To successfully develop and realise the personality, it is necessary to correctly understand the behaviour of people and adequately interact with the surrounding world. This ability first studied by E. Thorndike in 1920 was named the social intelligence. For 40 years, social intelligence has been studied as a theoretical construct, until in the 1960s J. Guildford operationalised the concept by creating a test to measure it. In the 1960s and 1980s, the social intelligence theory continued to be developed but empirical studies were also carried out that we are addressing today.
In the domestic research into social intelligence of this period, the focus was not on its dimension, as in the foreign research, but on understanding the essence of the given construct, its correlation with such concepts as social perception, communicative competence, practical intelligence, etc. [1,4,12]. Its functions and structure were analysed [7,15,16]. In a number of works, the psychological indicators of social intelligence were identified. These are accessible for observations or measurement manifestations of social intelligence, which make it possible to judge characteristics that are inaccessible to direct research. So, a high level of social intelligence is indicated by the high rates of social flexibility and empathy, the ability to reflect (A.A. Bodalev , A.I. Savenkov 10), self-control and self-confidence (V.N. Kunitsyna ), adaptability (A.L. Yuzhaninova ), the ability to build an internal action plan (E.V. Subbotsky ). Aggressiveness and shyness represent the indicators of a low level of social intelligence (V.N. Kunitsyna ) and low self-esteem (N.A. Kudryavtseva ).
However, in the 1990s, this problem got a much lower interest. This is strange because the social intelligence problem seems very relevant. Social intelligence contributes to the successful inclusion of people in the social environment and professional activities. Obviously, not only adults need this ability. As early as in primary school and even among pre-school children, the social intelligence development level has a tangible impact on various aspects of the child's life, therefore the formation of social intelligence should be among the priority areas of education and upbringing.
The ability to understand yourself and others becomes especially important in adolescence. At that time, an adolescent is influenced by many factors, both external (the relationships with peers, parents, teachers, the attitude of society toward development and maturation, new responsibilities in the family, etc.) and internal ones (changes in the body structure, hormonal changes in the body, comparing yourself with others, soul searching, etc.). The crisis age period is also complicated by the fact that a teenager's character has not been fully developed. Character's place is occupied by some outstanding, strengthened or accentuated features. These traits can become – and often do become – an obstacle to successful socialisation.
Thus, another key concept of our work is accentuation of personality traits. In shaping your character, under the influence of external and internal factors, a disharmonious development of its certain features, which is expressed in excessive expression, `sharpening' of some features and the weak manifestation of others, can be seen. This phenomenon was called by A.E. Licko `accentuation of personality traits' and is defined as the `hypertrophied severity of certain of its features, which causes an increased vulnerability of an individual in respect of certain influences and makes it difficult to adapt in certain situations' [8, p. 210]. If you compare personality traits and the indicators of social intelligence, then these lists will overlap in a significant way. In the research of A.M. Molokostova  it is directly shown that the level of social intelligence in various types of personality is expressed in different degrees.
Thus, the need for further research into the relationship between the accentuation of personality traits and social intelligence for different age and professional groups is evident.
It should also be noted that research in this area is complicated by the lack of unanimity of opinion on the nature of social intelligence (as discussed by famous psychologists of the mid-20th century) and ways to measure it. This is indicated, for example, by the analysis of the latest publications [2,12], whose authors note that social intelligence is still a category in the development and clarification, which requires further research with reference to specific areas of the professional activities. Some questions in respect of determining accentuations, their classifications, the limits of the norm, the connection with personality disorders, for example, also remain controversial. Perhaps, therefore, there has not been any detailed systematic study of the social intelligence development level in connection with accentuations of personality traits among teenage schoolchildren. Accordingly, there are not enough reliable data which can empower teachers who are obliged to enhance the social intellect of their pupils. All this determines the problem and relevance of our research.
As noted earlier, the quality indicators of high and low levels of social intelligence belong to various character types. This suggests that social intelligence and character variants are inter-related in a certain way. Since the character of the adolescent's personality is still being shaped, it is logical to assume that the accentuations of personality traits represent a significant factor in influencing the level of social intelligence achieved. Thus, the purpose of this study is to identify options for communication between social intelligence levels and types of accentuations of personality traits in adolescents.
The methodology of comparative analysis for the scientific literature is applied. Empirical data were obtained with the following psychodiagnostic methods: the methods of studying social intelligence by J. Guilford and M. O'Sullivan adapted by E.S. Mikhailova ; pathocharacterological diagnostic questionnaire by A.E. Lichko adapted by Yu. A. Parfenov . For statistical data processing, variational statistics methods were used.
The empirical base of the research was represented by the students of 7 grades, 74 people (42 girls and 32 boys) at the age of 13–14. Testing was organised in a group form.
• In adolescents with the labile and sensitive types of accentuation, the ability to foresee the consequences of behaviour and predict what will happen in the future is significantly higher than in those who do not express this trait of personality or is less pronounced (Х = 6.451 and 6.481, respectively, at Х = 3, p 0.05).
• Does this mean that the reasons for the same response are the same? – No. In the labile type, this ability is developed through adequate self-esteem. Adolescents of the labile type are well aware of their personality traits, they know that they are `moody people' and everything depends on their mood. In addition, the relations with their close people are very important for them; therefore, in order to avoid rejection on their part, which is a `weakness' for this type, they should be able to foresee the consequences of their own and others' behaviour. The opposite is also true, adolescents, who can foresee the consequences of their actions, are more susceptible to mood swings. So, the labile and sensitive accentuates are more able to foresee future people's actions based on the analysis of real-life situations (family, friends, business), predict events based on the understanding of feelings, thoughts, intentions of communication participants and also build strategies to achieve goals, although due to different reasons. In the sensitive type, this ability is a consequence of their desire to avoid conflicts, mockery and other problems in the relationships.
• Thus, both labile and sensitive accentuates are more capable of anticipating further actions of people on the basis of an analysis of the real situations of communication (family, business and friendly), predict events based on understanding the feelings, thoughts, intentions of the communication parties, and build strategies for one's own behaviour to achieve the goals set, although for different reasons.
• The accentuation of the excitable type is negatively associated with the level of development of such a parameter of social intelligence as the logical generalisation ability, the identification of common essential signs in various non-verbal human responses (Х = 6.699; Х = 3.843; p 0.05). That means that hyperexcitable adolescents have problems with the non-verbal logic of behaviour.
• The measurements of such a parameter of social intelligence as the ability to understand the change in the value of similar verbal responses of a person depending on the context of the situation, which causes them, is significantly higher in adolescents with a demonstrative type of accentuation of personality traits (Х = 4.536, Х = 3.843; p 0.05). They can correctly understand the speech expression of other people and correctly build their speech. They well feel how the meaning of the spoken words changes depending on the context change, and they `do not fall out' of the situation.
• The analysis of differences in the types of accentuations between the groups of boys and girls showed that girls more often have hypertensive (Х = 7.17; Х = 6.635; p 0.01) and demonstrative (Х = 3.85; Х = 3.843; p 0.05) accentuations.
A comparison of our results with the results obtained in studies on similar subjects showed the following:
• We found that the level of social intelligence is negatively related to the excitability of an adolescent. This, at least, does not contradict the outcome obtained by V.N. Kunitsyna , which showed a similar connection between social intelligence and human aggressiveness.
• We found that the level of social intelligence is positively related to the ability to manipulate people. This corresponds to the results of E.V. Subbotsky (14).
• We found that the level of social intelligence is positively related to various manifestations and components of the adolescent's sociability. This corresponds to the results of the research conducted by E.A. Kovalenko .
• For measuring social intelligence, like us, A.M. Molokostova used the Guilford–O'Sullivan test . Our results coincided in the sense that both studies showed that the level of social intelligence in different types of personality (character) is expressed in different ways. There are types of personality traits inherent in people with a high level of social intelligence, and there are personality types for which a high level of social intelligence is not inherent.
Thus, the facts we found concerning the connection of social intelligence and accentuations do not contradict the results of their predecessors, in general.
The detected difference in the incidence of demonstrative and hypertimic accentuations in girls appears to be quite a trivial result well-known among pedagogues. The hypertimism and demonstrativeness expressed, although not excessively, are the qualities necessary for leaders, and the school asset in the adolescent period is mainly represented by girls.
Our research is correlative, and this imposes obvious limitations on the results. As is known, any correlation found leads to two problems, the problem of direction and the problem of the third variable. If the principal way of solving these problems is known, then its practical application requires greater efforts. In any case, a special study is required. We have identified the links between the levels – high and low – of social intelligence and the accentuations of personality traits. But what influences what? What is the cause, and what is the consequence? Here is a question that we cannot answer directly (the problem of orientation). For a researcher, sometimes the context may be helpful in identifying the direction of the research for an answer while it is life and professional experience for practitioners. In any case, you should use the results of correlation studies with caution, so as not to fall prey to the illusion of knowledge of the cause–effect relationship. We would like to remind our reader of that. However, it happens that knowledge of the fact of correlation itself is enough to get positive results based thereon. In our case, this means that, based on the discovered non-random connections of social intelligence with the accentuations of personality traits, it is possible to control the development of social intelligence through smoothing accentuations and vice versa.
We calculated the correlation between the four subtests of the Guildford–O'Sullivan test – four abilities in the structure of social intelligence and ten scales of the Lichko test. Four statistically significant correlations were found. Only four but labile, sensitive, excitable and demonstrative adolescent accentuates are not unusual. Therefore, knowing the relationship between their personality features and social intelligence, the development of which is one of the priorities of teachers, is an effective tool for teachers. We understand that this tool is not yet quite advanced and we plan further research in the area.