That is interesting. I would expect there to be some significant differences in relationship quality among MB types even if the types are only somewhat correlated (under the assumption that socionics is correct).
One of the better sites on the topic is Rick DeLong’s Socionics.us. He says there is only roughly a 30% correlation between MBTI types and Socionics types. Boulakov is also skeptical of the validity of MBTI typings. Perhaps the correlation is not high enough to obtain meaningful results here. I will be updating my beliefs on the matter, as this implies most MBTI types are mistaken if socionics is valid.
Honestly though, it really does look a lot like motivated cognition on part of socionists. I mean, they do have a coherently self-consistent theory but reference to external data points are suspiciously scarce. They seemingly start with the assumption (based on anecdotal observations of Augusta, socionics’ founder, and others after her) that these relationship preferences between distinct types exist, find subjective validation, and then go from there to assert that the MBTI is just not accurate enough at determining the traits socionics is based on. So for example if two people who are claimed to be ISFp and ENTp (where lowercase p is “irrational”) do not get along, Socionists will say the typing is invalid rather than that the theory is wrong. But if relationships are the only acid test of a typing, and relationships are the only thing predicted by the typing, it’s turned into a vague “if you like these kinds of people you will like these kinds of people”.
However, it’s not entirely hopeless because there are more specific predictions to to validate. As an example, given a valid ISFp/ENTp pair, socionics also predicts the ISFp will be a supervisor (“supervision transmitter” in DeLong’s terms) for the INFj type, whereas ENTp will be the “request transmitter” or beneficiary for the INFj. So if you could design a set of experimental test situations where supervision and request are distinguishable from other types of interaction (perhaps a game of some sort), you could perhaps set up a series of meetings between test subjects and see if it checks out. You could verify a given dual pair by their interactions with a given supervision/request receiver first, then arrange a meeting between them and see if they have more compatibility than the control group. The same thing could be verified with the diad’s supervision/request transmitter type.
That is interesting. I would expect there to be some significant differences in relationship quality among MB types even if the types are only somewhat correlated (under the assumption that socionics is correct).
One of the better sites on the topic is Rick DeLong’s Socionics.us. He says there is only roughly a 30% correlation between MBTI types and Socionics types. Boulakov is also skeptical of the validity of MBTI typings. Perhaps the correlation is not high enough to obtain meaningful results here. I will be updating my beliefs on the matter, as this implies most MBTI types are mistaken if socionics is valid.
Honestly though, it really does look a lot like motivated cognition on part of socionists. I mean, they do have a coherently self-consistent theory but reference to external data points are suspiciously scarce. They seemingly start with the assumption (based on anecdotal observations of Augusta, socionics’ founder, and others after her) that these relationship preferences between distinct types exist, find subjective validation, and then go from there to assert that the MBTI is just not accurate enough at determining the traits socionics is based on. So for example if two people who are claimed to be ISFp and ENTp (where lowercase p is “irrational”) do not get along, Socionists will say the typing is invalid rather than that the theory is wrong. But if relationships are the only acid test of a typing, and relationships are the only thing predicted by the typing, it’s turned into a vague “if you like these kinds of people you will like these kinds of people”.
However, it’s not entirely hopeless because there are more specific predictions to to validate. As an example, given a valid ISFp/ENTp pair, socionics also predicts the ISFp will be a supervisor (“supervision transmitter” in DeLong’s terms) for the INFj type, whereas ENTp will be the “request transmitter” or beneficiary for the INFj. So if you could design a set of experimental test situations where supervision and request are distinguishable from other types of interaction (perhaps a game of some sort), you could perhaps set up a series of meetings between test subjects and see if it checks out. You could verify a given dual pair by their interactions with a given supervision/request receiver first, then arrange a meeting between them and see if they have more compatibility than the control group. The same thing could be verified with the diad’s supervision/request transmitter type.