What's Wrong with Socionics (And Analytics...)

[jason_m]

Recently, out of both career and personal interest, I have been studying analytics - the application of statistics and computers to analyze data for different purposes (e.g., to forecast events). I have found a few problems with it. For instance, let's say you want to determine which goods in a market basket are purchased together. You then find that pencils and paper are highly correlated - 'found together.' But, according to analytics, that relationship isn't interesting - because this is exactly what you'd expect people to buy! However, if you were to find that diapers and beer were often purchased together, now that would be interesting! What is wrong with this picture? In using this approach, the problem is that cause and effect relationships are completely ignored. The possibility of people commonly buying both diapers and beer together might just be a statistical fluke... In these cases, analytics techniques then just become a blind formula that you plug values into without using any reasoning or logic. But isn't that exactly what some of us are doing with socionics? Sometimes people completely ignore what they like in a person (the key cause and effect relationship) and just blindly follow the formula in order to determine they're dual. That's also the problem with some people's typing of others: "He must be Type X, so why isn't he going with Type Y?" Maybe it's because he doesn't like Type Y... That person who is criticizing them would then be making the same mistake of just blindly following a formula without focusing on the logic behind the other person's self-typing.

Now, does that mean that socionics (or even analytics) is severely misguided? I would say 'no.' I think that socionics can be used with success if you focus on what you like in a person. Even with analytics, I can think of all kinds of applications for it. The absolute key to both is that they cannot be used blindly without any thinking or reasoning. And even when you do use them logically, that does not guarantee success for either endeavour.

jason_m

[strangeling]

If anybody uses socionics to explain other people, rather than getting to know them, then lol.

Personally, I'd say my interest in socionics is to find over-arching themes between different people and their behavior, but not explain why they do the things they do. That takes getting to know someone. But I'll be honest and say that there are some socionics themes I instinctively avoid because I have little or no interest in them. Si themes where people care so much about impressions and how something looks or feels to them is completely foreign to me. So are Fe values where everybody expresses themselves in some kind of particular way, instead of just being real. But I'm probably ILI, even though a lot of people here seem pretty keen that I'm LII or alpha NT or some shit (maybe not anymore).