Effects of cognitive variables on entrepreneurial intention levels

- 1 - Effects of cognitive variables on entrepreneurial intention levels Francisco Liñán University of Seville Dept. Economía Aplicada I. University of Seville Av. Ramón y Cajal, 1. E41018 - Seville (Spain) Tel.: +34.954554487. Fax: +34.954551636. flina[email protected] Juan C. Rodríguez University of Jaén Dept. Economía Aplicada. University of Jaen Paraje de las Lagunillas, s/n. E23071 - Jaen (Spain) Tel.: +34.953212069. [email protected] José M. Rueda Pablo Olavide University Dept. Economía, Métodos Cuantitativos e H.E. Pablo Olavide University Carretera de Utrera, km. 1, 41013, Sevilla Tel.: +34.954349354. E-mail: [email protected] Jesús S. Martínez University of Seville Dept. Economía Aplicada I. University of Seville Av. Ramón y Cajal, 1. E41018 - Seville (Spain) KEYWORDS: Entrepreneurship Entrepreneurial intention Factor analysis Regression - 2 - Effect of cognitive variables on entrepreneurial intention levels Abstract Entrepreneurship is becoming a very relevant instrument to promote economic growth and development in different regional and national economies. However, social scientists have not still agreed on the determinants of the decision to become an entrepreneur. Therefore, there is some concern that policies may not be sufficiently efficient in achieving this objective. From a psychological point of view, the intention to become an entrepreneur has been described as the single best predictor of actual behaviour. Hence, some studies have started to analyse the entrepreneurial intention and its determinants but however, methodologies and research instruments used so far differ widely. Then, the availability of a validated instrument to measure abilities, attitudes and intentions towards entrepreneurship could be of much help. In this paper, we use an Entrepreneurial Intention Questionnaire (EIQ), which has been recently validated, to measure entrepreneurial intentions in a sample of students from two different Spanish universities. After a brief description of the characteristics of the EIQ, most important results are discussed. Additionally, we pay attention to the influence of different personal and family variables on entrepreneurial intention and its antecedents. 1. INTRODUCTION Economic development of a territory could be understood as the process of growth in average production per capita, maintained in the long term. In this sense, it is the consequence of the introduction by enterprises of -mainly technical and organizationalinnovations that allow productivity increases. These, in turn, help improve the retribution of factors used in the production process. The entrepreneur, therefore, is the main responsible for economic development, as it is understood nowadays. Most authoritative conceptions about the entrepreneur's figure (Knight, 1921; Shumpeter, 1934; Kirzner, 1998) stress his/her promoter role in the economy, above and beyond other more extended roles as manager and property owner. This economic function of entrepreneurs allows us to highlight their important role as development agents. According to it, entrepreneurs are responsible for the promotion of enterprises and businesses; they infuse dynamism in economic activity within their territory; manage organizational and technical change; and also promote the innovation and learning culture on such environment. Territories with higher increases on entrepreneurial initiative indexes tend to show a greater fall in unemployment levels (Audretsch, 2002). However, the entrepreneurial resource is scarce. In 2001, less than 10 percent of the OECD adult population was starting a new venture (Nolan, 2003). Therefore, a considerable agreement exists about the importance of promoting entrepreneurship to stimulate economic development and employment. - 3 - Entrepreneurship is an attitude that reflects an individual’s motivation and capacity to identify an opportunity and to pursue it, in order to produce new value or economic success (European Commission, 2003, p. 5). This attitude is crucial for competitiveness, because new entrepreneurial initiatives raise the territory’s productivity -increasing competitive pressure- and encourage innovation. In this sense, backward regions such as Andalusia tend to have low levels of firm creation rates and entrepreneurial activity (Westall et al., 2000), although it needs not be always the case (Nolan, 2003). If the entrepreneur is a development agent, as it seems, the Andalusian situation of relative backwardness might be a consequence of the low propensity of their people to be entrepreneurs. As Rodríguez-Pose (1998) affirms, some territories create institutions that favour economic development, while those built by other territories limit it. Perhaps Andalusian people’s attitudes -as an economic institutiondo not favour firm creation and economic development. In this paper, we try to address theses issues. Using the EIQ, which core elements have been recently validated, our main objective focuses on a wider analysis of entrepreneurial attitudes and intentions of Andalusian university students. In particular, we have concentrated our analysis in Seville, the largest metropolitan area in the region. Using intention models as the basis for research, we test the contribution of a broader set of variables to the intention of becoming an entrepreneur with a sample of university students. This paper has been structured in five parts. After this introduction, the second section presents the entrepreneurial intention model applied in the study. The third section describes the Entrepreneurial Intention Questionnaire used in the empirical analysis. The fourth part analyses the results obtained. Finally, the paper ends with some conclusions about the empirical work and its implications. 2. ENTREPRENEURIAL INTENTION MODEL In this section, we briefly summarise the entrepreneurial intention model. For a more detailed analysis, several studies cited along this paper may be considered. In particular, Liñán (2004) offers an easy-to-understand description. After many authors looked for the existence of certain personality features or traits that could be associated with the entrepreneurial activity (McClelland, 1961), other works have been carried out remarking the importance of different characteristics such as age, gender, origin, religion, level of studies, labour experience, etc. (Reynolds et al., 1994; Storey, 1994), which are usually called “demographic” variables (Robinson et al., 1991). Both lines of analysis have offer interesting results. However, many authors have criticized those approaches (Ajzen, 1991; Shapero & Sokol, 1982; Gartner, 1989; Santos, 2001; Veciana et al., 2000), so much for their methodological and conceptual limitations as for their low explanatory capacity. From a third perspective, since the decision to become an entrepreneur may be plausibly considered as voluntary and conscious (Krueger et al., 2000), it seems reasonable to analyze how that decision is taken. In this sense, the entrepreneurial intention would be a previous and determinant element towards performing entrepreneurial behaviours (Fayolle & Gailly, 2004; Kolvereid, 1996). In turn, the intention of carrying out a given behaviour will depend on the person's attitudes towards that behaviour (Ajzen, 1991). More favourable attitudes would make more feasible the intention of carrying it out, and the other way round. In this sense, this “attitude approach” would be preferable to those traditionally used in the analysis of - 4 - the entrepreneur, such as the traits or the demographic approaches (Robinson et al., 1991; Krueger et al., 2000). Thus, attitudes would measure the extent to which an individual values positively or negatively some behaviour (Liñán, 2004). The entrepreneurial intention model used in this paper is the result of integrating two different contributions: the theory of the “entrepreneurial event” (Shapero & Sokol, 1982) and the theory of “planned behaviour” (Ajzen, 1991). Both models present a high level of mutual compatibility (Krueger et al., 2000), and have had a notable influence on recent research. The theory of the entrepreneurial event considers firm creation as the result of the interaction among contextual factors, which would act through their influence on the individual's perceptions. The consideration of the entrepreneurial option would take place as a consequence of some external change -a precipitating event- (Peterman & Kennedy, 2003). People’s answers to that external event will depend on their perceptions about the available alternatives. There are two basic kinds of perceptions: Perceived desirability (attraction to become an entrepreneur) and Perceived feasibility (consideration of being able to create the firm). In turn, both types of perceptions are determined by cultural and social factors, through their influence on the individual's values system (Shapero & Sokol, 1982). Therefore, external circumstances would not determine firm-creation behaviours directly, but rather they would be the result of the (conscious or unconscious) analysis carried out by the person about the desirability and feasibility of the different possible alternatives in that situation. Along the same line, but much more detailed, Ajzen (1991) develops a psychological model of “planned behaviour”. It is a theory that may be applied to nearly all voluntary behaviours and it provides quite good results in very diverse fields, including the choice of professional career (Ajzen, 2001; Kolvereid, 1996). According to it, intention becomes the fundamental element towards explaining behaviour. It indicates the effort that the person will make to carry out that entrepreneurial behaviour (Liñán, 2004). And so, it captures the three motivational factors that influence behaviour (Ajzen, 1991): Perceived behavioural control (perception of the easiness or difficulty in the fulfilment of the entrepreneurial behaviour); Attitude towards the behaviour (positive or negative personal valuation about being an entrepreneur); and Perceived social norms (perceived social pressure to carry out -or not to carry outthat entrepreneurial behaviour). These three elements would constitute the explanatory variables of intention. Their relative contribution to the configuration of intention is not established in the model, as it may change from case to case. In particular, in the sixteen empirical works analyzed by Ajzen (1991), subjective norms tended to contribute very weakly to the intention of carrying out different behaviours. Finally, the model assumes the existence of interactions among the three explanatory elements. If we compare these explanatory variables with those considered by Shapero & Sokol (1982), we can see that perceived feasibility -as it has been mentioned abovecorresponds quite well with perceived behavioural control. On the other hand, the willingness to carry out entrepreneurial behaviours (perceived desirability) could be understood as composed by the personal attitude and perceived social norms. In this sense, it may be remembered that Shapero & Sokol (1982) considered desirability as a result of social and cultural influences. - 5 - Figure 1 Entrepreneurial intention model Source: Liñán (2004: 15), Figure 2. Additionally, as shown in Figure 1, a greater knowledge of the entrepreneurial environment will surely contribute to more realistic perceptions about the entrepreneurial activity and would help identify adequate role models. This latter element would have an influence on perceived feasibility and possibly on desirability as well (Scherer et al., 1991). In general, greater knowledge will also directly provide a greater awareness about the existence of that professional career option, and will make the intention to become an entrepreneur more credible. 3. THE ENTREPRENEURIAL INTENTION QUESTIONNAIRE (EIQ) The entrepreneurial intention model developed above will serve as the basis for analysing which factors may have a relevant effect on entrepreneurial intentions. In particular, last year undergraduate students of Business Sciences and Economics in the two existent universities of Seville, i.e. University of Seville and Pablo de Olavide University, have been studied. In a first stage of this research project, we used a preliminary version of a questionnaire, which contained mostly yes/no questions or items with four maximum different response options (Liñán & Rodríguez, 2004). Results were clearly encouraging and tended to support the entrepreneurial intention model developed above. However, the robustness of the results were not as desired. For those reasons, a second version of the questionnaire was developed, specifically designed to allow for a full validation of the questionnaire. The detailed process of construction and validation of the Entrepreneurial Intention Questionnaire (EIQ) has been explained in a different contribution to this conference (Liñán, 2005). The first version was used as the basis for this EIQ, but it has been carefully crosschecked following Kolvereid (1996), Krueger et al. (2000) and Veciana et al. (2000). Along the whole construction and design process, Ajzen’s (1991, 2001, 2002) work has been carefully revised to solve any discrepancies. Items have been built as 7-point likert-type scales. In particular, this has been true for the part of the EIQ measuring those latent variables that are crucial for the Entrepreneurial Knowledge Perceived Feasibility (self-efficacy) Perceived Desirability Personal Attitude Perceived Social Norms Entrepreneurial Intention - 6 - entrepreneurial intention model (see Figure 1): i.e., entrepreneurial knowledge, personal attraction, social norms, self-efficacy and intention. The EIQ has been divided in ten sections. Sections two to six corresponds with the elements in the entrepreneurial intention model and they include only one yes/no question for comparison purposes, which is the following: “Have you ever seriously considered becoming an entrepreneur?” as similar items like this one this have been already used by Krueger et al (2000), Veciana et al. (2000) and others, its inclusion in the EIQ would allow establishing comparisons. Section two asks interviewees to rate their knowledge of entrepreneurs and their perceptions about how good those entrepreneurs are. As Scherer et al. (1991) pointed out, having access to role models is one key element in explaining entrepreneurship. However, they consider that knowledge alone is not enough. It has to be completed with the evaluation made about how successful those entrepreneurs are. In this paper, interviewees’ evaluations of their role models have been included in the analysis together with demographic variables, in order to explain their effects on the entrepreneurial intention model. The first (education and experience) and ninth (personal data) sections ask for demographic variables that should not affect intention directly, but could be very useful in identifying their effect on perceived control, personal attitudes, social norms, and knowledge. The questionnaire also includes a seventh section centred on entrepreneurial objectives. Its purpose is to analyse students’ concept of “success” and the importance they ascribe to business development and growth. Guzmán & Santos (2001) defined entrepreneurial quality as the behaviours performed to develop the firm and make it dynamic. This section tries to measure the intention to perform such behaviours. The eighth section asked about participation in entrepreneurship education courses and the extent to which they have helped to increase entrepreneurial intention of interviewees, or any of the antecedents of intention. Finally, in section ten, we asked students to voluntarily provide contact data so as they may be studied again in the next future. This follow-up will hopefully allow for an analysis of the intention-behaviour relationship. 4. RESULTS Empirical analysis is supported by a sample of 354 last-year course undergraduate students from Business Sciences and Economics, at both two public universities in Seville (Spain). Most of them correspond to the University of Seville and the rest to Pablo de Olavide University, which presently does not include the Economics degree. There are two main reasons why such a sample may be selected. Firstly, lastyear students are about to face their professional career choice and secondly, these students belong to the empirically highest entrepreneurial inclination segment of the population, according to Reynolds et al. (2002). That is, those individuals between 25 and 34 years old with high level of education tend to show a greater propensity towards entrepreneurship. - 7 - Business Sciences students represent 69.21% and the rest corresponds to Economics degree. Pablo de Olavide University provides only 31 students whereas the University of Seville, 323. This is because the former university was founded just eight years ago and it is still relatively small. 55% of the interviewees are female, while the average age is 23.7 years old. Factor-Regression procedure (FR procedure) The empirical procedure developed in this paper can be defined as a mixture of factor analysis and regression. SPSS is the statistical software used for factor analysis and Econometric Views for regressions. The first step consists of carrying out a factor analysis with the aim of distinguishing all different factors influencing entrepreneurial intention. At this initial stage, most available items were included in the analysis. Therefore, the so-called antecedents of intention (knowledge, personal attitude, social norms and self-efficacy) were entered together with a whole set of other variables measuring possibly relevant concepts. Only those items regarding degree studied, labour experience and demographics (gender, age, place of residence, …) were excluded at this step. The main purpose is to check whether the inclusion of those other factors would lead to a different set of explaining variables for intention. The factor analysis was carried out using a principal axes factorization for extraction and a promax rotation. Pardo & Ruiz (2002) suggest these options as the most adequate when a regression is to be performed. After factors have been identified, a regression analysis is estimated including the entrepreneurial intention factor as the explained variable and the rest of factors as explanatory variables. The regression is tested to solve, if that would be the case, heteroscedasticity, multicollinearity and autocorrelation problems. As a result, some of the explanatory variables may not be significant at a 95% confidence level, so they would be removed from the model. Once the regression model is adjusted, the items that correspond to the remaining factors are again included in a second step factor analysis. In this way, we obtain a new set of factors that may be reflecting the underlying structure. Once more, these factors are related through a new regression model, and so on. Hence, the final result would be a model which can explain how certain factors may influence the entrepreneurial intention. However, as different measures have been used to evaluate each antecedent of intention, and some additional variables have also been included (especially those related to entrepreneurial objectives), there is a risk of finding factors with no easy interpretation. Nevertheless, as it is shown later, these factors are fortunately found to be closely related to our theoretical a priori expectations (see Figure 1). Our point of departure is the following different features in which the EIQ is organized: (a) Entrepreneurial knowledge (11 items in 3 groups). It includes “knowledge of the institutional business framework” and other items regarding knowledge of different kinds of entrepreneurs. (b) Personal attitude (11 items in 3 groups). It includes “personal attraction” and other items regarding valuation of different career options in the short and long run. - 8 - (c) Perceived social norms (11 items in 3 groups). It includes the core “social norms”, together with items regarding social valuation of entrepreneurship as a career option and existence of social obstacles to be an entrepreneur. (d) Perceived feasibility (12 items in 2 groups). It includes “self-efficacy” and items regarding perceived level of specific abilities. (e) Entrepreneurial intention (7 items in 2 groups). It includes the “intention” scale together with the yes/no question (have you ever considered…?). (f) Objectives (17 items in 4 groups). This additional group includes items regarding desired size of eventual future firm, importance attached to different dynamic behaviours and concepts of “entrepreneurial success”. The first factor analysis is performed with just 313 individuals due to nonanswered questions. It yields up to 19 different factors (see Table 1). It is interesting to remark the negativity of the three following factors: growth as a key feature for success, preference for organizational employment and personal attitude to entrepreneurship. That is, individuals with high entrepreneurial intention do not really think about growth as a key feature for success, do not prefer organizational employment, and do not have much attraction for being entrepreneur. This latter relationship is against our a priori expectations. Perhaps, reasons different than their personal view of the advantages and disadvantages of entrepreneurship may explain the entrepreneurial intention. For instance, other factors as the perceived feasibility or the perceived social norms, among others, are positively related with an entrepreneurial behaviour. The determination coefficient of the model is 0.538. Table 1 Factors Sign of Coefficient Entrepreneurial intention Perceived feasibility * Knowledge of support measures Foreign trade and innovation Perceived social norms* Growth as a key feature for success* Friends approval for entrepreneurship* Leadership, communications and professional contacts Friends valuation of entrepreneurship Social approval and turnovers as key features for success Planning, alliances and formation for employees* Innovation, creativity and detecting opportunities abilities* Preference for organizational employment* Entrepreneurs in the family* Preference for being independent professional* Personal attitude** Preference for continuing education Size, development and entrepreneurial enlargement Knowledge of non-family entrepreneurs (explained) Positive - - Positive Negative Positive - - - Positive Positive Negative Positive Positive Negative - - - * Significant at a 95% confidence level. ** Significant at a maximum 93% confidence level. The second step of the factor analysis was carried out including the remaining 10 significant factors plus the entrepreneurial intention factor. The results are shown in Table 2. In contrast to the first step, personal attitude changes its sign in the expected direction, becoming positive. That is, in this case, individuals with high entrepreneurial intention find more advantages than disadvantages in being entrepreneur. - 9 - Regarding the sign of the coefficients associated to growth as a key feature for success and the preference for organizational employment, they remain unchanged. Notice that planning, alliances and promoting higher formation on employees are positively related behaviours with respect to the entrepreneurial intention. Lastly, personal attitude and perceived feasibility (as theory indicated) are two of the main explanatory factors regarding entrepreneurial intention in these last-year course undergraduate students. Table 2 Factors Sign of Coefficient Entrepreneurial intention Perceived feasibility * Personal attitude* Perceived social norms Planning, alliances and formation for employees* Friends approval for entrepreneurship** Growth as a key feature for success* Ability of innovation, creativity and detecting opportunities Preference for organizational employment* (explained) Positive Positive - Positive Negative Negative - Negative * Significant at a 95% confidence level. ** Significant at a maximum 93% confidence level. However, perceived social norms are not really significant at a 95% confidence level, which is coherent with Ajzen (1991). In this case, with 337 observations, a 0.691 determination coefficient is obtained after solving some heteroscedasticity problems using the White consistent standard errors for the estimated coefficients. No problems of multicollinearity or autocorrelation were found. In a third step, we made another factor analysis including the remaining 6 factors plus the entrepreneurial intention factor as the explained variable. The results are shown in Table 3. At this stage, the friends’ approval for initiating entrepreneurial activities is not significant and therefore will be removed, while all the rest of factors (and their signs) remain unchanged. Table 3 Factors Sign of Coefficient Entrepreneurial intention Perceived feasibility * Personal attitude* Planning, alliances and formation for employees* Friends approval for entrepreneurship Growth as a key feature for success* Preference for organizational employment* (explained) Positive Positive Positive - Negative Negative * Significant at a 95% confidence level. Finally, a last factor analysis is carried out with the remaining 5 explanatory factors and the corresponding explained variable entrepreneurial intention. The final results are shown in Table 4. At this stage, 338 observations were used. The regression model yields a 0.70 determination coefficient, which is satisfactory.