Teacher personality: a review of psychological research and guidelines for a more comprehensive theory in educational psychology

ABSTRACT The current review aims to demonstrate that findings from personality theories can help educational psychology craft a more thorough explanation of the role of teacher personality in the educational process. This topic seemed to have been inadvertently omitted. The following five groups of studies in psychology and related fields (classified based on their research objectives) are critically analysed: studies of teacher typologies, studies of teachers’ desirable and undesirable features, studies of teachers’ professional behaviours and their influence on students, studies of teachers’ professional identity and studies of teacher personality within the framework of personality theories (particularly within the Five-Factor Model of personality). Arguments in support of personality theories, and methodological dilemmas in assessing teachers’ personalities involving data collection and research design are also discussed. The article suggests that those findings on teachers’ personalities acquired in the context of personality theories might currently serve as the best starting point for a more comprehensive psychological theory of teacher personality in educational psychology – particularly when considered alongside current knowledge regarding small social group management and aspects of learning and development theories. Thus, the current review might serve as the basis for a separate field within school psychology focused on the psychological aspects of the pedagogical profession.


Introduction and aims of the review
In the educational disciplines, there are relatively distinct research fields that focus on investigating different aspects of the teaching profession and teachers personalities. For example, within pedagogy, deontology focuses on the responsibilities and rights of teachers with regard to their students, whereas pedeutology involves the features of teachers as determined by their roles (Rosić, 2011). In the psychological sciences, there are no such specified research fields for teaching, which is surprising because there is a long-running consensus that the teacher's personality is the most important and complex variable in the educational process (Evans, 1959;Getzels & Jackson, 1963;Sanders & Rivers, 1996) and because the study of personality is primarily the task of psychology. Even in educational psychology, the role of personality characteristics of teachers is typically emphasized only in research addressing the following topics: (1) management styles in small social groups (these studies inevitably use Lewin, Lippitt, and White (1939) as the starting point), which remains primarily a social psychological issue (see, for example, Alexander & Winne, 2006;Vizek-Vidović, Rijavec, Vlahović-Štetić, & Miljković, 2003); (2) social learning theory (Bandura, 1999); (3) developmental theories such as those proposed by Piaget, Vygotsky or Bruner and their followers (Lutz & Huitt, 2004) or (4) school docimology, i.e. the study of how knowledge is assessed and measured (Grgin, 2001).
De Raad and Schouwenburg (1998) posited that studies that apply new insights from personality psychology to explain the role of different patterns of teachers personality traits in educational endeavours in a more comprehensive manner fall within a distinct group of research (De Raad & Schouwenburg, 1998). This approach was first introduced in the middle of the twentieth century (Lamke, 1951) and has been regularly engaged in since that time (Mount & Barrick, 1998;Rushton, Murray, & Paunonen, 1983) even experiencing an uptick in recent years (Aidla & Vadi, 2010;Genc, Pekić, & Genc, 2014;Göncz, Göncz, & Pekić, 2014;Li & Wu, 2011;Rushton, Morgan, & Richard, 2007). Nevertheless, this approach has never become dominant in the field.
The current review aims to demonstrate that it is precisely the research that relies on developments in personality theory that can help educational psychology establish a more thorough explanation of the role of teachers personality traits within education. In particular, the Five-Factor Model of personality, especially his variant based on the lexical hypothesis (Goldberg, 1992), can provide one of the best foundations for a relatively coherent framework for a more comprehensive psychological theory of teacher personality. This aim was formulated on the basis that this approach, more than other approaches, discusses and offers more potential solutions to several dilemmas concerning data collection techniques, to appropriate assessment instruments that allow for the selection of teachers, and raises the question of adequate research design. What might be at least equally important, due to this approach earlier objections that the investigations of teacher personality were conducted in a theoretical vacuum (Getzels & Jackson, 1963;Li & Wu, 2011), will be remedied to some extent. Because we consider the Five-Factor Model to be one of the most important starting points in researching teachers personalities in the teaching/learning process, some introductory remarks about this model needs to be given at the outset of this article.
In the multifaceted field of personality psychology, many researchers consider the Five-Factor Model of personality for the leading approach. This view is also accepted in this review.
Numerous sources have reported on the history of the Five-Factor Model (John, Naumann, & Soto, 2008) and discussed both arguments in its favour (Digman, 1990;McCrae & Costa, 1999;McCrae & John, 1992;Saucier & Goldberg, 1996) and against it (Block, 2010;McAdams, 1992;Paunonen & Jackson, 2000). However, it should be emphasized that the Five-Factor Model is a model of personality in the descriptive sense (Srivastava, 2014) that is based on the consensus among researchers in two separate research traditions: the tradition associated with studies of traits using personality questionnaires and the tradition associated with studies of personality traits used in natural language.
These two traditions have yielded largely consonant models, which state that the best way to describe the structure of personality is to arrange personality traits along five dimensions. Some researchers use the label 'Big Five' instead of Five-Factor Model. The acronyms 'NEOAC', 'OCEAN' or 'CANOE' refer to the same dimensions, which are as follows: neuroticism (N), extraversion (E), openness to experience (O), agreeableness (A) and conscientiousness (C). A description of the personality traits that are representative for the extreme ends of these dimensions is given in the corresponding subsection of the article.
To achieve the above-stated goal, we first conducted an extensive analysis of psychological and educational studies in which the notion 'teacher personality' is used as a variable (not only as an independent or dependent variable but also as a moderator or mediator variable). This analysis included handbooks, monographs andprimarilyresearch articles (with keywords such as teacher personality characteristics, teacher ratings, school teachers, teacher identity and teacher education) in representative journals in the fields of educational psychology and educational studies (see References). Because the selected 80 sources can be structured in many ways, thereafter we searched for criteria by which the analysed studies could be classified into groups, and we opted to use the research objectives of these studies as the main criterion.
In accordance with this decision, we divided the studies into the following five groups: (1) teacher typologies; (2) studies of teachers' desirable and undesirable features; (3) studies of teachers' professional behaviours and their influence on students; (4) studies of teachers' professional identities and (5) studies of teacher personality (or on the different patterns of teachers personalities) within the framework of personality theories. These groups can be conditionally considered as both relatively distinct and simultaneously interconnectedand to a certain degree overlappingstages that show how (mainly psychological) knowledge on this topic has developed over time, beginning in the 1940s. Within groups, some representative sources have been organized chronologically. Where it was deemed appropriate or possible, further subdivisions of studies within groups were made using a similar methodology or using research trends. The merits and weaknesses of the presented approaches were also specified, particularly with regard to the proposed approach based on personality theories. As one of this review's goals, a separate chapter is devoted to the methodological dilemmas and open questions faced in assessing teachers' personalities, including data collection techniques and possible research designs. Finally, in the conclusion to this review, guidelines are included for future research and for the development of a more comprehensive psychological theory on teacher personality.

Teacher personality as a research topic in educational psychology and related fields
Studies on teacher personality in educational psychology and related fields are rather diverse with respect to their methodologies and their assumptions and aims. All forms of observation have been used, ranging from casual to experimental. Moreover, studies have been based on teachers self-evaluations and on assessments of teachers by their pupils, university students and other teachers or experts in the educational field. In addition, data were collected from essays on teachers personalities, evaluation scales, checklists, sociometry methods and personality and ability tests. Their classification by means of a single, well-defined criterion (and without overlapping between exhaustive categories) as demanded by formal logic is impossible. If a research objective is used as a criterion, it is conditionally possible to classify the studies into the following groups (within which further classifications into subgroups have been made): (1) mainly (but not exclusively) theoretical studies that aim to describe those personality types (or behaviour patterns) that are more or less suitable or unsuitable for a teacher, i.e. teachers typologies; (2) field research that aims to establish desirable and/or undesirable features of real teachers; (3) empirical studies addressing the professional behaviours of teachers and their impact on shifts in student personality or behaviour; (4) studies devoted to questions involving the professional identity of teachers and (5) research considering teachers personalities in the context of personality theories.

Teachers typologies
Research focusing on teachers typologies is dominated by descriptions of teachers' characteristics and their effects on students based on contemporaneous knowledge and the conceptual apparatus of educational psychology and pedagogy, psychiatry and psychopathology, or anthropology. Studies also differ in how they interpret the concept of type: within the five presented typologies, the first three types are considered categories, whereas the last two types are conceived of as dimensions. When outlining these studies, we have deliberately ignored open questions regarding the scientific status of the notions and categories they focus on. Many of these are rarely used in contemporary science or have changed the meaning. Nevertheless, the value of these studies, which were (mostly) undertaken in the second half of the twentieth century, is not merely historical, and it seems reasonable to assume that modern researchers should be familiar with them.
One of the most famous typologies of teachers was proposed by Caselmann (1949Caselmann ( / 1970 and describes both the shortcomings and the advantages of the suggested types. The author differentiates between 'paidotrop' teachers, who emphasize upbringing (thereby interested in individuals or groups), and 'logotrop' teachers, who emphasize education (teaching). The latter could be philosophically or professionally scientifically oriented. Philosophically oriented teachers prefer monologues, have poor contact with students and are oblivious to their frequently high demands. The scientific-professional orientation is characterized by good communication with students and teacher demands that are more adjustable to student abilities. From today's perspective, one might say that they create an environment for engaging dialogues and have a dynamic communication style. According to this typology, the style in which a teacher manages a group of students can be either authoritarian or social. With regard to didactic tools, a teacher can lean towards scientific-systematic, artistic or practical (pragmatic) methods in teaching. Their preference will depend on how much they are convinced that it is possible to manage teaching by relying on scientific knowledge or to what degree they consider teaching to be a type of an art or whether they show a readiness to apply all those actions they believe will increase the efficacy of teaching. Caselmann emphasized the importance of the teacher's awareness of the type they belong to, thus helping them prevent inappropriate actions.
Despite comprehensiveness and originality, which also largely applies to the four other typologies described herein, the Caselmann typology has several shortcomings (in addition to using three main criteria and three accessory or secondary criteria for constructing types): it is not empirically validated through cluster analyses or by other statistical methods in a satisfactory manner, and it is not supported by psychological instruments that allow for the selection of teachers.
The typology described by Adelson (1961) differentiates between teachers depending on whether they prefer their profession, their students or their institution. Those falling within the first group are characterized by outstanding professional knowledge; these teachers will enthusiastically convey information to students in a charismatic and narcissistic way. As a consequence, there is often a distance between the (expert) teacher and the (poorly informed) student. A teacher-facilitator is student-oriented and will encourage students to develop their potential. According to the author, this role demands outstanding altruism and dedication to the profession; in addition, it is exhausting and rarely successful. The third group includes those teachers who serve an acknowledged institution or idea. Such teachers will enthusiastically propagate a particular value system and convey it to students, which might diminish spontaneity in students and limit their ability to make judgements independently.
The advantage of this typology over other presented typologies, particularly Caselmann's, is that the author used only one criterion: teacher preference. However, as a result, this typology is also insufficiently comprehensive. Additionally, there is neither empirical validation nor operationalization of this typology.
According to the typology developed by Donáth (1977), individuals who have a schizoid, epileptoid or hysteroid characteror who could be described as school tyrants or instinctive educatorsshould not become pedagogues. A person with a schizoid character is reserved, cold and distanced, and such a person lacks spontaneity and is unable to express his/her own emotions. Such persons can occasionally lose their temper; more importantly, they are incapable of love, are often pessimistic and are cynical regarding others. An individual with epileptoid tendencies is characterized by frequent outbursts of anger (instead of muscle spasms) and (importantly) uncontrolled irritability. A hysteroid character refers to an egocentric person who behaves affectedlya 'poser' who is easy to see through and who is not sincere about his/her supposed ideas or perspectives. Donáth (1977) considers that neither 'school tyrants' nor the so-called 'instinctive' educators should teach in childcare or educational institutions. The former are full of bitterness and desire for revenge and power, but they are afraid of adults and therefore address such needs in a less dangerous school environment, taking it out on children. These teachers perceive students as constantly rebelling enemies, they set unrealistic demands of them and they are cynical and extremely pedantic, sometimes even mocking their students. On the other hand, the 'instinctive' educator is overly attached to students and aims not to lose their love. Such teachers do not approach realizing childcare needs and educational aims rationally. Parting from each class at the end of the year is a painful experience for them. It is as if such teachers want to follow the students through their entire lives, thus satisfying their need for parenthood through school.
G. Donáth used one criterion (although, from today's perspective, it involves questionable character traits) in her typology, but empirical validation and accompanying instruments are missing.
The dimensional view on teacher classroom behaviours offered by Ryans (1960aRyans ( , 1960b might also be understood as descriptions of typologies. Ryans argues that the manner in which teachers behave with students can indicate whether the teacher belongs to a friendly type, full of understanding and warmth, or whether he/she is more inclined to being distant and cold. A distinction is also made between teachers characterized by responsible and systematic behaviour and those who behave chaotically and recklessly. In addition, stimulation can be dominated by behaviour that is either routine or creative and rich in ideas. Ryans applied proper data gathering techniques and statistical methods (trained observers of teachers behaviours in the classroom, factor analysis), and the typology is supported by an instrument (Classroom Observation Record) that is suitable to make an assessment of certain aspects of teacher effectiveness. However, it is not an exaggeration to say that the question of how to select, train for or evaluate teacher effectiveness remains open until today. Baumrind's (1991) perspective regarding parenting styles can also be broadened to include teachers (Bernstein, 2011). As with parenting styles, the nurturing (or classroom management) actions of teachers can be positioned along control and affective dimensions. The former refers to the educator's desire to control (physically and psychologically) the behaviour of children, and the latter refers to manifested love and encouragement for students. Authoritative educators apply control within reasonable limits while showing warmth and affection. Authoritarian educators do not express enough love and control children excessively. In the case of indulgent educators, warmth is common, along with excessive freedom and few demands. Indifferent educators show a lack of control, followed by neglect of student needs and the absence of expressed affection. Research has shown that an authoritative teaching style is correlated with socially competent and responsible students. Authoritarian teaching is associated with inefficient social interaction and passivity in students. Indulgent and indifferent teaching frequently leads to immature, impulsive behaviour in students, along with poorly developed social skills. Correlations have been established in many studies (see Šašić, 2011), and the teaching style has been measured with instruments based on the typology described above (see, for example, Bosworth et al., 1996).
However, Baumrind's model of parenting, as an influential developmental model, is more concerned with the effects of parent-child interactions, and their consequences on the child development, but first of all in the family, and less in institutions. For a teacher it is, of cause, of the utmost importance to be familiar with this model, which is, in addition, many times checked and accompanied by verifiable instruments. Still, a teacher has several other roles, which are different from parent's roles, especially if he/ she is working in higher grades of regular schools, and not in correctional institutions. The five-factor model provides a far more elaborate taxonomy of personality traits having a role in education in comparison with Baumrind's styles of parenting. Because of that, the five-factor model is in many respects more important for issues concerning teacher personality than Baumrind's model of parenting.

Studies of the desirable and undesirable features of teachers
This type of research has the longest and most prolific history, but it is also the most heterogeneous with regard to applied methodology and its findings. The heterogeneity of findings can be explained to a significant degree by the fact that pedagogical activity is highly saturated by the ecological factors of the teaching environment. In particular, the expectations of a teacher, the functions and roles a teacher should play, and thus the desirability and undesirability of certain personality traits are significantly determined by the specific context in which teachers work. Consequently, it is understandable that in one particular educational context certain features of a teacher are highly appreciated, whereas in another setting, these same features might be considered irrelevant or even unwanted (and sometimes the teacher-professional should adjust his/her behaviour accordingly).
The participants in this type of research are typically asked to select certain desirable and undesirable characteristics of teachers from a predetermined list of options. In some studies, the participants are asked to freely state such characteristics or write an essay on the features of their most or least favourite teacher. Teachers self-evaluation is also sometimes employed.
In the early studies of Jersild (1940) and Witty (1947), which focused on primary and secondary school teachers, students' evaluations referred to the following: personality traits, temperament characteristics, the physical features of teachers and their management style (students have a preference for kindness, readiness to help, sense of humour, natural behaviour, good mood, kind-heartedness, young looks and good health) and teaching (didactic) qualities (i.e. clear and organized presentation of material without digressions and details). The significance of the latter increases with the age of students (Feldman, 1986). In a recent study, Vujačić (2008) emphasized that primary and secondary school teachers are aware of the expectations students have of them, and align their behaviours in accordance with these expectations. However, teachers judge themselves more harshly than students and are generally rather self-critical in regard to their practice.
A previously neglected issue regarding the expectations of college lecturers and university teachers was emphasized more in studies from the 1980s. Thus, Morgan and Knox (1987) noted that the most significant expectation from a teacher in the (medical) profession was to be a good role model. Ben-Chaim and Zoller (2001) emphasized the high level of agreement between a teacher's self-perception of personal style (the way they react to people and events in their environment) and the same evaluation made by their students.

Studies of teachers' professional behaviours and their influence on students
Some research regarding teachers' professional behaviours argues that it is illusory to expect that a general desirable profile of teachers' professional behaviours that is appropriate to all educational situations can be established. On this basis, an approach focusing on the concrete actions and procedures of teachers in a classroom seems justified. Such procedures include teaching methods and styles, communication, socio-emotional relations with students and management styles within a class. Typically, this research involves cataloguing the behaviours of efficient teachers and subsequently anticipating students' achievements (teacher-student interaction studies). These investigations were more prominent in the field at a time when family circumstances were thought to have a larger influence on students' school achievements than school experiences. Flanders (1974) has made a particularly important contribution to this line of research by elaborating upon how to conduct a micro-analysis of class interactions using the immediate observation of conversation in class. (In particular, 10 types of verbal behaviour of teachers and students in the class could be treated as a kind of typology, and from this perspective, there is some overlap with the first group of studies presented in this review.) Flanders' approach had a substantial impact on this line of research (Abrami, Perry, & Leventhal, 1982;Brophy & Good, 1986;Gage & Needels, 1989;Madike, 1980;Mercer, 1996;Šašić, 2011). More recently, Darling-Hammond (2000) and Wayne and Youngs (2003) found that a teacher's verbal skills were a good predictor of student success.
To fully comprehend the influence of teacher actions, it gradually became clear that more attention should be given to the mediating role of student characteristics. However, individual actions and teachers' interventionsthe effects they have on studentsare also determined by the teacher's personality to a great extent. Thus, it has been reported that university professors who suffer pronounced anxiety associated with teaching actually prefer minimized interactions with students, such as pair work or student presentations (Houlihan, Fraser, Fenwick, Fish, & Moeller, 2009). In one of his interviews, Mervis (2010) addressed the findings of American researchers who reported that the interactive teaching of physics at universities, along with the application of pair work and digital technology, was much more efficient than traditional teaching methods. It was suggested that such interactive teaching may be more efficient because it is based on students' thinking processes. Indeed, this idea was previously advocated by the critics of programmed teaching in their insistence upon the need to make 'the black box' more 'transparent' by harmonizing students' thinking activities and learning. The personality characteristics of teachers are seemingly not in the foreground here, although, of course, openness to novelty is a precondition for a lecturer considering this form of teaching. Again, we are led to the conclusion that the influence of teachers' actions on students (together with numerous other factors) is largely based also on teachers' personality traits and dimensions, which can best be explained within the context of psychological theories of personality.

Studies of teachers' professional identity
A separate group of contemporary (mostly qualitative) studies in this research field involves investigations of teachers personalities in the professional context. In these studies, the key variables are identity, self-concept, values and attitudes, and depth interviews and analyses of life stories are the most dominant techniques employed. One subgroup of these studies emphasizes the possible definitions and identification of professional identity characteristics (Beauchamp & Thomas, 2009;Beijaard, Meijer, & Verloop, 2004;Korthagen, 2004), whereas another subgroup highlights questions involving teacher formation through training and development, which are connected to the problems of teacher selection and education and to the possibilities of influences on teacher identity (Day, Kington, Stobart, & Sammons, 2006;Geijsel & Meijers, 2005;Korthagen, 2004). In these attempts at determining professional identity and its components, it is also emphasized that professional identity is a type of social identity that indicates the degree to which individuals define themselves as members of a profession and that is closely bound together not only with personal identity, job satisfaction, self-actualization and self-concept but also with the expectations of others towards the members of the profession. The professional identity of teachers also inevitably involves what teachers consider important in their profession. Some teachers emphasize expertise in the domain of subject matter knowledge and skills (subject matter experts), others emphasize expertise in the domain of didactical knowledge and skills regarding the planning, execution and evaluation of teaching and learning processes (didactical experts), and still others emphasize expertise in the domain of knowledge and skills that support students' social, emotional and moral development (pedagogical experts; see Beijaard, Verloop, & Vermunt, 2000). There is a consensus that issues related to the professional identity of teachers, the selection of candidates for the teaching profession, and improving the professional skills of teachers are crucial for raising the efficacy of teaching/learning processes, which also largely dependtogether with many other factorson the structure of the teacher's personality. Researchers in this field are mostly proponents of social investment theory (Srivastava, 2010). In accordance with this theory, some authors argue that people become more agreeable, conscientious and emotionally stable and less extraverted as they age (Caspi & Roberts, 2001), and, in addition, that conscientiousness, extraversion, agreeableness and emotional stability are positively associated with commitment to work (Hudson, Roberts, & Lodi-Smith, 2012). These arguments should be considered when reviewing the performance of teachers during different periods of their professional careers.

Personality theories and teacher personality
The main argument of this article is that findings from personality theories deserve a closer attention of educational psychology in their attempt to better explain the role of teachers personalities in the educational process. Currently, this approach is not sufficiently present in educational psychology, although psychology is populated with an abundance of personality theories that offer certain solutions regarding structural, developmental and motivational issues in human personality. In his review of several leading personality theories on the basis of six criteria (comprehensiveness, precision and testability, parsimony, empirical validity, and both heuristic and applied value), Cramer (2013) convincingly demonstrated that a combination of findings from different personality theories can provide the best solutions to some important life problems, which also includes the role of teacher personality in the educational processes. This is the best solution because there is not one single personality theory, or such a group of similar personality theories, which would meet all the requirements of a good theory. To be specific, a good personality theory 'should describe, explain, predict, and control phenomena and behaviour' (Cramer, 2013, p. 9). Some theories are better or less good in some requirements then other theories. That is the reason why in this review, which is limited to discussion of several investigations which rely on Cattell's, Eysenck's and Jung's theory, and on the Five-Factor Model, there is also an attempt to shortly point at findings from a different angle, suggested by other personality theories.
Namely, in some theories traits are treated as fixed, the recognition of cultural differences in the formation of personality is not sufficient, wider contextual issues, like situational pressures that may alter individuals' behaviour, are neglected. Developmental issues of children, which are so important in education, are in many personality theories ignored or insufficiently treated as well. Lamke (1951) was the first to assess teacher personality using questionnaire statements grounded in personality theory, namely Cattell's personality theory (Cattell & Mead, 2007). This early self-report provided a description of the traits and behaviours of successful and unsuccessful teachers. Using 16 dimensions, this author found that more-successful teachers achieved above-average results for the F and H dimensions. On the F dimension (liveliness), they were more talkative, typically in a better mood, more cheerful, calm, open and impulsive, whereas on the H dimension (social boldness), they were more gregarious, adventurous, artistic, they had more sentimental interests and they were more interested in the opposite sex than the norm. Less-successful teachers scored below average on the F dimension.
Surprisingly, Li and Wu (2011) found no differences between good and poor teachers in any of the four dimensions of Eysenck's personality model (Eysenck, 1970), namely psychoticism, extraversion, neuroticism and social conformity. In this study, college students evaluated the classes in different courses held by the same teachers during three years. Based on their assessment, the teachers were classified into three groups, which included 'good teacher', 'ordinary teacher' and 'bad teacher'. The self-reported measures of teachers in the 'good' and 'bad' groups on Eysenck's dimensions were not statistically different. Regardless of the results, the authors does not deny the importance of teachers personality patterns in teaching, and as one of the possible reasons for these results they consider that 'when college students evaluate teaching, they paid more attention to the teaching quality than to the teacher personality' (p. 763). However, Petrović-Bjekić (1997) refers to Birkinshaw's studies (which relied on the same theory) and demonstrated that extraverted and more emotionally stable teachers are more efficient in many aspects of their work than poor teachers.
Considering the dimensions like extraversion and neuroticism in the educational context, it should be noted that these constructs can have different meanings, and therefore different consequences in education. For example, Downes (2003, pp. 108-110) draws attention that as early as 1921, Jung characterized extraverted thought as formulaic and merely 'programmatic'. He holds that extraversion is basically cultural adaptation and conformity of thought and feeling. Extraverts draw energy from the external word and not from within, lacking a capacity to challenge the environment. So an education system simply promoting extraversion is a very limited one for producing thought and feeling that brings change to the environment and also deeper emotional communication. Jung treated introversion as having many positive qualities. As for neuroticism, it can be also understood within a Freudian framework of repression. As Sincoff (1992) pointed out, in that case, neuroticism is not necessarily linked to psychopathology and poor physical health, because there is evidence that 'repressors', who score low on trait anxiety and high on defensiveness, avoid negative self-relevant information, and also may be better at health behaviours that they perceive as under their personal control (for instance, repressors report less substance use). Rushton et al. (2007) used the Myers-Briggs Type Inventory (MBTI; Boyle, 1995), which is a questionnaire based on Jung's theory of psychological types (Jung, 1921(Jung, /1971. This theory posits that there are four psychological functions (sensation, intuition, feeling and thinking) through which the world is experienced. One of these is dominant and determines the preferences regarding how people perceive the world and make decisions. Notably, the four functions are expressed in either introverted or extraverted form. The theory is turned to practical use by constructing the MBTI for classifying people into types based on their preferences, and the characteristic features of these types are described by Myers and Myers (1995). An examination of a select group of the best teachers in the U.S.A. using this questionnaire revealed a predominance of the 'ENFP' type profile (Extraversion, Intuition, Feeling and Perceiving). For 'typical' teachers, there was a predominance of 'ESFJ' types (Extraversion, Sensing, Feeling and Judging). More-successful teachers were likely to be oriented towards their environment, be inclined to make friends, think aloud and be rather active (E). Such teachers are oriented to the future, prefer to live in a world with many possibilities and options, are focused on complex abstract problems, and consider the complete picturesometimes at the expense of details (N). They accept other peoples' values and are good at estimating their influence on decision-making (F). In addition, they prefer a flexible and adaptive life style, they lean towards spontaneity, they require more information when making decisions and they often do things at the last moment (P). By contrast, the 'average' group of teachers was likely to be dominated by S (Sensing) rather than N (Intuition) characteristics. Individuals in the 'average' group rely on sensible experiences; they are oriented to practical problems and believe that well-established routines should not be changed. In addition, the J (Judging) type is more likely than the P (Perceiving) type to have a preference for order and to be self-disciplined and well structured, with an organized lifestyle.
The five-factor model and teacher personality As already mentioned in the Introduction, where a brief description of the development of this model is also given, many researchers reflect on the Five-Factor Model as today's leading personality model. Popkins (1998) evaluated this model based on five criteria: compatibility, originality, application, taxonomy and universality. He holds that the model is consistent with other factor-analytical models (Cattell, Eysenck), and also even with Freud's psychoanalysis, because 'people who have a certain characteristic that falls at an extreme on the chart of one or more of the five variables are likely to have some sort of psychological abnormality associated with that trait' (Popkins, 1998, p. 1). He highly evaluates the model also with respect to taxonomy, originality, application (in the academic and experimental forum), and judges that the Five-Factor Model holds very well across cultural and linguistic lines. His opinion is in line with Srivastava (2010), who claims that this model is also a model of social perception because the evaluation of personal perception is made along these dimensions to predict one's behaviour in most cultures. However, because the variables of the model are too broad, Popkins (1998) admits that the model fails to anticipate behaviour in many situations.
In his critical appraisal of the Five-Factor Model, McAdams (1992) also emphasizes that the Big Five operate at such general level of analysis that trait scores may not be especially useful in the prediction of specific behaviour in particular situations. He also highlighted some more limitations of the model, for example, the failure to provide compelling causal explanations for human behaviour. Because of that he holds, that the Five-Factor Model, although an important model, is not an integrative theory of personality.
A description of the personality traits that are representative of individuals at the extreme ends of these dimensions is typically given in introductory personality textbooks. Therefore, we discuss below only those personality characteristics in this model that are essential to education, especially to the expectations for teachers.
With regard to the possible applications of the Five-Factor Model, it has most often been considered within the field of organizational behaviour and particularly in professional selection and occupational choice. For example, Liao and Lee (2009) revealed that engagement in the workplace was positively correlated with nearly all the dimensions, except for neuroticism, which presented negative associations. Barrick and Mount (1991) found that conscientiousness was a good predictor of success in different groups of professions and that extraversion was a good predictor of achievement in those professions in which social interaction was important. Judge, Heller, and Mount (2002) conducted a meta-analysis to establish the reported positive correlation between job satisfaction and conscientiousness, extraversion and agreeableness, and a negative correlation with neuroticism, although only the relations with neuroticism and extraversion were found in all the studies.
The Five-Factor Model can also be applied to research on teacher personality. Descriptions of the dimensions in connection with the expectations of effective teachers in contemporary schools and with regard to the applications of the Model within the field of organizational behaviour provide a good basis on which the preferred or undesirable personality characteristics of teachers can be examined. Thus, it seems justifiable to assume that pronounced neuroticism (which encompasses negative emotions, pessimism, low tolerance for frustration, impulsiveness and so forth) is not a desirable trait for teachers to have in most pedagogical situations or in interactions with students. It is not very likely that teachers with high neuroticism are capable to build dynamic communicative relationships to their students, and many of them will need a great deal of time to improve their social skills. With regard to the other dimensions, it is assumed that they would be more pronounced in a good teacher than in the general population. Moderate extraversion is desirable as it is associated with friendliness, self-confidence and positive emotions. Efficient teachers should be open to novelty, curious and creative and have a well-developed vocabulary. Moreover, such teachers should be receptive to unconventional ideas and beliefs and accepting of different cultural backgrounds and the various feelings and behaviours of their students. A high degree of agreeableness is also necessary for successful pedagogic work. A good teacher should be cooperative, willing to compromise, mild-mannered and benevolent with faith in mankind. In addition, a good teacher must also have a reasonable level of conscientiousness so that characteristics such as competence, order, sense of duty, planning, self-discipline, impulse control and dedication can emerge in his/her work.
On first sight, it could be said that these assumptions can be taken for granted. However, their acceptance includes also some risk for education: the danger to accept them unconditionally in education is that they prescribe a monolithic personality that is to be prescribed for teachers. This is not only a huge force towards cultural homogenization of an education system but also may not be suited to varying teaching contexts with different cultural groups. Some of the above assumptions have been subjected to empirical validations and have on the whole been confirmedat least insofar as they are not out of step with reality. For instance, Aidla and Vadi (2010) established that teachers from Estonia (particularly male teachers) were more agreeable and conscientious, and have lower scores on the neuroticism scale than the general population, which was consistent with the features attributed to them by both the public and recently graduated students. Genc et al. (2014) confirmed the assumption that students expected good teachers to have less emotional lability and to show more extraversion, openness, cooperation and conscientiousness than the general population. For neuroticism, the difference (Cohen's d ) is either small or medium in size, for cooperativeness of a medium size, and for extraversion, consciousness and openness to experience the difference is large. Also, the gender of students does not influence their expectations. Göncz et al. (2014) suggested that students preferred that their teachers have characteristics that contributed to pleasant interpersonal relations. In addition, students' selfevaluations for the dimensions of openness, agreeableness and neuroticism were better predictors of expectations of good teachers than self-evaluations of extraversion and conscientiousness. The authors concluded that students with a high degree of openness to experience and agreeableness, and low neuroticism (i. e. expressed emotional stability) requiered these domains to be even more pronounced in a good teacher, and conservative assessors prefered even greater conservativism from teachers (Göncz et al., 2014, p. 84).
This study also revealed that students of the social and humanistic sciences preferred more extraverted, open, agreeable and conscientious teachers than students of the natural and technical sciences. Furthermore, the self-assessed personality traits of natural and technical sciences students were better predictors of their expectations of good teachers than the personality traits of social and humanistic sciences students, which might indicate that this group of students has clearer and more unambiguous expectations of a good teacher.
Other relevant studies have not explicitly referred to the Five-Factor Model, but the features attributed to teachers in those studies can be considered characteristics of the various domains within this model and might nonetheless be of interest to the present overview. For instance, Suplicz (2009) found that students attributed emotional coldness and a lack of humour (most present in neuroticism and introversion) to the secondary school teachers they considered inadequate and considered poor knowledge of the profession and lecturing style (low conscientiousness) to be less relevant characteristics. Sánchez, Pecino, Rodríguez, and Melero (2011) investigated the expectations of social sciences students at universities in Andalusia and established that students expected their teachers to treat them with respect and understanding (which might be linked to agreeableness). These students also expected teachers to be open to cooperation, to disregard their personal characteristics (openness to experience) and to have good lecturing skills (conscientiousness).
With regard to earlier studies, similar expectations of teachers by students were reported by Feldman (1976). Gorgonia (1971) also stated that students predominantly favour emotional stability in an efficient teacher (which is comparable to the scale of neuroticism), followed by characteristics that might be associated with conscientiousness. The remaining expectations, which seemed to be of lesser importance, could be classified in terms of the other three dimensions. Three decades ago, Rushton et al. (1983) investigated university professors that were considered to have adopted two opposing professional roles: a lecturer and a researcher. These authors established certain differences in personality traits that depended on the degree of the professors' orientation towards the educational or the research aspect of their profession. For the education-orientated professors, features associated with openness to experience (e.g. being liberal, friendly, objective, supportive and non-authoritarian) were in the forefront. For the research-orientated professors, features characteristic of conscientiousness (e.g. being ambitious, persistent, dominant, less supporting and inclined to leadership) were dominant.
For purposes of the present review, studies on the personality correlates of success in other professions are also notable. It has been suggested that persons with different clusters of features prefer different professions (Rubinstein & Strul, 2007). This suggestion has been shown to be true for helping professions (Dimitrijević, Hanak, & Milojević, 2011) and is also characteristic of teachers (Harris, Vernon, Johnson, & Jang, 2006), suggesting that teachers may have a distinctive personality profile as a professional group. Focusing specifically on success, studies have shown that neuroticism is negatively correlated with success at work (Smithikrai, 2007) and with striving for management and efficacy (Niehoff, 2006). A positive correlation was revealed between extraversion and work success (Smithikrai, 2007). Niehoff (2006) suggested that a person who is open to experience addresses the need for curiosity and explores new possibilities through the choice of his/her profession. In the workplace, an agreeable person pays attention to interpersonal relations and respects rules; furthermore, he/she is interested in the welfare of others (Ashton & Lee, 2001) and is conscientiously dedicated to work (Li, Lin, & Chen, 2007).

Methodological dilemmas and some open issues in assessing the personality of teachers
There are many dilemmas in this area of research; most of these are linked to data collection techniques, to appropriate assessment instruments that allow for the selection of teachers, and to the problem of adequate research designs.
Some authors (Genc et al., 2014;Göncz et al., 2014) have recently drawn attention to the fact that the question of who can be considered a good or poor teacher and what his/her dominant personality traits are generally discussed in the context of social perception in the social psychology of education or in personality psychology and the psychology of individual differences in the context of knowledge about the structure of personality. The choice of research technique depends on the starting point.
The most commonly used technique until the mid-1980s was described by Feldman (1986), who catalogued 72 studies in which some researchers asked respondents to write a term paper on the characteristics of their most or least favourite teacher, whereas other researchers opted for the technique of free guidance of characteristics or offered respondents a choice from a list of characteristics. This instruction implies that a teacher is considered good or bad if the respondents perceive him or her as such. The question of which factors the attributed traits depend upon (i.e. what else, other than teachers' traits, is reflected in these traits) did not arise.
The problems associated with classifying responses and limited opportunities for indepth statistical analysis of the data are weak points in most such studies. Although these procedures continue to be used (e.g. Suplicz, 2009, or in research on the professional identity of teachers), other studies have made progress by employing personality questionnaires. This approach is essentially a modified version of the approach in which respondents choose from a given list of characteristics. However, a significant advantage of a technique that uses standardized personality questionnaires is that the same list of items (stimuli) is always offered. This standardization facilitates the replication of studies and allows comparisons to be made of results obtained from different samples. In addition, it can also enable results to be more generalized. On the other hand, it can be argued that the description of personality is limited only to those characteristics that the applied questionnaire measures. However, all the techniques have the same purpose, i.e. they are used for data collection in situations involving social perception, when both the assessor and the subject of the assessment are humans. The end result of all social assessments always features a significant proportion of variance that is not attributable to objectively existing external reality. Although high intersubjective agreement of independent appraisers is often taken as sufficient criterion to justify assessment, the issue of the proportion of subjective variance (due mostly to the personal characteristics of the appraiser) remains open. This problem encroaches on the core issues of the validity of introspective data in psychological research and, as such, is beyond the scope of this article.
Regarding the issue of the appropriate research design in this field, it is first necessary to stake out a position regarding the relationship between assessment and self-assessment. Some believe that these are quite different mediums that reflect the influence of a variety of factors on the obtained results, but Srivastava (2010) treats self-assessment as a variant of assessment in which the perceiver and the target are the same person, but both serve for social perception.
In contemporary research, the description of the personality of a (good) teacher is mostly undertaken by teachers' self-assessment on some type of personality questionnaire, and the results are compared with the self-assessment of a non-teacher group, or with a group of poor teachers on the same questionnaire to obtain the differences. In this manner, it is implicitly accepted that the best research design requires the comparison of two self-assessments. However, in real school situations, the influence of teachers depends on how the students perceive the teachers and not how teachers see themselves. From this perspective, students' estimates of teachers would be more relevant. Concurrently, the question of reference values for interpretation of these estimates arises. If we accept the above-mentioned opinion expressed by Srivastava (2010) for obtaining the description of the personality of a good teacher, then it is acceptable to compare students' evaluations of a good teacher with self-evaluations of the representatives of general population ('norms') as well. This comparison would represent another design in this field of research with the same weaknesses and strengths as other designs and data collection techniques but more in accordance with real educational practice. This research design has been so far relatively rarely used, probably because of the primarily theoretical interest of researchers on teacher personality.
Besides methodological problems, there are many open questions regarding the interpretation of results in this field of research. Every personality theory or model has its inherent limitations. Some theories are more successful in explanation of the structure, some others in explanation of development or dynamics of personality; some are more, some less contextualized. Because of that, it would be desirable to combine and supplement knowledge from different theories and models when interpreting regularities observed in some empirical investigation. Thus, the role of teachers personality in the educational process could be seen from the perspective of different theories and models.
Guidelines for future research and for the development of a more comprehensive theory Psychological research on teacher personality can undoubtedly increase the effectiveness of the educational process. The results of the outlined studies have numerous implications for pedagogical practice, particularly with respect to the selection of future teachers, their education and subsequent professional improvement, and the evaluation of their success. This review provides a brief outline of five groups of studies in this field. We believe the most promising of these groups of studies is the type of research that employs insights from personality theoriesand particularly the Five-Factor Modelas its starting point. Beyond the fact that this is one of mainstream models, such an approach has a number of (mostly methodological) advantages over other types of research. For instance, using this framework allows one to formulate more specific hypotheses that can be empirically verified. Thus, Göncz et al. (2014) found that evaluations of personality qualities of a teacher are determined by the personality characteristics of the assessors: correlational analyses of the relationship between self-assessment of students and their evaluation of a good teacher for the five domains of Five-Factor Model indicated about 20% of the common variance, i.e. a strong effect of the predictors on the criterion. In addition, this approach resolves certain methodological dilemmas regarding data collection, replicability and the generalization of findings. The common techniques for teacher data collection (such as student essays, freely stated responses regarding the features of a good or a bad teacher, or the selection of characteristics from a list), provide less-developed and elaborated possibilities for statistical data processing (in spite of the numerous advantages of qualitative research and non-standard techniques of data collection). However, when standardized theory-based personality questionnaires are used, research can be repeated under the same conditions, which enables the comparison of results obtained for different samples and, consequently, the possibility of generalizing the results is also increased. In addition, a theory-based questionnaire approach has greater explicatory power than the other approaches. Such an approach also has higher criterion validity because of the huge number of publications on personality instruments. On the basis of such advantages, we recommend the more often use of this approach in future studies.
Besides the Five-Factor Model, it would be desirable to rely also on other newer models or theories of the factor-analytic family, like HEXACO (Ashton & Lee, 2007;Lee & Ashton, 2013). This six-dimensional model of human personality was developed through similar methods as the Five-Factor Model and other trait taxonomies, but it is unique mainly due to the addition of the Honesty-Humility dimension. The six dimensions, or factors, include Honesty-Humility (H), Emotionality (E), Extraversion (X), Agreeableness (A), Conscientiousness (C) and Openness to Experience (O). The names of four of the HEXACO dimensions (all except H and E, of which E has much in common with Neuroticism dimension of Big Five) were adopted from the existing labels for the Big Five factors. An individual who stands high on H dimension is sincere, honest, faithful, loyal and unassuming, in contrast to individuals with low position on this dimension, which are sly, deceitful, greedy, hypocritical and boastful. This would also mean that low levels of H corresponds to higher levels of psychopathy, Machiavellianism and narcissism, the Dark Triad personality construct (Paulhus & Williams, 2002), which is represented inconsistently on measures of the Big Five. The HEXACO model can be successfully used in research when behaviours and traits found on the Honesty-Humility, Agreeableness and Emotionality dimensions are of specific interest, including the study of teacher personality. An individual teacher who scores low on the H factor may have a proclivity for anti-social acts. It is also probable that a teacher with a personality pattern of high levels of H and A, and low level of E, would have a tendency for pro-social altruistic behaviours and inclined towards forgiveness and tolerance, which is an important propensity in every aspect of the educational process. Unfortunately, there is a lack of investigations that would shed more light on these connections. The proposed framework for researching teacher personality relying on (factor-analytic) personality theories and models could at least somewhat alleviate this state of affairs.
Designing studies around an established theory of personality, such as the Five-Factor Model or some other factor-analytic trait models or theories, supplemented by findings from personality theories that have a different approach, is likely to ensure an abundance of empirically verified and verifiable findings on teacher personality at different levels of education, in various systems of education, and in relation to specific conditions, purposes, pupils and classroom dynamics. These insights might provide the core of a more comprehensive psychological theory on the role of teacher personality in the educational process.
In such a theory traits should not be treated as static, unchanging features of personality but rather as dynamic communicative styles, and it should be borne in mind that typologies are not essences but categorizations constructed for certain goals or purposes. The development of such a theory should entail considerations regarding leadership in small social group functioning as well as the guidelines offered by learning and development theories of pedagogical interventions that optimize students' potential. Sociocultural approaches to learning and development in a Vygotskian framework should be the cornerstone of a theory of that kind (see especially John-Steiner & Mahn, 1996). Namely, statements that during the process of socialization the cultural achievements of earlier generations are internalized (transferred to childs' mental level) through the interactions between the child and individuals who are more capable of child, if the interactions are in the zone of proximal development, provide useful information not only for teacher personality, but also how to organize the educational process, and how to assemble educational programmes using representative samples of cultural knowledge. An effective teacher should have interventions aligned to the level of childs' functioning, and be aware of individual differences. He or she also must be able to externalize its own thinking ability and to insist on dialogue between teacher and student and the students themselves. These personality characteristics are included in the theories and models of personality that were mentioned in this review. A theory that would include all these matters might develop over time into the psychology of teachers within educational psychology and into the psychology of the pedagogical profession within school psychology.

Disclosure statement
No potential conflict of interest was reported by the author.

Notes on contributor
Lajos Göncz is professor of Psychology at the University of Novi Sad, Serbia, and a full member of the Academy of Sciences and Arts of Vojvodina. His main interest is in minority studies, particularly in psychological consequences of multilingualism and multiculturalism, and in pre-service and inservice teacher personality and teacher education.