was missclassified as an invertebrate by showing the flaws in the test that failed to identify the spinal cord that exists. If you disagree with your MBTI label, you can just take the test again and answer the questions differently, at which point your label changes. Or does it? Nobody knows, because MBTI is a game masquerading as science.

When zoologists label something 'frog' or 'mouse', they observe specific traits that differentiate the mice from frogs and group them into categories. The specific observable traits and the definition of the species are essentially the same thing. It's possible to discover an animal with traits that correspond neither to frog, or to mouse. It's possible (though obviously unlikely because our skills at observing and describing physical traits in animals are far superior to our skills at observing complex patterns in our own behavior) for a test to mis-classify a frog as a mouse, and that a subsequent test will show the flaws of the original test.

MBTI doesn't do that. MBTI tests assign unspecific predefined labels to test-takers, and that's why I say it is not science. There is no real discovery.

Disagreeing with your MBTI label doesn't mean you can take the test again and answer questions differently - if you can honestly answer a question in a different way, you fall on the spectrum between things.

What many people miss is that the labels are merely points on a spectrum, each of the four factors is given a percentage rating showing how strongly that factor applies (70% E, for example, implies 30% I - you're more extraverted than introverted, but still have that introversion in your personality).

I'm not saying MBTI is 'science', as there's no way to scientifically prove personality traits, but your specific argument against it is flawed.

Actually I must contradict that assertion. It is theoretically possible to observe cellular life that is not adequately described by those three labels. It is highly improbable because the descriptions are the result of thousands of man-years of study, and covers everything we've been able to discover so far. But because the definitions are strict and based on empirical observations, it can be done.

In fact, Achaea is a relatively new sub-category of Prokaryote, because the increasing number of empirical observations stretched the limits of the existing definitions. Scientists found life that failed the Eukaryote test, nominally passed the Prokaryote test, but were really sufficiently different from other Prokaryotes such that a new term was assigned. This is because real effort was made to make the definition of Prokaryote strict and precise based on empirical observations. That leaves open the possibility that unexpected observations (discoveries) will challenge the usefulness of existing definitions. That won't ever happen with MBTI. You can never take a Meyers-Briggs test and have the test fail to categorize. Every test gives you an answer, a personality type that reads like a horoscope. The types are not strict and precise like the definitions of cellular life, they are vague and inclusive like your horoscope. You don't read it and think "interesting new knowledge!" you read it and think "this is so true! This is me!"

Consider the human conditions that MBTI fails to identify. It cannot identify a psychopath. It cannot identify a schizofrenia. It cannot identify alzheimers disease, or any sort of mental illness even due to physical damage to the brain. It cannot, technically even determine whether the subject is human (although that could be a reasonable assumption). If this methodology can not and will not ever be able to distinguish between recognized psychiatric diseases, is it really reasonable to expect that it can differentiate subtle personality traits based on the results of a questionaire? Is it really telling you anything useful about yourself? You could have a serious, potentially dangerous mental condition like bipolar disorder and all the MBTI will tell you is that you "enjoy being with people."

Incidentally whether it's a spectrum or not doesn't change the argument at all. A person who scores 50%/50% introvert/extrovert is not a discovery. The test fully accounts for this.