ENTREPRENEURiAL PERSoNALiTY TRAiTS ANd SMEs PRoFiTABiLiTY iN TRANSiTioN ECoNoMY

Creativity and innovation have been identified in the literature as the main drivers of small and medium size enterprises’ development. The aim of this study was to determine whether these factors also affect the profitability of the SME sector in Serbia. The conceptual model was set and six hypotheses were proposed. The survey was conducted in South and South-Eastern Serbia, on a sample of 717 small and medium enterprises. Confirmatory Factor Analysis (CFA) and Path Analysis were used to test conceptual model. The results showed high reliability of the data and confirmed hypotheses.

competitors, in order to strengthen its market position and increase sale (Keh et al., 2007).
The aim of this study was to determine the mechanisms by which entrepreneurial activities and traits (entrepreneurial creativity, knowledge transfer, entrepreneurial self-efficacy, data collection, teamwork and organization' innovativeness) affect the profitability of SMEs in Serbia.Serbia was chosen for this study due to the fact that it was among the first Balkan countries restructured to a market economy during the 1990s.Since then, Serbia has been going through a transitional period and is currently a candidate for the membership in the EU.During that period a lot of things happened in Serbia which had different impact on the performance of organizations compared to the changes the countries with stable economies were undergoing.According to authors' knowledge not many research were done regarding SMEs activity in Serbia.Statistical Office of the Republic of Serbia conducted survey for period [2004][2005][2006] on innovation activity in SMEs in Serbia and the results have shown that SMEs are faced with a large number of obstacles which adversely affect their activities (Nikolić et al., 2015).In their paper Cvetanović et al. (2014) applied multicriteria analysis in order to investigate the impact of information technologies on SME innovativeness and further on productivity growth and competitiveness.The authors of the study did not find the research that investigate the similar aspects, therefore this paper makes an attempt to identify the key factors which affect SMEs' profitability in Serbia.South and South-East Serbia was chosen as one of the less developed part of Serbia and future research will include other regions in Serbia.The results obtained from this region will be analyzed and compared to the results obtained from other regions, establishing thus universally valid connections that might be important for the development of entrepreneurship in Serbia.
Although in his work Naude (2010) underlined that research in the field of entrepreneurship should be limited only to the developed economies, Engelen et al. (2009) pointed out that similar researches in various cultures would only promote and improve entrepreneurship, because they could highlight links that were valid in relation to the links that applied to individual cultures.Also, some authors (Leskovar-Spacapan & Bastic, 2007) emphasized the transferability of conclusions across countries, and pointed out that the conclusions driven for developed countries could not be generalized and were not necessarily relevant for explaining situations in transition and developing countries.
In Serbia, the transition from centrally planned to a market-oriented economy was a radical change, making it impossible for companies to simultaneously carry out the internal reorganization and to successfully adapt to new environmental conditions.Although transition economies are becoming more similar to Western economies, however, the competitive ability of some organizations still remains limited by the legacy of the previous political and economic system.The way the organizations used to operate is no longer in accordance with modern way of doing business that requires flexibility, strengthening the innovative capacity and entrepreneurial spirit.
Taking all the previously stated into consideration, a conceptual model was developed and six hypotheses were proposed, which explore the factors affecting the profitability of SMEs.(1934) in his theory of "creative destruction" stressed the key role of entrepreneurial creativity in initiating and developing technological innovation and economic changes.Since then, many studies have been examining the phenomenon of entrepreneurial creativity.Maslow (1968) underscored that creativity was an inherent feature of all people, but only a small part of the population used that creativity.Nystrom (1993) said that entrepreneurial creativity was "an invention of the future".In their study, Baron and Tang (2011) emphasized the positive impact of the entrepreneurial creativity on the organizations' performance.Entrepreneurial creativity can be defined as the creation of new and useful ideas in any domain (Amabile et al., 1996).People who are creative in one area are, probably, also creative in other areas, just as people who are highly intelligent have good results in any cognitive tasks (Nunnally, 1978;Silvia et al., 2009).Creative individuals are important in an organization for achieving sustainable and competitive advantages (Shalley, 1995), but also their presence can create a spillover effect and they can act as role models for the rest of employees (Shalley & Gilson, 2004).

Schumpeter
Innovation is a key element of competition and dynamic efficiency of the market (Hall et al., 2009).In the long run, an innovative organization will grow faster, be more efficient and more profitable compared to a non-innovative one.Our expectations are that the entrepreneurial creativity is positively related to organization innovativeness, therefore we suggest the following hypothesis: H1.Entrepreneurial creativity has positive impact on organizations' innovativeness.

Knowledge transfer and organizations' innovativeness
In organizational terms, knowledge transfer is a process in which the experience of one group affects the second group, having in mind that this process involves two or more parties (Hasan et al., 2013), and these parties are organizational units.In organizations where there is good communication between employees, knowledge transfer is something normal and happens indirectly (Sparkes & Miyake, 2000).Organizations themselves are not capable of creating knowledge because knowledge is created by individuals, and it doubles the moment individuals share it with others (Zhang & Ng, 2012).Knowledge transfer is better when there are no intermediaries in the relationship between two organizational units (Hansen, 2002).
Proper knowledge transfer enhances organizations' innovativeness, and thus the organizational performance (Szulanski, 1996).Soo et al. (2007) underlined that each organization would improve its performance if they were able to absorb knowledge, use that knowledge, act creatively by this knowledge and at the end, create new knowledge.Our expectation is that knowledge transfer will increase organizations' innovativeness and thus their profitability, so we suggest the hypothesis: H2.Knowledge transfer has positive impact on organizations' innovativeness.

Entrepreneurial self-efficiency and team work
Entrepreneurial self-efficacy refers to the extent to which an individual believes that he/she can complete tasks and roles of entrepreneurship (Boyd & Vozikis, 1994).Entrepreneurs with the same skills may have weak, normal or outstanding performance, depending on whether their own beliefs about the self-efficacy strengthen or diminish their motivation and attempts to solve the problem (Wood & Bandura, 1989).There are higher chances that individuals with high self-efficacy for a specific task will seek and stay on this task longer than individuals with low self-efficacy (Bandura, 2001).A person may have high self-efficacy in one area, but low self-efficacy in other areas, especially pertaining to certain, very specific tasks and/or skills (Wilson et al., 2007).Entrepreneurs with strong beliefs about their entrepreneurial self-efficacy tend to associate challenging situations with awards such as profit, recognition and psychological fulfillment, because selfefficacy is not related to the past, but to what could be achieved in the future (Hmieleski & Corbett 2008;Yang & Cheng, 2009).
According to Bandura (1977), there are four sources of information that are associated with self-efficacy: current effect, observation of another's behavior, support from others, especially important people, and emotional excitement or anxiety in contact with a certain type of behavior.In homogenous groups people with high level of self-efficacy are encouraged to be even more self-efficient.Our expectations are that the entrepreneurial self-efficacy has a positive impact on teamwork, and thus on better business performance, hence we propose the following hypothesis: H3.Entrepreneurial self-efficacy has a positive impact on teamwork

data collection and team work
Information from the market is defined as data related to the current and potential stakeholders, stemming from various external sources (Moorman, 1995).In their paper, Kawakami et al. (2012) underpinned that the importance of the data collection process was enormous for both large and SMEs.Data collection is a big challenge for SMEs, because of the limited resources at their disposal and less experience in market research.Therefore, most entrepreneurs rely on informal sources of information, such as relatives or existing customers (Narver & Slater, 1990).Training and additional education of employees will enable employees to collect and use information from the market (Kawakami et al., 2012).
Quality decision making requires the integration of large amounts of knowledge dispersed throughout the firm (Grant, 1997).This means that those who have to make decisions first have to find the information within the organization, and then to transfer that knowledge to work teams when necessary (Dougherty, 2001).Internal learning in combination with external (fairs, exhibitions, seminars, training programs) creates the necessary precondition for successful teamwork in the organization, and thus the conditions for quality decisionmaking.We expect that the systematic collection of market information will have a positive effect on teamwork in the organization, and we suggest the following hypothesis: H4. Data collection has positive impact on team work.

organizational innovativeness and profitability
SMEs that have previous experience in innovative activities have higher chances to be successful than non-innovative firms (European Commission, 2010).In many studies, there is evidence suggesting a positive relationship between the organizational innovativeness and performance of organizations, both in production and in service companies (Golovko & Valentini, 2011;Love & Roper, 2015).The same authors emphasized the positive impact of innovations in SMEs to their profitability, the growth of the organizations and increased export.We suggest the following hypothesis: H5.Organizational innovativeness has positive impact on profitability.

Team work and profitability
The basic elements of each organizational unit are people and teams.By using teamwork organizations achieve synergetic effect, because higher scores are achieved working together rather than any individual contribution, and also each individual in a team achieves more than he/she ever could (Tata & Prasad, 2004).Today, teamwork is becoming important increasingly and modern business without it is almost unimaginable.
In almost all studies it was confirmed that teamwork increased employees' satisfaction and loyalty (Stewart & Barrick, 2000) and leads organization closer to the set goals.Using teams is a part of social responsibility, because it enhances communication, trust and stimulates profitability (Levine, 2007).We suggest the following hypothesis: H6. Teamwork has positive impact on organizational profitability.
According to hypotheses H1-H6, the conceptual model has been defined to measure the influence of certain parameters on the organizational profitability (Figure 1).

METHodoLoGY
The questionnaire consists of two parts.The first part comprises 8 control questions (Table 2) and second part contains 25 questions divided into 7 groups, which are related to entrepreneurial creativity,  1).For the assessment of the answers, we used a five-point Likert scale, where 1 means "strongly disagree"; 3, "neutral" and 5, "strongly agree".

Sample and data collection
This research was conducted from January to May 2015, in South and South-Eastern Serbia.The authors personally conducted the survey, trying to explain the meaning of every issue to every respondent, in order to avoid misunderstanding.The survey was anonymous, but either way there were some respondents who did not want to fill out a survey.Survey covered a total of 856 small business owners, 717 questionnaires were properly completed (83.76%).Ratio between sample size (717) and number of questions (25 from questionnaire) is 26.68 and is much larger than the prescribed value 5 (Hair et al., 2006).
The demographic characteristics of the sample are described in Table 2.The results have shown that survey included 504 male and 213 female entrepreneurs.81.3% from total number of respondents belong to the age group between 26-55, which represents the best period for entrepreneurship.According to the size of the company, 64.3% are owners of micro enterprises (less than 10 employees) and regarding the firm age, 75.2% are companies not older than 20 years, which is due to the fact that until 1990s economy in Serbia was centrally planned and SMEs started to develop during the transition period.
66,5% of the total number of respondents said that they had had previous entrepreneurial experience, which means that 33.5% started a new venture.Out of those who said that they had had previous entrepreneurial experience, 86.6% (413/477) stated that in the past they had been doing the same business as they did now.Also, it is notable that 87.9% of organizations are domestically owned and 57.8% are service organizations.

descriptive statistics
Descriptive statistics belongs to the group of statistical methods for research of mass phenomena, which includes collecting, sorting, calculating, displaying and  describing the main characteristics of the statistical series.Descriptive statistics was performed using software package SPSS 18.0.Results of descriptive statistics for all 25 questions (variables), are shown in Table 3.
Table 3 shows some basic parameters of descriptive statistics only: the mean value of the sample as a measure of central tendency and standard deviation.

Factor analysis
First we conducted an exploratory factor analysis to uncover the underlying factor structure and the distinctiveness of latent variables.Table 4 4.
Before evaluating the fit of the conceptual model presented in Figure 1, it is necessary to define a measurement (control) model to verify that the 25 measurement variables written to reflect the seven unobserved constructs-groups, do so in a reliable manner.Important issues of defined model functionality are its validity and reliability.Therefore, Confirmatory Factor Analysis (CFA) was applied on all latent groups of the measurement model at the same time, forcing correlation relationship between 7 defined groups.The obtained values are shown in Table 4.
Although, PCA analysis had already uncovered unidimensionality within 7 groups, CFA with maximum likelihood estimation, confirmed that the overall measurement model fit appeared quite good.The reliability and fit measures of measurement model are also shown in Table 4.
The CFA parameters of unidimensionality and reliability (Cronbach`salfa) of the scales, indicate that all the groups are unidimensional and highly reliable.
Furthermore, CFA was used to assess convergent validity (Živković et al., 2010) and if all factor loadings of indicators on their constructs were significant, convergent validity was attained.The values of factor loading, t-value and p-value (p<0.05),shown in the Table 4, imply that all constructs have strong convergent validity.
A number of goodness-of-fit criteria are available to assess the overall fit of the models.Goodness-of-fit criteria measure the extent to which the actual or observed covariance input matrix corresponds to that predicted from the proposed model (Ho, 2006).Some commonly used measures of absolute fit include: goodness-of-fit index (GFI) and root-mean-square error of approximation (RMSEA).GFI is a nonstatistical measure ranging from 0 (poor fit) to 1 (perfect fit).The higher the value of GFI indicator, the better the match.Good fitting is indicated by a value above 0.90 (Molina et al., 2007).Therefore, this indicator is acceptable in our model (GFI =0.85), its value is just below the threshold, and it is assumed that it would increase with an increase in the number of respondents.Other indicators of correspondence of the model are given in Table 5.
The final aspect to be studied is the proposed model's parsimony.Of the measurements proposed, only the relative chi-square is of use in the confirmatory analysis.This measurement must take values above one and below three or even five to ensure the data is not over fitted (Hair et al., 2006;Molina et al., 2007) and to be truly representative of the data.In our case, the value reached is 2.48 and therefore within the accepted limits.

Correlation matrix
Results of correlation analysis are shown in Table 6.
All the values of correlation coefficients should be above the recommended value of 0.33.In this way it is ensured that there is a positive correlation between the latent variables.

Path model
The conceptual model was tested using software package LISREL 8.8.SEM analysis was chosen over regression analysis, because SEM can analyze all the paths in one analysis (Gefen et al., 2000;Tenenhaus et al., 2005;Živković et al., 2010).
The results of SEM analysis are shown in Figure 2. Regression coefficients are presented above the arrows and they explain the strength of the relationship between dependent and independent variables.It can  be seen that all coefficients have positive value, after that the t-test was performed in order to test statistical significance of the results.The results of the t-test demonstrated values over 1.96 as it had been prescribed (Hair et al., 2006;Ho, 2006), so we can conclude that all six hypothesis are confirmed.
The Squared Multiple Correlations (R 2 ) value represents the percentage of variance in an endogenous construct explained by other constructs connected to it directly.Interpreted as multiple regression results, the R 2 value indicates the amount of variance explained by the model.The overall model explained 58% of the variance in organizational profitability.

diSCUSSioN
Results of descriptive statistics (Table 3) show that the mean value of respondents' answers ranges from 3.0 to 4.1, and standard deviation from 0.81 to 1.22.Mean value for all 25 questions, is 3.67, and mean value for standard deviation is 0.95.Based on these results it can be concluded that entrepreneurs have a positive opinion regarding all researched aspects and that they can meet all essential preconditions of entrepreneurial activities, which are defined in the model.
In all the groups (Q1-Q7) KMO indicators are far above the recommended value of 0.6.Bartlett indicator of sphericity is significant and lower than 0.05, which indicates that there are correlations among the items in the questionnaire, that is, correlation matrix is not one.In this way, the results confirmed that the data set was suitable for factor analysis.
The results of factor analysis (PCA) confirmed unidimensionality in all groups of questions that were posted in the conceptual model, because all the test items grouped in one factor were set with eigenvalue greater than 1.0.Percentage of variability described by each factor is shown in Table 4.The factor loading of variables is in the range from 0.67 to 0.90, which is above the recommended value of 0.6.This suggests that the group of latent variables defined by Q1 to Q7, can be reliably described using research questions defined by Q1_1 to Q7_3.
Based on the results of the CFA analysis (Table 4) it can be seen that the values of Cronbach' alpha coefficient for all groups of questions is over the prescribed value 0.7.Also, Cronbach' alpha coefficient for the entire survey population was calculated (α=0.91) and is above the recommended values.These results suggest that there is an internal consistency and good possibility of modeling results for the survey population.
Table 5  The relative chi-square (χ 2 /d.f.) of the measurement model is 2.30, and 2.48 of the structural model, which fits in the recommended range from 1 to 3. Thus it was confirmed that the initial data were truly representative.
Based on the results of the fit indicator it can be concluded that good fitting of both models (both measuring and structural) was achieved.Furthermore, a general conclusion can be made that all 25 variables (questions) can be a reliable and valid way to describe the formed 7 groups of latent variables (sets of questions) of the conceptual model, which is shown in Figure 1.

CoNCLUSioN
The transition from centrally planned to a market-oriented economy demanded some changes within the organization and in the external environment.The biggest efforts were made to create political and economic conditions in the country that would attract foreign capital but also enable people with capital in the country to launch their own small business.In all of these efforts to adjust the situation in the external environment little attention was given to people within the organization and to certain personality traits carrying with it the spirit of entrepreneurship.Entrepreneurship requires a high level of creativity, innovation and risk taking, everything not specific for Serbian national culture.Entrepreneurship was created in the Western culture and precisely reflects that way of thinking and doing business, which relies on creativity and innovation.
This research is a contribution to the field of entrepreneurship, related in particular to creativity and innovation.Serbia, which was mainly characterized by centrally planned and mono-structural economy until 2000, has a very short history of entrepreneurship.Similarly to all transition countries, entrepreneurship is still in its development in Serbia, due to economic conditions and also because of the national culture, which is not inherently risk-taking that is on the other hand inevitably linked to the concept of entrepreneurship.
The implications that this research could have on the business policy makers, and especially to those to which the research refers, for entrepreneurs, is that creativity and innovation must be encouraged.Maslow (1968) said that creativity was a personality trait that could be developed and nurtured.One of the biggest myths of entrepreneurship is that entrepreneurship cannot be taught.Yes, it can, considerably, we just have to learn to accept the risk that goes with it and every innovation introduced into the business.Although most entrepreneurs and small business owners have very little free time, however, they should find time for creative thinking and a positive attitude to change.
This research examined how owners' personality traits affected organizational profitability.As with any study, there are certain limitations.First of all, entrepreneurs evaluated their creativity and self-efficacy by themselves hence the entire investigation is based on their subjective assessment which must always be taken with a grain of salt.We tried to alleviate their subjectivity regarding the innovation, forcing them to selfassessment in relation to the biggest competitor.Future research could also reduce subjectivity using a different scale, which will exclude self-assessment of entrepreneurs.Another limitation was the fact that the survey was conducted in South and South-Eastern Serbia, which represents a less developed part of Serbia, thus the obtained results cannot be generalized for the entire country, but can be a good basis for further research in other regions, and used for their comparative analysis.

Table 1 .
Questionnaire items

Table 2 .
Demographic characteristics of the sample

Table 3 .
Descriptive statistics of variables

Table 4 .
The results of the EFA and CFA statistics for measurement model

Table 5 .
The values of the indicator fitting for measuring and structural model