Test of the Relationship Between Traits and Their Reinin Images

So what, after all, do we obtain for the correlations between traits and their images constructed from products of other traits?

(more precisely, the square roots of the absolute values of these pairwise products, multiplied by the sign of the product)

PRESENTING THE COMPLETE EXPERIMENTAL DATA

Unfortunately, they are discouraging for the theory of Reinin traits (which assumes nonlinear interaction of traits) and for Model A (which is based on the discrete nature of psychotypes).

The point is that although, “on average,” each trait correlates with pairwise products of other traits, this correlation is achieved exclusively and solely due to respondents who are well acquainted with socionics theory. But for socionics-naive respondents (for whom both the possible influence of autosuggestion effects associated with unconsciously adjusting their answers to fit a type, and the increase in interest in socionics when their actual type coincides with a classical one, are minimal), these correlations are equal to zero, which points rather to the absence of a real nonlinear interaction between socionic traits in the human psyche.

Also indicative is the high correlation (+0.75) between, on the one hand, the correlational relationship of a Reinin trait with its images constructed from other Reinin traits and, on the other hand – the sigma of this trait in actual questionnaire measurements. The point is that the sigma of a trait in actual measurements indicates respondents’ sensitivity to the perception of this trait (consciously or unconsciously).

In fact, there is nothing unexpected in the absence of correlations between traits and their images. Such a nonzero correlation could arise, as can be proved mathematically, only in one case – if there were at least some boundaries between sociotypes, even if only in the form of small depressions in statistical density between them. If, however, the multidimensional trait space (15-dimensional) is absolutely homogeneous in its density (that is, devoid of boundaries between types), then a necessarily zero correlation between traits and their Reinin images follows directly from this. The presence of positive correlations in the subsample of more experienced socionists indicates precisely that only in this subsample is a decline in statistical density observed at the boundaries between types.

AND NOW, THE MAIN CONCLUSIONS OF THE STUDY IN SIMPLE WORDS:

In Aušra’s theory, the set of poles of all 15 Reinin traits for any sociotype is fixed. For example, an LII is not only an introvert, a logical type, an intuitive type, and a rational type, but must also necessarily be Judicious and Static, and cannot be Dynamic. The Reinin relations follow from this: Static=Extravert*Irrational=Introvert*Rational (and so on). Accordingly, if we take the Extraversion/Introversion and Rationality/Irrationality traits and multiply them, their product should correlate with Static/Dynamic. These correlations are what were tested in the study. It was tested whether Static correlates with the product of Extraversion and Irrationality, whether Democratic (measured directly) correlates with the product of Intuition and Logic, and so on, and so forth.

Until now there had been hope that the Reinin relations between traits (where each trait is generated by the product of others) held at least partially, that is, that the above correlations were not “empty” - this would mean that between the types, in the psychological 15-dimensional space of 15 factor-traits, there were at least hints of boundaries between types (in the form of a decline in the population density of respondents at those boundaries).

But now it turns out that, alas, they do not. The hope has vanished. That is, although a weak correlation between traits and the “Reinin” products of other traits does exist, and therefore a decline in density at the boundaries between types does after all exist, this decline is observed only in the group of more experienced socionists who are well acquainted with socionics theory. And it IS NOT OBSERVED in the group of people who are only very slightly acquainted with socionics (for them, the correlation between traits and pairwise products of other traits in their prescribed “Reinin trait pairs” - zero).

The presence of boundaries between types (with partial fulfillment of the Reinin relations) specifically and only in the “experienced” group is, in that case, explained not by the actual physiology of the brain in all people, nor by the laws and rules of some “information metabolism” in the brain, but either by the fact that the “experienced” partially adjust their answers (unconsciously) to Aušra’s theory, or by the fact that those for whom the descriptions of “classical” sociotypes fit best are simply the ones who become “experienced.” (and those for whom they fit poorly quickly become disappointed in socionics and simply do not remain in it).

How great is this disappointment for socionics? How dangerous is it for its scientific future?

It is not great and not frightening, the concerns are unfounded.

Although one of the important postulates of the original “classical socionics” (namely, the existence of boundaries between types) has been completely refuted, socionics by no means ends there, because it can perfectly well exist without this postulate, just as it can exist without Model A altogether.

It has functions of the psyche – this is its most important and most valuable innovative contribution to the psychology of individual differences.

It has 16 types of intertype relations, including the asymmetric relations of supervision and benefit. Intertype relations, their differing degrees of comfort, and their role in the formation of romantic and friendship pairs - all of this has also already been tested and confirmed (including in our studies) and constitutes one of the greatest contributions of socionics to modern psychology.

Finally, socionics has long used 15 personality factors with real substantive content, rather than 5, as does the rest of modern academic psychology of individual differences. And therefore – socionics is capable of providing a much more detailed predictive description of the personality and behavior of any person.

That is, the difference between them, between socionics and other modern psychological approaches in differential psychology, is the same as the difference between a 15-year-old adolescent speeding on a motorcycle and already kissing on a bench, and a 5-year-old child playing with blocks.

And that cannot be changed.

Table 1

Linear correlation between traits and their representations - in the form of the square root (with sign correction) of pairwise products of other traits

(1338 respondents to sociodiagnostic questionnaires, forms 078, 089, 090 and 091)

Table 2

Linear correlation between traits and their representations - in the form of the square root (with sign correction) of pairwise products of other traits

(1338 respondents to sociodiagnostic questionnaires, forms 078, 089, 090 and 091)

Table 3

Linear correlation between traits and their representations - in the form of the square root (with sign correction) of pairwise products of other traits

(1338 respondents to sociodiagnostic questionnaires, forms 078, 089, 090 and 091)

Table 4 - summary

Linear correlation between traits and their representations - in the form of the square root (with sign correction) of pairwise products of other traits. Correlations for the sum of 7 pairwise products of traits

(1338 respondents to sociodiagnostic questionnaires, forms 078, 089, 090 and 091)

Linear correlation (across the 15 traits) between the correlation values in the 4th column of the summary table and the sigma (standard deviation) of the measured values of the corresponding trait in samples of respondents with different levels of knowledge of socionics and in the total sample of 1338 respondents:

V. L. Talanov, Feb. 2021