Is the American statistics of types and intertype relations under the test of I.Myers-Briggs reliable?

Bukalov A.V.

The test of I.Myers-Briggs is widely used in America for definition of Jung-Myers-Briggs psychological types. The following results (percentage to a total number of tested) were obtained [1] on testing the large groups of the population:

(ENTP)	5	(ENFJ)	6	(ESFP)	13	(ESTJ)	18
(ISFP)	5	(ISTJ)	6	(INTP)	1	(INFJ)	1
(ESFJ)	13	(ESTP)	13	(ENTJ)	5	(ENFJ)	5
(INTJ)	1	(INFP)	1	(ISFJ)	6	(ISTP)	6

This statistics is regularly used in the American literature on typology, and the attempts of its critical consideration are not known for us.

Abnormally small number of some types (1% each) is characteristic for the given statistics of type allocation in a population. Mean time our studies show that the dispersion of type quantity really exists, but never reaches such abnormal values (for example, 1% and 13% ). Such anomaly of type allocation can be explained either by presence of the constant error in use of MBTI, or by nonrepresentative sample.

The elementary analysis of group of types, which percentage in a population, by the opinion of the American researchers, makes 1% comes to the same result. These are , , , , which all are intuitive introverts, and either they assisted in testing minimally (that quite probably by virtue of their psychological features), or the MBTI gives an error in this group of types. This is not surprising: it is very difficult to make the test adequate for all 16 psychological types. However MBTI is the gauge, on which basis all researches on Jung-Myers-Briggs typology are created, with the exception of socionics, developing independently.

Therefore we can study the described results from positions of the socionic analysis, which considers not merely typology, but holistic system of 16 types as socion. Socion is known to consist from 4 quadras (4 types each), and quadras, in turn, consist of 2 dual diads (2 types each). The statistics on quadras should be most equable from the side of socionics. And really the number of types belonging to every quadra in the American population makes:

quadra a (, , , )	24
quadra b (, , , )	26
quadra c (, , , )	25
quadra d (, , , )	25

As we see, the deviation on allocation of 4 quadras is minimal and practically is equable (25 %).
Let's consider allocation of dual diads (expected equable allocation is 12,5 %)

α 1 (, )	10	α 2 (, )	14
b 1 (, )	12	b 2 (, )	14
c 1 (, )	14	c 2 (, )	11
d 1 (, )	14	d 2 (, )	11

Thus the deviation from equable allocation on percentage of a-diads makes 20%. For b-, c- and d-diads the deviation does not exceed 12%.

So, the abnormal allocation of types in a-2, b-2, c-1 and d-1 diads not reflects a real situation, but is a consequence of defined response of the examinees group (namely, of intuitive introverts) on the MBTI questions. This conclusion proves to be true also by type allocation close to equable, in a-1, b-1, c-2 and d-2 diads, for which the deviation from norm makes:

a1:	20%	a1:	20%
b1:	4%	b1:	4%
c2:	20%	c2:	4%
d2:	20%	d2:	4%

Thus the deviation does not exceed 20%, that coincides with deviation for diads of all socion. It is connected in our opinion with sharply expressed extraverted, expansionary behavioural stance, cultivated in the American culture, which influences a person from childhood. Therefore some types, like intuitive introverts, "mimic" their duals – extravert sensorics, because this helps their adaptation in the American society. Certainly, it does not mean, that their type of the person change: "mimicry" is rather superficial and manifests in extraverted social behaviour (including the answers to the tests). However true type is determined not by the answers to the tests, but by internal psychological structure, which can be detected not by the test, but by individual experts' work with the people.

		EGO
		SuperID

As I.Myers-Briggs has placed, allocation of marriages in America for psychological types, with coincidence of Jung scales of the partners, makes:

The conversion of these data in socionic relations gives the following:

It means, that approximately 28±2% on statistics of I.Myers-Briggs fall on intraquadral socionic relations (dualization, activation, mirror and identity). 28±2% fall on the relations between orthogonal quadras (quasi-identical, full opposition, superego and conflict). 28±2% fall on the relations between related quadras (social order, revision, congeneric and semi-dualization). 22±2% fall on the relations between unrelated quadras (social order, revision, business and mirage).

Because of detected error of the MBTI, the amount of the intertype relations becomes other. In fact, the equal content of the intraquadral relations and more unfavorable relations with orthogonal quadra calls questions.

Actually results of the MBTI demonstrate almost equal probability of a marriage in couples with various socionic relations, that is randomness of the partner choice. Such result can be explained only by large systematic error of the MBTI, and points on its nonvalidity.

At the same time empirical results on statistics of married couples, obtained by us, quite correspond to socionic regularity [3]. The amount of marriages with dual relations makes 45% from total, and intraquadral relations makes 64% of all relations in the investigated married couples. These results are well explained from the side of information metabolism model (model А) -- an interaction on horizontal blocks of a model A predominates over interactions on vertical blocks.

Only 4% of the marriages on the American data fall on the conflict relations. D.Keirsey [1] recommends these relations (together with the relations of a revision) as best at that. The sharp contradiction of dogmatic stance of the author on attractiveness of types with completely opposite scales to the real facts is present. From data of many researchers-socionics it is easy to understand the harm, which such "recommendations" of D.Keirsey have done and continue to do: because conflict and revision relations are the heaviest, unpleasant and, moreover, are fraught with psychosomatic diseases of the partners [2]. Let's note that D.Keirsey has based his recommendations for compatibility on his own ideas and postulates [1], but not on the real facts, unlike other studies.

Here methodological value of socionics, which shows an objective picture, irrespective of ideological, ethno-cultural and other stances of the researches, becomes apparent.

1. David Keirsey, Marilyn Bates. Please understand me. - Gnosology Books Ltd. 1984.
2. Aushra Augustinavichute  Human Dualistic Nature // Socionics, mentology and personality psychology. — 1996. — №№ 1–3.
3. Bukalov A.V., Karpenko O.B., Chykyrysova G.V. Biodata of Married Couples in the Light of Socionics // Socionics, mentology and personality psychology. — 1999. — № 1. — P. 3–7.