THE BEHAVIOURAL AND PERSONALITY CORRELATES OF TRANSACTIONAL AND TRANSFORMATIONAL LEADERSHIP

The article argues that there is a statistically significant positive relationship between leadership style and creativity. It was found that there is a higher statistically significant positive relation between transformational leadership style and creativity than between creativity and the other leadership styles. Creativity (fluency of thought and originality) can partially be accounted for by the Multifactor Leadership Questionnaire. To demonstrate this in the study, the researcher uses, firstly the Multifactor Leadership Questionnaire to identify leadership style, and secondly three measuring instruments that measure creativity. The argument is substantiated by the results of a statistical analysis of leadership style and creativity. Opsomming Hierdie artikel argumenteer dat daar ’n statisities-beduidende positiewe verband bestaan tussen leierskapstyl en kreatiwiteit. Daar is bevind dat daar ’n hoer statisties-beduidende positiewe verband tussen transformasionele leierskapstyl en kreatiwiteit is as tussen kreatiwiteit en die ander leierskapstyle. Kreatiwiteit (vlotheid van denke en oorspronklikheid) kan gedeeltelik verklaar word deur die Veelfaktorleierskapvraelys (MLQ). Om dit te illustreer, gebruik die navorsers in hierdie studie die Veelfaktorleierskapvraelys ten einde leierskapstyl te identifiseer en tweedens drie meetinstrumente wat kreatiwiteit meet. Die argument word gesubstansieer deur die resultate van ’n statistiese ontleding van leierskapstyle en kreatiwiteit.


Leadership traits
According to Ackermann, Schepers, Lessing, and Dannhauser (2000) research has made it clear that successful leaders are not like other people.The evidence indicates that there are certain core traits that significantly contribute to a business leader's success.While none of the traits reviewed by Stogdill (1948Stogdill ( , 1974) ) were found in all studies to be associated with leadership, the consistency with which some traits were found to be associated with leadership and the magnitudes of these associations are impressive.For example, the traits that most consistently showed high correlations with leadership were intelligence (Mann, 1959); dominance (Dyson, Fleitas and Scioli, 1972;Rychlak, 1963); self-esteem (Bass, 1957); task ability (Marak, 1964, Bass, 1961); and sociability (Kaess, Witryol and Nolan, 1961).Kreitner and Kinicki (2001, p. 555) states that we can no longer afford to ignore the implications of leadership traits because traits play a central role in how we perceive leaders.Kirkpatrick and Locke (1991) argue that there is less clear evidence for traits such as creativity, originality and flexibility.These authors believe that the key leadership traits help the leader to acquire the necessary skills; formulate an organisational vision, and an effective plan for pursuing it; and take the necessary steps to implement the vision.
If a positive correlation can be found between creativity and leadership style then the question arises as to whether some leaders are more creative than others.

Transformational and Transactional Leadership
It appears from the literature that Transformational Leadership as conceptualised by Burns (1978), is currently a major focus area.According to Ackerman et al., (2000, p.58) Transformational Leadership, probably serves as the most appropriate style in managing the changes presently taking place in South African organisations.
Transactional Leadership style focuses on interpersonal transactions between managers and employees.Leaders are seen as engaging in behaviours that maintain a quality interaction between themselves and followers (Kreitner & Kinicki, 2001, p. 567).According to Gibson (1997, p.313) the leader helps the followers identif y what must be done to accomplish the desired results, i.e. better quality output, more sales or services and reduced cost of production.In helping the followers identif y what must be done, the leader takes into consideration each person's self-concept and esteem needs.
Transformational Leadership is viewed as a special case of Transactional Leadership, and unlike the external rewards promised by Transactional Leaders, the employee's reward is internal (Gibson, 1997).

Transformational Leadership
According to Bass and Avolio (1985) Transformational Leaders motivate others to do more than they originally intended and often even more that they thought possible.They set more challenging expectations and t ypically achieve higher performance.

THE BEHAVIOURAL AND PERSONALITY CORRELATES OF TRANSACTIONAL AND TRANSFORMATIONAL LEADERSHIP Creativity
Although there is seemingly less clear evidence of creativity as a trait that leaders possess, an abundance of research have been done on creativity as such (Getzels, 1964;Greeno, 1978;Sternberg, 1999).

Divergent thinking and ideational fluency of thought
According to Guilford (1957) it is in divergent thinking that we find the most obvious indications of creativity.Divergent thinking ability require individuals to produce several responses to a specific prompt.The more responses the individual produces, the higher the ideational fluency ability of the person (Guilford,1962).

Associational fluency of thought
Creative thinking has been viewed by Mednick (1962) as the recombining of associative elements that either meet specified requirements or are in some way useful.He argues that solutions may occur through mediation (combination via common elements).Mednick (1967) argues that the greater the number of associations that an individual has with the requisite elements of a problem, the greater the probability of his reaching a creative solution.To assess creativity, Mednick (1962) devised the Remote Associates Test (RAT).

Originality of thought
Originality of thought is defined by Cougar (1995, p.370) as 'the capacity to produce unusual ideas, solve problems in unusual ways, and use things and situations in an unusual manner'.Originality consists of uniqueness and nonconformity with thought and action.Barron (1955) argues that original must be defined relative to usual, and the degree of originality must be specified in terms of incidence of occurrence.Thus the first criterion of an original response is that it should have a certain specifiable uncommonness.Sternberg (1999, p.450) states that there appears to be consensus that the two defining characteristics of creativity are originality and usefulness.
Creativity and Problem Solving Guilford (1957) argues that creative steps are necessary in solving new problems.A problem only exists because the situation presents the necessity for new productions of some kind.Factors are abstractions of components from total activities.Some of the components are recognised as being more creative than others, for example, qualities of originality and fluency.Creativity is therefore not a gift possessed by rare and exceptional individuals (though some components are undeniably more creative), but all individuals are creative to some extent.
This study has attempted to establish whether a correlation exists between creativity and behavioural leadership style.

METHOD
This section describes the method that was employed in the study, including a description of the target population, raters and measuring instruments.

Participants/respondents
All students currently enrolled for their MBA and M.Com.degrees comprised the population of interest.The unit of study (Cooper, 1998, p.215) consisted of students who are busy with their master's degrees in three different universities in the Gauteng area in South Africa.The sampling method used is non-probability sampling as defined by Cooper (1998, p.237).For the purpose of this research judgement sampling was used as a type of purposive sampling.The researcher selected sample members who conformed to some attributes.The most common attribute of masters' students is their work experience on managerial level.The researcher believes that although the use of students may imply homogeneous groups, this sample is more heterogeneous due to the fact that the respondents have managerial experience as a common attribute.The sample size (N) was 196.

Measuring instruments
Transformational Leadership was measured by the Multifactor Leadership Questionnaire (MLQ) developed by Bass and Avolio (1985).The MLQ measures a broad range of leadership style from Laissez-fair to Idealised influence.The MLQ consists of 45 items loading on four factors that represent the Full Range Model.Ideational fluency and originality of thought were measured by two divergent thinking ability tests similar to those developed by Chand and Runco (1993).
The first test assessed ideational fluency and originality of thought in respect of a concrete concept, eg.
List the uses of a paper clip, and the second test assessed ideational fluency and originality of thought in respect of an abstract concept, eg.List the possible solutions for crime in SA.
Associational fluency of thought was measured by the Remote Associates Test as developed by Thompson (1993).
Complete records were obtained in respect of 178 participants.

Procedure
The research procedure comprised the following steps: The researcher administered the questionnaires in person.
Respondents were told that confidentiality would be maintained.
Respondents were asked to complete the Multifactor Leadership Questionnaire answer sheet.It was explained to the respondents that the questionnaire consists of statements about typical leadership behaviour and they were asked to indicate how often they behave in a certain way.The items required that the respondents should indicate how strongly they identify with the behaviour.Because the questionnaire was self-explanatory the respondents did not have to provide their names.This took approximately 20 minutes.
Respondents then completed the Remote Associates Test.A number of practice examples were given before beginning the test.The amount of time allocated for the Remote Associates Test was 20 minutes.
Respondents then completed the two divergent thinking ability tests.Explicit instructions were given in order to ensure that respondents understood that the alternatives generated should be different from one another.The respondents were asked to be as creative as possible when generating alternatives.The amount of time allocated was three minutes for each test.
All four sets of tests were stapled together for each respondent.
The researcher and two independent raters then evaluated the divergent thinking ability tests in order to establish originality of thought.A five-point scale for individual components was used where 1= very common and 5= very original.
The researcher evaluated the Remote Associates Test.
All four tests were sent to the Statistical Consultation Service at the Rand Afrikaans University for processing.

RESULTS
The factor analysis of the MLQ Two different types of instruments were used.Firstly the MLQ was used to identify leadership style and secondly, three creativity measuring instruments, measuring fluency and originality of thought were used.The 45 items of the MLQ were subjected to a principal factor analysis.The analysis was beased on 196 complete records.As a first step, the items were intercorrelated and the eigenvalues of the unreduced intercorrelation matrix were computed.As fifteen of the eigenvalues were greater than unity, fifteen factors were extracted and rotated to simple structure by means of a Varimax rotation.
To overcome the effect of differential skewness of items, subscores in respect of each factor were computed by adding the items with high loadings on a factor, together.The subscores were then intercorrelated and subjected to a principal factor analysis.The eigenvalues of the unreduced intercorrelation matrix are presented in Table 1.Four factors were extracted, as four of the eigenvalues were greater than unity.Factor 1 had substantial loadings on 16 items and Factor 4 on 19 items yielding Cronbach alphas of 0,767 and 0,885 respectively.Factors 2 and 3 were poorly determined, having loadings on four and six items respectively.The reliability of the corresponding scales were 0,55 for Factor 2 and 0,479 for Factor 3. Accordingly a three-factor solution was tried.Factors 3 and 4 combined into a single factor with a reliability of 0,84.No items were lost.The reliability of Factor 4 fell, however, from 0,88 to 0,847.The obtained factor matrix was rotated to simple structure by means of a Direct Oblimin Rotation, as given in Table 2. Table 2 shows that three factors are reasonably well determined, with substantial loadings on all three.However, from an inspection of the communalities it is apparent that several of the subscores share only a small proportion of their variance with the other subscores.In order to identify the obtained factors, the items associated with each factoe were grouped into categories as given by Bass (1999).Bass's classification (1999) is presented in Table 3.  Bass (1999) Table 3 shows that Factor 1 is strongly representative of items categorised as Transformational Leadership and Leadership Outcome.There are only two items categorised as Transactional Leadership.Factor 1 can therefore be identified as Transformational Leadership/Leadership outcome.
Factor 2 is representative of items categorised as Transactional Leadership and Non-Transactional Leadership (Laissez Faire).There is also one item with a negative loading, categorised as Transformational Leadership.The items listed as Transactional Leadership, all deal with failure to take action when required, and keeping track of mistakes, irregularities, failures and complaints.Factor 2 was identified as Non-Transactional Leadership.
Factor 3 is representative of items categorised as Transformational Leadership.There are three items categorised as Transactional Leadership.Factor 3 is therefore identified as Transformational/Transactional Leadership.
For the purpose of this research the items in respect of each factor were compared to the items as given in Table 3.The three factors of this research differ from those of Bass (1999) in the sense that the factor analysis revealed some differences in the items grouped under the different factors.Although this research also postulated three factors, the items that make up the three factors differ in minor ways: Bass (1999) argues that Factor 1 represents Transformational Leadership/Leadership outcome.The items are similar to Factor 3 of this research.
According to Bass's research (1999), Factor 2 represents the Laissez Faire leadership style (a finding confirmed by this research).However, the fact that only five items remained in Factor 2 (after the three factor solution was done) decreased the reliability coefficients of both Factors 2 and 3.These items focus more on the leaders' ability to solve problems and rectify mistakes.When problems are solved and mistakes rectified, the leader achieves an outcome.These items therefore do not focus on any specific leadership style.Bass (1999)  Table 4 shows that Factors 1 and 3 are positively correlated.Three scales were formed corresponding to the three factors obtained.
To determine the reliability of these scales, they were subjected to item analysis, using the NP 50 program.The item statistics in respect of the first scale are presented in Table 5.Table 5 shows that the obtained reliability in respect of Scale 1, is 0,752.The item total-correlations range from 0,288 to 0,615.The scale is internally consistent.No items were rejected.The reliability coefficient is sufficient for the purpose of this research.The item statistics in respect of the second scale are presented in Table 6.
Table 6 shows that the obtained reliability, in respect of Scale 2, is 0,494.The item-total correlations range from 0,482 to 0,628.None of the items were rejected, but there are too few items.The item statistics of the third scale are presented in Table 7.  Table 7 shows that the obtained reliability of Scale 3 is 0,837.The item-total correlations range form 0,219 to 0,649.None of the items were rejected.The scale consists of 23 items.

Factor analysis of creativity measures
The three creativity ability tests measure fluency and originality of thought.The two divergent thinking ability tests measure fluency and originality of thought.The fluency of thought scores (Scores 2 and 4) were calculated by adding the number of responses, and these responses were then subjected to an evaluation of originality (Scores 1 and 3) by independent raters.Four scores were derived from these fluency tests.
Secondly, the Remote Associate test (RAT) measures associational fluency of thought.This ability test is divided into two sections (Scores 5 and 6) and comprises 15 items per section.Scores are calculated by adding the number of correct responses for each item.Two scores were derived from this test.
The six derived scores of the creativity tests were subjected to a factor analysis.The analysis was based on a sample of 178 participants.As a first step, the six scores were intercorrelated and the eigenvalues of the intercorrelation matrix were determined.The obtained eigenvalues are given in Table 8.From Table 8 it is clear that only one eigenvalue is greater than unity.Accordingly only one factor was extracted and is given in Table 9.From Table 9 it is evident that all six scores have substantial loadings on the same factor.This factor represents fluency and originality of thought.Scores 1 and 3 are comparable (originality of thought).Scores 2 and 4 are comparable (fluency of thought).Scores 5 and 6 are coded.The communalities of the six scores vary from 0,227 to 0,509.Scores 4, 5 and 6 have the lowest communalities, indicating that they have less in common with the other measures, but are not necessarily poor measures of creativity.

TABLE 9 FACTOR MATRIX OF THE CREATIVITY MEASURES
The means and standard deviations in respect of the creativity measures are presented in Table 10.
Table 10 shows that the obtained reliability of the Creativity Scale is 0,735.None of the scores were rejected.The mean of the Remote Associates Test 1 was lower than that of the other measuring instruments.Fewer correct answers were found in respect of this measuring instrument.The mean of the Crime Rater total (originality) is slightly less than that of the other measuring instruments.
Table 11 shows the Pearson product moment (Cooper,1998) correlations between the three scales of the MLQ and the seven scores of creativity.Some of the variables are positively related and others are negatively related.Two stepwise regression analyses were done.In the first analysis, Scale 1 (Transformational/Transactional) of the MLQ served as the independent variable and in the second analysis Scales 1 (Transformational/Transactional) and 3 (Transformational/ Leader outcome) of the MLQ were used.Table 12 provides a model summary of the regression analysis of the independent variables on the dependent variable.
The regression analyses indicate that only two predictor variables Scales 1 and 2 hold promise of the MLQ.A multiple correlation of 0,263 was obtained in terms of the first model, which indicates that 6,4% of the variance of fluency/originality can be accounted for by the first predictor variable, i.e.Scale 1.A multiple correlation of 0,332 was obtained in terms of the second model, indicating that 10% of the variance in fluency/originality can be accounted for by Scales 1 and 3.The combination of Scales 1 and 3 therefore accounts for a greater proportion of the variance of creativity.Table 13 gives the analysis of variance.
From the analysis of variance it is clear that the linear regression accounts for a statistically significant proportion of the total variance.
All the regression coefficients are statistically significant.

DISCUSSION
The researcher concludes that there is a statistically significant positive relationship between leadership style and creativity.
There is a higher statistically significant positive relation between Transformational Leadership style and creativity than between creativity and the other leadership styles.Approximately 10% of the variance of creativity (fluency and originality of thought) can be accounted for by the MLQ.
In the light of the above-mentioned finding the MLQ can be recommended as a test in the selection process.It would also be worthwhile pursuing means of upgrading the MLQ.For the purpose of this investigation respondents were asked to rate themselves in respect of the MLQ.It is recommended that if similar research is to be done in future, the superiors of the respondents should also complete the instrument and rate the respondents from their perspective (360°approach).
Despite the positive results obtained with the battery of creativity measures, certain reservations concerning them must be mentioned.Two of the measuring instruments were pertinent to divergent thinking ability.The main drawback from a psychometric point of view, is that the measurement of divergent thinking necessitates an element of subjectivity in the scoring of the tests.Even where the actual scoring procedure is objective, a subjective assumption is made to the effect that the score so obtained provides a measure of whatever the test is supposed to measure.Even more questionable is the assumption that originality can be gauged by weighting responses in terms of the rarity of their occurrence in the population.The possibility of rater bias must also be recognised.The correlations obtained by intercorrelating the creativity measures are generally so high as to suggest that the measures overlap to a considerable extent.This was also borne out by the general factor of creativity obtained in the factor analysis of the measures.Since the measuring instruments were not designed for factor analysis per se, the finding of a general factor cannot be cited as proof that a general creativity factor exists, but it does suggest that the different instruments used do not measure separate factors beyond fluency of thought and originality.Though there is evidence that creativity is manifested in leadership style, accurate judgements of creative ability can only be made with reference to an individual's actual output.
Ideally the dependent variable should be validated against real life work performance.This can only be achieved if use can be made of experts, who have the knowledge and skills required.But given such an opportunity, one must recognise the possibility of rater bias.The Remote Associates Test has shown promise as a predictor of associational fluency, an element of creativity.The biggest problem with this measuring instrument is the difficulty of establishing an objective and reliable scoring system for the South African context.Problemsolving is a concept that relates to both leadership style and creativity.A problem -solving orientation was identified by the MLQ (Factor 3), but the ability to solve problems as such was not measured in this research as the instruments focussed on fluency of thought and originality.It is therefore recommended that a test that measures problem -solving ability should be included in future research.

TABLE 2 ROTATED
FACTOR MATRIX OF THE MLQ (DIRECT OBLIMIN ROTATION) Note: Factor 3 has been reflected Bass (1999)actors of this research produced essentially the same categories of information as those produced byBass (1999)on leadership i.e.Transformational/ Transactional, Laissez faire and Transformational/Leadership outcome.The intercorrelations between the factors are presented in Table 4.
states that his Factor 3 represents the Transformational/Transactional Leadership style.The items are similar to Factor 1 of this research.It is interesting to note that 65% of the items under Factor 1 represent the Transformational Leadership style in its pure form.The researcher came to the conclusion that