The Big Five and HEXACO models have proven pretty statistically reliable. But yeah, most any predecessors to those were complete ass pulls, often by people not really versed in the field (like Myers and Briggs).

What does "statistically reliable" mean in context of a multiple choice test of preferences and predispositions?

Replicable, in this context.

It means you can ask people on a questionnaire "are you open to new experiences" and then later offer them a new experience and get some papers off this incredibly shocking finding.

Someday psychology might be a science but it isn't now; it just rejected its alchemical mystics like it saw the cool sciences doing, but without any paradigm to follow them.

You'll be shocked when you realise that neuroimaging is entirely based on the results of cognitive measurements. i.e.: correlations between whats in the image and the kinds of tests you're traducing.

Fallacy. Just because your preferred belief framework has no supporting evidence or research doesn’t mean an actual science that you don’t like isn’t actual science. In other words, you can create your own alternative facts but they still aren’t facts.

I am not sure what "preferred belief framework" you are projecting onto me, but you may be interested to note that I mentioned no such thing.

The findings of psychology tend to be unreplicable just-so stories, like all the findings so popular in pop-psychology books that failed to replicate in the 2010s, or journal-filling tautologies, like "Big 5" papers showing introversion correlates with introversion or what have you.

Alchemy reformed into chemistry. You can make real sciences out of mysticism. Psychology is not a real science reformed out of mysticism, it's just academic bureaucracy set up like the ones associated with real sciences, producing useless papers.

Please don't enact your personal psychodrama on random people on the Internet, or at least don't do it to me.

> Please don't enact your personal psychodrama on random people on the Internet

I wouldn't dream of competing with you.

Myers Briggs has a significant overlap with the Big 5. They’re mostly the same if you think about it from a Venn Diagram perspective

They're somewhat correlated, but not as well as you'd like:

https://en.wikipedia.org/wiki/Myers%E2%80%93Briggs_Type_Indi...

> But yeah, most any predecessors to those were complete ass pulls, often by people not really versed in the field (like Myers and Briggs).

Which is quite a shame because the underlying groundwork of Jung is quite interesting.

Statistically reliable, in what sense? That they frequently produce the same outcome on repeated application of the test? That's not a surprise. As a wise man said: the gold standard for repeatable outcome is astrology.

Suppose I make a test and ask questions like "Do you like coffee?" and "Is your favorite color blue?". That would get a very high test-retest reliability. If I then score e.g. +coffee -blue as "intelligent", does that really mean people who like coffee but prefer orange are intelligent?

> Statistically reliable, in what sense?

Let's just say that if astrology was reliable, people would follow their horoscopes.

Astrology has a great test-retest reliability. Every time someone tests your sign, the same answer comes out. That's what it means. Same for personality inventories: people's replies don't vary much over time. As astrology shows, that doesn't mean you can draw conclusions.

But star sign doesn't really correlate with personality traits. People with similar horoscopes can have vastly different personalities. People with similar Big 5 scores have similar personalities.

> were complete ass pulls

Boy is that a dismissive overstatement. Briggs began thinking about personality in 1917, intentionally read biographies to get more data on personalty, and incorporated Jung's ideas in 1923 when his book was published, albeit with her own interpretation. Meyers did an apprenticeship with a test developer to learn how to make tests. So both Meyers and Briggs were, in fact, reasonably well versed in the field. They did not even make the test until 1944, after they'd been thinking about these ideas for over 20 years. Note that the Big 5 was not developed until the 1980s, almost half a century after they developed their theory. Ass-pulling this is not.

Wikipedia hates on the MBTI, but the actual test that you have to pay for is hardly "Barnum effect, flattery, and confirmation bias": you have 50 or so questions that you have to choose between two options of what you do. Now, you do have to answer what you actually would do rather than what you think is best or what you should do or what you should do culturally, so there is some confounding effects there.

https://en.wikipedia.org/wiki/Myers%E2%80%93Briggs_Type_Indi...

More recently I’ve been considering neuro-divergence in the context of these personality types. Specially ADHD and Autism. ADHD by example has a whole system effect for someone, with resultant strengths and challenges very much informing part someone’s personality.

Folks with Autism often have traits of being: direct, seek deep details, are able to stick with a few topics to deep excellence, can sometimes struggle reading people’s emotions.

And yes while generalisations, I wonder on how different neurodivergence’s may be hiding in plain sight within some of these personality systems.

I’ve def seen jokes online that notice that certain MBTI profiles are commonly gateway diagnosis for say ADHD/Autism.

Interested in others thoughts and takes on this.

Things like MBTI and enneagram and such are frameworks to talk about cognitive functions and all that. Holding them to the same standards as physics is a bit silly.

They don't talk about cognitive functions. Those would be things like vision and memory. And I'm not holding them to the same standards, but that doesn't mean that any old balony is fine. People attach a certain value to psychological tests, and they should realize that the whole thing is rather arbitrary.

You're right that they don't talk about cognitive functions. That's only when looking at these personality traits as a categorization of the outputs, which would feel like arbitrary curve fitting.

To be more nuanced, the underlying basis for MBTI on cognitive functions is about describing the input processing of various personality types, which isn't obvious (and completely obfuscated) if one is reading descriptions on a site like 16personalities.com.

It's a subtle but nuanced difference, but provides an avenue for therapists to understand how inputs in a client's life may affect them and their responses to situations. This is a more valid dive into the input-processing aspects of MBTI: https://www.typeinmind.com/type-theory

If you read the type descriptions on the typeinmind.com website, you'll see that the approach is very different. It may seem arbitrary and foo foo, but I'd challenge you to look deeper before completely writing it off as bologna.

Not all things fall under the purview of science, and not all things should. Whether someone makes claims of this being empirical is, for me and other pragmatists, aside the point: if it serves to help you understand yourself, or gives you some advantage in navigating interpersonal dynamics with others, it is valid. The proof is in the pudding. Nobody would question a literal pudding's empirical validity. They care only whether the recipe works.