Personal Experiment using Reinin Dichotomies for finding type
I once looked through all the Reinen dichotomies, picked one of each for myself without knowing which of the 16 types most of them were associated with, and came out with an actual type. Someone said that's Forer effect but I thought Forer effect was about reading something that's supposed to be about you that's vague, not trying to describe yourself. If Reinen dichotomies were really vague and I was picking them randomly then I should have had about a 16 * 215 (or 1 in 2048) chance of landing on an actual type, and shouldn't have gotten even remotely near any of them. So, I'd like for everyone who doesn't know which dichotomies are associated with which type to do the same. Even better, I'd like for people to gather people who know nothing about socionics to describe themselves in terms of the dichotomies and I'll try to do a statistical analysis of it all and see if people hit or cluster around type values much more than chance. If there are really 15 completely independent dichotomies then that shouldn't happen, but if they're not independent it won't. I largely think it doesn't matter if people are actually describing their behaviors accurately or not, because if the Reinen dichotomies cluster around types, they'll end up describing an idealized type and the same effect will still be there.
I'll probably have to talk to some of my science/engineering friends (or acquaintances, but that's an awkward word) to figure out how to analyze the data rigorously. I do figure we'll be able to get enough data to find an effect if even like 20-30 people here can ask questions of an average of 3 people each. I don't see why we can't do grassroots science since we don't really need fancy equipment for this, and some scientists actually want more people to do grassroots science. Of course it won't be the last word since the testing conditions are not super stringent but we need to get this started. This would basically validate the typology system (although not necessarily all of the other tidbits of the theory) if it were proven.
Last edited by ex nihilo; 10-02-2016 at 08:56 AM.
It's only meaningful how many of RDs have matched from the ones you did not know beforhand to which types they relate. Taking into account you could to forget you knew about them. There is also a probability of randomness.
Originally Posted by Schildmaid
Reinin's dichotomies are baseless bs. Your actual types is also unknown as there are a lot of mistakes.
So your "experiment" is bs too.
There were experiments to link RDs with types, - by tests. At least, some of them (Lytov) have shown no correlation, while Talanov claims he've found it but without good basis for this.
There is no even theoretical single-valuedness how to interpret some of RDs between different typers. It's low quality hypothesis.
> I'll probably have to talk to some of my science/engineering friends
You should look on lame theoretical basis of RDs. And what are normal experiments.
Last edited by Sol; 10-02-2016 at 09:16 AM.
Landlord of the Dog and Duck
I think including the 4 Jungian dichotomies as "Reinin" dichotomies in such a test would inflate things, and should be avoided. There are other dichotomies such as merry/serious and static/dynamic that are either not true Reinin dichotomies or are so well known that they could not be independently assessed.
When I first attempted such an exercise for myself years ago, I would have been typed ILE by the method.
Most people don't even know what socionics is though, so I doubt they could be influenced by these. Just go around giving people questionnaires about themselves without telling them what it's about and see which of all of the dichotomies they pick as describing themselves. Then we compare the tendency of these collections of dichotomies to match types to a good P value for these kinds of "soft science" experiments and if it's significant that means it's onto something. It's sort of silly to dismiss something out-of-hand just because you find it unlikely. Most people can easily get their friends and even acquaintances or strangers to do silly online personality tests like "Which Game of Thrones House are You?" and "What Country should You Really have been Born in?" and those tend to have about or even more than than 15 questions, so this shouldn't be hard at all to get people to do. I think I'll write up a questionnaire and then we can find some way to distribute it to people without giving them too much information on what it's for but not so little that they're completely confused. We could even survey people on the Internet (all you shyer folk) as long as they don't have too much information on what it's for.
Originally Posted by Subteigh
I think I'm going to have to write up the questionnaire to get it as value-neutral as possible. For example, positivist vs. negativist could become really value-laden at least in American culture because you're supposed to "see the glass half full" as people say, but if the question was solely worded as "I prefer to look for problems to fix" vs. "I prefer to look for strong points to improve" that seems a lot more value-neutral even though the glass half full/half empty thing is implicit in that. Asking/declaring seems like it could also be value-laden if asked wrong, but asking people if they prefer to engage with groups of people vs. one-on-one seems to be implicit in that dichotomy so we could ask for that. Yielding/obstinate seems to be the hardest one to ask about, but that might just be because of my personal values rather than any common cultural values.
Last edited by ex nihilo; 10-02-2016 at 11:16 PM.
A year ago, I took a C# class and for the final assignment, I created a working reinin dichotomy test in comand prompt. I gave a paper version to seven people who I knew their type, and inputed their results into my program and each of them got their correct type. I was in the process of working with someone to make this into a online tool, but they fell through. I know exactly how my algorithm works, the problem is the actual website. If anyone would like to help, we could make a complete working test over thanksgiving break.
@sindri Wow... I'm so late. That seems like a weird thing to do for a class, but probably much more useful and cool than what most people have done for that kind of thing as well.
@Barubary Yes, it does...