EMOTIONAL INTELLIGENCE MODEL FOR DIRECTORS OF RESEARCH CENTERS IN MEXICO

The objective of this article is to show the results of IE-Martruj, a model designed for measuring emotional intelligence in directors of research centers (RCs), which was applied to forty-three directors of research centers in Mexico. The validation model and reliability tests are described in detail, as well as the mathematical formulae used. As a general conclusion of this research, we show that: The model studied is an adequate instrument for the task proposed, and that the Emotional intelligence (EI) of directors of RCs is positively associated with: motivation, self-assessment

The validation model and reliability tests are described in detail, as well as the mathematical formulae used.As a general conclusion of this research, we show that: The model studied is an adequate instrument for the task proposed, and that the Emotional intelligence (EI) of directors of RCs is positively associated with: motiva-tion, self-assessment, self-regulation, empathy and social skills.Also, the following hypotheses were tested with positive results: H1 -Motivation, initiative, optimism and involvement are conditions positively correlated with the effectiveness of directors of RCs.H2 -Self knowledge, emotional awareness, precise self evaluation and self awareness are positively correlated with each other, and are essential parts of intrapersonal intelligence.H3 -Self-regulation, self control, adaptability and self awareness are positively correlated with each other, and are also part of intrapersonal intelligence.
H4 -Empathy, comprehension of others, development of subordinates' skills, readiness to serve, capacity to manage diversity, and political coherence exhibited by an RC director are positively correlated with each other, and are part of interpersonal intelligence.
H5 -Social skills exhibited by directors, that are also part of interpersonal intelligence, allow a director to exert a greater influence on the working group, facilitating communication, conflict management, leadership, collaboration, cooperation and development of team skills.

KEY WORDS Emotional intelligence, Interpersonal intelligence, Intrapersonal intelligence, Models of intelligence, Assessment of intelligence, Research centers
Clasificación JEL: M12

INTRODUCTION
The concept of Emotional Intelligence (EI) was proposed by Salovey and Mayer in 1990, and has gained considerable currency.It was Goleman (1995), with his accessible style, who made the concept familiar to a significant number of readers, thus sparking considerable interest in the concept of EI.This successful popularization was the point of departure for the publication of many books on the subject, some of them proposing instruments of uncertain validity.According to Goleman, EI is a set of learnable meta-abilities that enable an individual to achieve better adaptation in personal, social and work settings.
Interest in the study of EI favored the proliferation of models, many of which lacked scientific power and did not include a description of validation procedures.
This article describes the IE-Martruj model, resulting from a study that proposed to measure the correlation between the EI of directors or RCs and the productivity of those centers in Mexico.In the first part, we present the history of EI.Later we describe models that have assessed EI in different settings.By means of the design of a measurement instrument, we propose a model for determining the El of academic leaders and directors of RCs in this country.We describe the statistical underpinnings of the model, and the subsequent range of stereotypes, defining the capacity and limitations of the resulting mathematical model, along with pertinent conclusions.

HISTORY OF EI
The development of the concept of EI over the years is heavily indebted to the work of three great authors: Galton, Thorndyke and Gardner.
In 1870, Galton was one of the first scientists who became interested in the systematic study of the "individual differences" between the mental capacities of people.To achieve this, he proposed a statistical analysis applicable to mental phenomena.One of his first attempts used questionnaires and nontraditional methods.(Roback andKierman, 1990) Cattell (1903) was the first American psychoanalyst to quantify stress by means of his experimental work.In 1905, Binet was commissioned by the French government to develop an instrument to assess intelligence in children.In 1916, the instrument developed by Binet-Simon was modified by Terman Lewis, and the new instrument became known as Stanford-Binet.Here, the concept of intelligence quotient appears for the first time.(Feldman, 1980) In 1888, Thorndyke proposed the law of effect, suggesting for the first time the existence of social intelligence, a precursor concept of emotional intelligence.Apart from social intelligence, the same author described abstract intelligence and mechanical intelligence.(Thorndyke, 1920) In 1930, Wechsler presented the Wechsler Adult Intelligence Scale (WAIS), and in 1949 the Wechsler Intelligence Scale for Children (WISC).Both scales are still in use.(Evoys and Wechsler, 1981) As from 1960, a weakening of behaviorist positions in Psychology is observed, along with the emergence of cognitive procedures within the framework of structuralism.Piaget developed a point of view that differed from the then current ideas on human cognition; according to Piaget, study of human thought should be approached from the viewpoint of an individual that is trying to grasp the sense of the world.(Gardner,1993) In 1978, Vygotsky found that intelligence tests provide an indication of potential development.
As from then, there have been a number of reformulations introducing the concept of rationality as a complement of the satisfaction principle, establishing the "reality principle" and the "logic of feelings".(Ribot, 1925;Freud, 1971) Hull proposed a physiologistic approach and Skinner developed it on an operationalist basis.(Hall, 1951;Skinner, 1974) Another approach to the law of effect from the cognitive viewpoint was proposed by Dollar, Millar and Galperin. (Dollar and Millar, 1977;Galperin, 1979) Later, Piaget and Maslow formulated a dynamic development approach to the principle.(Piaget, 1972;Maslow, 1973) In 1997, Sternberg proposed a conception that is associated to the "adaptation capacity" linked to emotion, memory, creativity, optimism and, to a certain extent, to mental health.
In 1983, Gardner reformulated the whole concept of intelligence through the theory of multiple intelligences, in which he identifies seven types of intelligence, not necessarily inter-dependent: musical-auditory intelligence, kinesthetic-corporal intelligence, visual-spatial intelligence, verbal-linguistic intelligence, logicalmathematical intelligence, interpersonal intelligence and intrapersonal intelligence.
If we examine Gardner's 1995 propositions, we find that two of his types of intelligence correlate closely with Thorndyke's social intelligence: interpersonal and intrapersonal intelligence.Later, in 1997, Gardner includes naturist intelligence, and in 1998 he modified his theory yet again, adding a new type: existential intelligence.
In 1990, P. Salovey and J. Mayer structured their conception of EI building on Gardner's interpersonal and intrapersonal types of intelligence; however, it was Goleman's merit to have extended the concept in 1995 with his book directed at the executive world that introduced the benefits and achievements of the study of EI to the field of administration.(Goleman, 2000) The concept of EI proposes to answer the issue: why are there people who adapt better than others to different situations in everyday life?EI is made up of various meta-abilities: awareness of one's own emotions, capacity for controlling emotions, capacity for self-motivation, recognition of the emotions of others, and control of relationships.
EI as a study field was developed by psychologists; however, there has been much important work carried out on a biological basis, such as that of LeDoux. (1987LeDoux. ( , 2002) ) ESTUDIOS GERENCIALES Vol. 24No. 106 • Enero -Marzo de 2008

EMOTIONAL INTELLIGENCE MODELS ASSESSMENT
As we already stated, the popularization of EI led to the creation of many models, of which even the most carefully structured suffered from limitations.
We therefore set out to analyze these models.
Both in the literature and on the consultant market there are a variety of tests that assess individual differences in the various components of EI.If we select only those that present empirical evidence obtained by formal methods and validated models, there are still fifty-three.The application of an inclusion criterion that the model should be listed in Art and Humanities Citation Index left us with fourteen.Once all models had been limited to the administration field, there were five left, that we proceeded to examine in detail.This selection procedure is shown in Table 1.

EMOTIONAL INTELLIGENCE MODELS IN THE ADMINISTRATION AREA
As mentioned above, the concept of EI has acquired many adepts in the field of administration, and is usually interpreted as the association of behaviors and emotions that are exhibited in the performance of an executive.The concept has been appropriated with enthusiasm, as it is considered vital in the business milieu to establish when an "academically intelligent" university graduate may not possess emotional intelligence or vice versa.Likewise -and notwithstanding what the compensation theory says-some individuals blend in themselves both types of intelligence (logical-mathematical and emotional).
Recent studies in the entrepreneurial field have made clear that the professionals who require a greater degree of EI are top managers, (Enebral, 2003) but it is also necessary for psychiatrists, engineers, teachers and social workers.On the other hand, programmers, lab technicians or accountants are less dependent on EI, although it is always useful.
EI models have been highly varied, including very different conceptions and abilities.(Bar-On, 1997;Cooper and Sawaf, 1997;Elias et al, 1999;Goleman, 1995Goleman, , 1998;;Guttman, 1997;Martin and Boeck, 1997;Shapiro, 1997) Of course, all the self-help books, the media, and the huge number of websites on EI that appeared at the high point of the emotional fashion, have hardly con-

ESTUDIOS GERENCIALES
tributed to give the term a scientific image.EI must establish clearer objective measures, like the perception of emotions (analogous to identification of colours, sounds and faces).So far, however, the empirical evidence is weak, as most of it was obtained using consensus methods.
We found five EI models applicable to management, with two basic types: mixed and based on abilities.Among the mixed models, those of Cooper and Sawaf, and Goleman's are especially noteworthy.Among models based on abilities, that of Salovey and Mayer is particularly worth mentioning.

Models based on abilities
Models based on abilities, as a rule, are centred on the emotional context of information and the study of abilities involved in processing it.
The first model evolved by Salovey et al, known as the Trait Meta-Mood Scale (TMMS) permits an assessment of EI, analyzing the more stable characteristics of an individual's awareness of his/her emotions and his/her capacity to control them.TMMS takes stock of the beliefs individuals entertain concerning their own emotions and their capacity to control them.Similarly, the scale evaluates an individual's beliefs referring to his/her attention capacity, clarity and ability to repair his/her emotional states.It is made up of 48 items in its "long" version, with three subscales: awareness of emotions, 21 items; clarity regarding emotions, 15 items; and emotional regulation, 12 items.(Salovey and Mayer, 1990).This test has indices of internal consistency and convergent validity.It has shown predictive capacity regarding emotional adjustment, and the predisposition of individuals to adapt successfully to stressful experiences; however, it must be pointed out that this instrument was designed for the educational field, where study subjects were upper level students.
Another model created by Mayer and Salovey (1995) has many points in common with the previously described instrument.It is really a revised version of the original model.It assesses EI from the point of view of a series of emotional and adaptive abilities, related conceptually according to the following criteria: a) Evaluation and expression of emotions; b) regulation of emotions; c) utilization of emotions in an adaptive way.This version is also called Trait Meta-Mood Scale (abbreviated TMMS with a subindex linked to the number of items contained in it) and constitutes an instrument to measure self-information on El, elaborated on the base of 48 items arranged in three subscales that assess three fundamental dimensions of intrapersonal intelligence: emotions, 24 items; clarity concerning emotions, 12 items; and emotional reparation, 12 items.This instrument has shown adequate indices of internal consistency and acceptable convergent validity.
The authors presented a further adaptation of their model in 1997, where they enumerate, in ascending order, the different emotional abilities that constitute the concept, from the more basic psychological processes (emotional perception) to the more complex (regulation of emotional states).Vol. 24No. 106 • Enero -Marzo de 2008   The questionnaire is organized as a trait scale that evaluates metaawareness of emotional states by means of 48 items.In other words, it rates the abilities by which we can be aware of our own emotions, as well as our capacity to regulate them.

ESTUDIOS GERENCIALES
The model built by Extremera and Fernández-Berrocal (2001) evaluates three variables: perception, comprehension and regulation.It has been used empirically in a number of studies with upper level students.The instrument has also been validated with different populations, proving to be useful both in schools and clinical settings.
As shown in Table l, TMMS 24 is based on the TraÍt Meta-Mood Scale (TMMS) by Salovey, Mayer and their group.This scale contains three key dimensions of EI, with eight items for each of them: emotional perception, comprehension of emotions, end emotional regulation.

Mixed models
Mixed models combine personality traits such as optimism and the capacity for self-motivation with abilities for regulating emotions.These are established according to the context under study.
The EQ-map by Cooper and Sawaf, made up of 21 scales, is considered a 360º degree instrument (also known as a "wide spectrum" instrument, covering general personality traits).Although its reliability is a matter of debate, it is frequently used in the middle level administrative milieu in the United States and Canada.The model permits the exploration of EI starting from personal aptitudes and vulnerabilities vis á vis performance, thus identifying individual and interpersonal patterns for success.(Cooper and Sawaf, 1997) The Goleman model presents ten situations with four alternatives.It is focused toward the areas of publicity, marketing, administrative services and corporate image service.This model offers an assessment of potentials or strengths and their corresponding limits, from the point of view of information on performance.This tool covers the whole range of emotional capacities that are generally present in professional development, by means of providing a general factor of EI. (Goleman, 1995) Another model, by Bar-On (1997) defines the following traits: intrapersonal abilities, interpersonal abilities, adaptability, stress management, and general state of mind.These traits are subdivided into fifteen major components.Due to its subcomponents -such as acceptance of reality, stress management and impulse control, among others-it is classified as a mixed model.However, as its authors have stated, it is an inventory that takes into account a wide range of emotional and social abilities.This instrument contains 133 items on five scales and 15 subscales.In its validation process, this model uses four different tests that could, each one used separately, have provided adequate validation.
Table 2 shows models, authors, type of model, definition and name of model.

METHOD AND MODEL DESIGN
Analysis of existing models persuaded us that they had been designed for environments and study subjects

Field of analysis
Authors Definition Abilities

Name of model
Education area Mayer and Salovey (1997).
El is a combination of abilities that accounts for individual differences in the way people perceive and understand their emotions.
More formally, EI is the ability to perceive, assess and express emotions precisely, the ability to obtain access to and/or generate sentiments that facilitate thought and understanding of emotions and to reason emotionally.(Mayer and Salovey, 1997;p. 10) • Perception, assessment and expression of emotions.El includes self-control, enthusiasm, persistence and the capacity to motivate oneself.There is an outmoded word that encompasses all the range of abilities that make up El: character.(Goleman, 1995;p. 28) • Awareness of one's own emotions

Table 2. Emotional intelligence models
Source: Prepared by the author with material from researchers mentioned above that did not coincide with the study subjects of our research.It must be stressed that RC directors must carry out very specific functions in their day-to-day activities.These functions can be grouped as teaching, research and service (meaning, chiefly, administration of research projects).
The study was carried out in fortythree research centers supported by CONACYT (National Council for Science and Technology) in Mexico.It involved all the centers, meaning it covered the whole Mexican Republic.Due to the small size of the sample, a census was carried out to determine its significance.Vol. 24No. 106 • Enero -Marzo de 2008   Even though we attempted to used one of de Emotional Intelligence models reported in literature, this was not possible for the following reasons: 1) The environment of the RC directors necessarily required the use of a mixed model.2) The instrument had to meet certain specifications, according to the subjects under study.Considering this challenge, the possible instruments could only be Bar-On's EQ I, Goleman's ECI, and Oriolo and Cooper's EQ map.However, all these were rejected.2a) Bar-On's EQ-I was discarded because it was designed for use with middle management.2b) Oriolo and Cooper's EQ map wasn't used because it is focused on the managerial area in services and marketing.2c) Goleman's ECI was also unsuitable because it is an instrument designed for areas such as services, marketing and publicity that have little in common with RCs.

ESTUDIOS GERENCIALES
The subject of our study was defined as a director of an RC, whose main functions are: teaching, research and service.Due to this many-faceted function, the models mentioned did not cover the multifunctional nature of the subject's work.
In Table 3, we show the characteristics of the existing models in the education and administration areas.It also shows the variables and values for validation and reliability, and subjects of study, including, as a final item, the model designed for this study (IE-Martruj).

INSTRUMENT RELIABILITY
Reliability was tested in three steps: validation of the instrument by experts; reduction of the instrument (statistics), and validation and statistical reliability tests.
A group of national and international experts was convened, according to the following profile: psychologists (preferably social psychologists), directors of educational institutions, and methodologists, to ensure that the instrument possessed validity of content.An instrument made up of 266 items was given to this group for validation.After their comments, the instrument was reduced to 176 items.

REDUCTION CRITERIA
One of the more common problems observed in instruments that measure EI is their low discriminating capacity.To overcome this shortcoming, a reproducibility coefficient for scaleograms and the Cronbach coefficient were applied.
As a result of statistical testing, we built a new instrument that retained 48 of the original items, plus 18 new ones, making a new 66 item questionnaire.This new instrument was applied to the same pilot group after four months (experts suggest a minimum interval of three) to obtain its validation and test its reliability.

McNemar's test
Tables 4 and 5 show the results.

• This model obtained better factors of validation and reliability IE-MARTRUJ has its corresponding mathematical model. The model is supported by an extensive statistical base.
The model has its own software that allows it to obtain results in seconds.

Study subjects and areas
Mayer and Salovey (1997) •Perception, evaluation and expression of emotions.
• Assimilation of emotions in thought.
• Understanding and analysis of emotions.
• Reflective regulation of emotions.

MATHEMATICAL MODEL (EI-MARTRUJ)
In the IE-Martruj model, intra and interpersonal EI are defined as dependent variables, while motivation, self-knowledge, self-regulation, empathy and social abilities were defined as independent variables.
The final EI mathematical model was worked out considering the ortogonality of the variables, as there was no similar instrument to follow.
Thus the model responds individually to the calculation of EI, interpersonal intelligence, intrapersonal intelligence, and also allows the calculation of each dimension included in the model; that is, motivation, self-knowledge, self-regulation, social abilities and empathy.The respective equations follow (See equations l, 2 and 3).
Model for EI in general: Model for interpersonal intelligence: Equation 3In which: Intra = intrapersonal intelligence; Inter or interpersonal intelligence; SK = Self knowledge, Mo = Motivation; SR = Self regulation; SA = Social abilities; Em = Empathy.
Following we show the general equation for each variable (see equation 4).Mathematical model is shown in equation 5. Var; = Value of the j th variable P J = Weighting factor of the j th variable Thus, the model can be used in general or for particular traits.
The proposed IE-Martruj model can be used to establish the independent value of the dimension used in the research, such as self-regulation (SR), self-knowledge (SK), motivation (Mo), social abilities (SA) and empathy (Em).To achieve this we have equation 6.

General model for dimensions
Where:

HYPOTHESES TESTS
The general hypothesis and the working hypotheses listed in the abstract served as guidelines to carry out this study.These hypotheses were confirmed by the instrument we ESTUDIOS GERENCIALES Vol. 24No. 106 • Enero -Marzo de 2008   designed, validated and tested for reliability, as can be observed in the following table.
High correlations were found to support formulated hypotheses.

CONCLUSIONS
The emotional intelligence between director researches centres are bad explained by traditional models of emotional intelligence for this reason has been proposed a specific model which in a contribution the field.
The motivation, self knowledge, empathy, and social abilities are the explanation variables to describe more precisely the emotional intelligence of director's research centres.
The intrapersonal intelligence is described how the capacity to manage the others emotions and the interpersonal is the capacity to manage the own emotions the emotional intelligence is consequence to sum of the inter personal and intrapersonal intelligences.
This research has been shown that the emotional intelligence for manage a research centers are different to other organizations.This research open the possibility to analyze the emotional intelligence in another organization, is a valid speculation expected that the explanation variable for ONG, universities requires different kinds of emotional intelligence.
Is necessary continue this research in another organizations such as universities, and international agencies to prove the explanation capacity of this model.

Is valid the speculation that emotional intelligence is a contingent variable whit organizational environment
The study of emotional intelligence is complex and requires simulation models that is not longer exist the actual models are all for diagnosis.
The universe of study is restricted to México so is necessary still going with a transnational study in order to probe if the national culture has a determinant influence in emotional intelligence.
The assumption of general intelligence is not guarantee to has emotional intelligence.
A great surprise in research directors centers is realized that not always the possession of PhD is associate with emotional intelligence.
of the EI indicator or index Any of the dimensions: SR, SK, Mo, SA or Em (1-5) Vi =Value of response to question bi = Sense of question I: 1 for posi-

Table 1 .
Our documentary search yielded Results of documentary search two hundred and forty models that can be divided into two basic types: ability models and mixed models.

Table 3 .
Characteristics of modelsEmotional intelligence model for directors of research centers in Mexico ESTUDIOS GERENCIALESVol. 24No. 106 • Enero -Marzo de 2008

Table 4 .
Reliability of the measurement instruments

Table 5 .
Reliability of instruments according to variable (48 items) 1 Variance inflation factor * Minus item 28 for reasons of balance

Table 6 .
Response to working hypotheses Emotional intelligence model for directors of research centers in Mexico

Strengths and limitations of IE-Martruj •
It is unique in its type • It has high reliability and validity • It has its own operating software • Its design method was tailor made for it, and recommendations from a wide range of experts were taken into account for its development.•Theapplication time is comparatively minimal (ten minutes) • The instrument offers: validity of content (validated by experts); predictive validity (based on correlations); construction validity (base on Cronbach coefficient; Test-Retest, etc).•Themodel has a rigorous statistical backing • The model is an ad hoc instrument for the study subjects • The model has an index for each variable • The model has an index for EIESTUDIOS GERENCIALESVol. 24No. 106 • Enero -Marzo de 2008