Hacker News Reader

Comment by maze-le | Hacker News Reader

They might be heuristics, but how can they ever be useful? They can neither be used in the analysis of a person nor are they predictive...

View on HN

Managers like them as they try to make sense of how their reports behave and work together, people they know but not really. Sounds like some people like them as a way to make sense of issues with their personal relationships. People outside yourself make decisions that are frequently confusing, and people like this sort of categorization to make sense of them.

People liking something doesn't mean they are actually useful, though. Some folks like fake medicine because they feel like they are doing something useful - even when we can prove that no, "memory water" isn't curing your cold. People like horoscopes for your same reasons - because people are confusing and some sort of categorization is a way to make sense to them.

It doesn't actually mean it is useful or even accurate in any way - it just means you are doing something. You could probably put the energy into managing the environment or better communication or coming to terms with "Other people are complicated".

Sure but it's worth exploring why these are popular beyond simply writing it off as "people are dummies."

People have been exploring why astrology etc. are popular for as long as they have existed. We don’t have to redo it all from scratch and have the whole argument from first principles every time some some new popular pseudoscience comes along. We can just declare it bunk and move on.

IDK i don't see this stuff as all that different from doing like a k-means clustering or factor analysis on a data set, it's just an attempt to do it on human personality, which we know has variance and correlations. We don't freak out at market segmentation analysis that involves labels like "early adopters" and "laggards" which are imperfect but informative.

I think people just find the entire idea of labeling humans in this way offensive, and latch onto some of the claims that it's "science" as an avenue of attack.

Edit: And I think it's slightly better than just categorizing people based on their birthday like you do in Astrology, as you are using at least responses to a personality test in most instances.

Personalities doesn't cluster so k-means would not give anything interesting. Big 5 is pretty close to a factor analysis on a questionnaire data set though, so that's a reasonable though. Four colours is not based on anything like that though.

I'd say that astrology is better than DISC in that it's obviously bunk, the problem with the colours is that while some people claim to enjoy it as a pastime or conversation starter, some seem to take it further and base actual decisions on it. People reading horoscopes in a magazine wouldn't typically actually follow the advice therein.

> People reading horoscopes in a magazine wouldn't typically actually follow the advice therein.

How do you know? See also, by the way, Japanese people’s beliefs about blood types:

https://en.wikipedia.org/wiki/Blood_type_personality_theory#...

If predictive power doesn't matter why do we need any of these systems at all? We could just randomly assign people with numbers from 1-10 and be done with it. You are type 7. I am type 3. Now you can go ahead and assign whatever properties you like to these types and be happy.

Reply on news.ycombinator.com