1. Introduction
In the last decades, entrepreneurship has been a phenomenon that has attracted increasing interest from some of the different regional actors in a globalized and knowledge-based global economy [
1,
2]. Entrepreneurship is a driver of sustainable regional economic development and a catalyst of innovation, which in turn increases job creation [
3,
4].
In this context, policy makers are generally aware that entrepreneurship should be a priority for the development of any region. Thus, government officials constantly try to accelerate the creation of entrepreneurial ecosystems, providing financial support through policy development, and education programs for entrepreneurship and innovation [
5,
6,
7]. As such, they intend to create an environment that favors and assists students in higher education to start their own companies [
3,
8]. However, the number of companies created by graduate students has not increased significantly [
9], on the contrary, it is falling below the level idealized by regional governments and universities [
10].
Over time, universities have evolved to adapt to the market needs and their students’ expectations. Nowadays, the mission of universities is no longer focused on teaching and research (the universities’ first and second mission). A third mission of the universities emerged, the entrepreneurial university, which ensures that they are involved in the transfer of knowledge to the local community, through entrepreneurial activities [
11,
12,
13]. The third mission of the universities expects to give contributions to the social and economic development of the regions where they operate [
14].
Within this context, in 2014, European Union policy makers changed their regional development policies, which are called research and innovation strategies for smart specialization (RIS3). RIS3 is based on the concept of smart specialization in areas and regions that have traditions [
15]. In this manner, it is expected that the regions will be able to increase their innovative performance, and also seek to have a regional economy, progressively based on technology [
16,
17].
Studies on academic entrepreneurship have been performed over time. Some studies were focused on the personality traits of individuals and how they influenced the promotion of entrepreneurial intentions [
18,
19,
20]. Other studies highlighted the relevance of the interaction between individual and contextual factors in the investigation of entrepreneurship and innovation [
21,
22]. Nonetheless, there are already studies that focus on individual characteristics, such as entrepreneurial mindset, traits, and personality [
23], and gender [
24]. Some studies consider family history as parents’ employment status and occupational status [
25], as well as previous exposure to companies that are family owned [
26]. Socio-environmental factors, such as regional policy, have also been considered in the study of academic entrepreneurship [
27].
Despite the previously mentioned facts, societies are evolving, and the shifting conjunctures bring new challenges. With this evolution, the competition between companies is becoming fierce, innovation and entrepreneurship is being constantly encouraged, consequently leading the world to become more and more global. Thus, the climate of uncertainty has grown and the accelerated development of technology brought changes in social behavior. Education for entrepreneurship has a fundamental role to prepare citizens to be talented and prepared to face the impact resulted from globalization and social transformation [
1,
28,
29]. Education should always be adjusted to these new needs. Being so, it is still necessary and pertinent to further investigate the various factors that affect academic entrepreneurship [
3,
30]. According to Anjum, et al. [
31] and Dentoni, Pinkse and Lubberink [
3], it is important to study the entrepreneurial intention in students from different countries and different educational backgrounds.
That said, the present research aims to evaluate the determinants that influence the interest of Portuguese higher education students to be entrepreneurs. This research brings new contributions as it contemplates variables that have been little explored in academic entrepreneurship such as business growth, skills, strategy, and success. From a launched questionnaire, 1114 responses were collected, and the results indicate that business growth has a positive impact on students’ willingness to become entrepreneurs. It also indicates that greater business skills have a positive impact on students’ willingness to become entrepreneurs. It was also confirmed that the use of business strategy for business expansion has a positive impact on the entrepreneurial attitude. Moreover, we have also corroborated that the appreciation of business success factors has a positive impact on students’ willingness to become entrepreneurs.
This research begins with an introduction where the problem under study and the respective objective is presented. In part two, the literature on academic entrepreneurship is extensively reviewed and hypotheses formulated. In part three the full methodological process and data collection methods are detailed. In part 4 the results are presented, discussed, and compared with the literature. Finally, the conclusions and main findings are presented as well as theoretical and practical contributions, followed by limitations and clues for future investigations.
3. Methodology and Data Collection
To accomplish the main objective of this research, a quantitative methodology was chosen. This type of methodology has been widely used to develop studies on entrepreneurship [
88,
89]. Its main advantages are to allow the validation of theories and relationships between variables, generalize results, and replicate methods and techniques in new samples [
90,
91,
92]. Furthermore, a questionnaire was used as a tool for data collection.
The questionnaire applied in this research was previously used in studies of Linan, Urbano and Guerrero [
37] and Lopes, Teixeira, Ferreira, Silveira, Farinha and Lussuamo [
30]. We collected a sample of 1114 observations through an online questionnaire that was applied to higher education students in Portugal between April 2017 and December 2020. Students attend courses in business science (management and related) in which the subject of entrepreneurship is taught, and as such, their course includes in its syllabus all important and emerging topics of entrepreneurship. A convenience sampling method was used. Initially, we search for HEIs located in Portugal. For this purpose, an online database containing Portuguese HEIs was used (
https://www.universia.net/pt/universidades.html, accessed on 20 August 2021). From the outcome of the online database, 60 Portuguese HEIs were contacted. The objective of this contact was to request the distribution of a questionnaire to their students via email. From the HEIs contacted, 14 responded positively, referring that they would distribute the questionnaire amongst their students, 30 responded negatively denoting that they do not send questionnaires to their students, and 16 did not reply at all. Moreover, the questionnaire was also disseminated in social networks, namely Facebook and Twitter.
The questionnaire consisted of nine groups of questions: (1) Entrepreneurial Intention with four questions; (2) Perceived Behavior with four questions; (3) Entrepreneurial Attitude with three questions; (4) Business Strategy with eight questions; (5) Business Growth with a question; (6) Business Success with seven questions; (7) Business Skills with six questions; (8) Being an Entrepreneur with one question, and (9) Sociodemographic characteristics with four questions related to age, sex, residence, and employment experience.
All the groups of questions mentioned above use a 7-point Likert scale, except for questions related to sociodemographic characteristics. In the groups of questions related to Entrepreneurial Intention (G1), Perceived Behavior, and Entrepreneurial Attitude a scale where 1—strongly disagree to 7—strongly agree is used; the Business Strategy questions group uses a scale where 1—not likely to 7—extremely probable; the Business Growth and Business Success question groups use a scale where 1—not important and 7—extremely important; the Business Skills question group uses a scale where 1—no aptitude and 7—total aptitude and the question about Being an Entrepreneur uses a scale where 1—not at all attractive and 7—extremely attractive.
Regarding sociodemographic characteristics (
Table 1), respondents are mostly women (65.6%) and have an average age of 26.53 years. A total of 76.8% respondents are under 30 years old, 31.6% are under 20 years old, and 45.2% are over 21 years old and less than 30 years old. In terms of residence, 83.8% live in mainland Portugal, 9% in the Azores, and 7.2% in Madeira. Most respondents are studying or have an undergraduate degree (67.6%) and 68.9% have professional experience.
Regarding the question groups of Entrepreneurial Intention, Perceived Behavior, Entrepreneurial Attitude, Business Strategy, Business Growth, Business Success, Business Skills, and Being an Entrepreneur, the mean and standard deviation of the answers are found described in
Table 2.
In the Entrepreneurial Intention group, in average terms, the issues of greatest agreement amongst the respondents were the ones related to the commitment of all efforts to start and manage a new business (A06—4.89) and the determination to create an enterprise in the future (A13—4.47); in the questions of Perceived Behavior, the questions related to the perception of success (A14—4.64) and the perception of the ability to control the process of creating a new business (A07—4.6) were the ones that generated the greatest agreement, being that respondents demonstrated, on average, to have little knowledge of all the practical details needed to start a new business (A20—2.93).
At Entrepreneurial Attitude, most respondents agree that they would love to open a new business (A10—5.74) and that being an entrepreneur would give them great satisfaction (A15—5.07). In terms of Business Skills, respondents showed more aptitude in terms of problem-solving skills (D3—5.39), leadership and communication skills (D4—5.28), as well as creativity (D2—5.02). As for Business Success, respondents consider it, on average, more important to maintain a positive growth path (7.g.—6.24), to do a type of work that they really appreciate (7.c.—6.16), and keep the business alive (7.f.—6.24).
Continuous development and business growth (Business Growth) is, on average, very important for respondents (8.—6.09) and they consider it, on average, to be an entrepreneur—Being an Entrepreneur (B2—5.33). Finally, in terms of Business Strategy, it is very likely, on average, that respondents will offer specialized training to their employees (9.g.—6.07), grow their business (9.h.—5.98), plan in detail the different areas of the company (9.e.—5.79), and establish cooperation agreements or partnerships with other companies (9.f.—5.75).
4. Results Discussion
The structural model shown in
Figure 1 was estimated using the Partial Least Squares (PLS) method in the Smart PLS 3.0 software [
93]. PLS is one of the approaches of Structural Equation Models (SEM), which allows estimating the causal relationships between variables, defined by a theoretical model. The nature of these relationships is not directly observable, so one or more indicators are used to measure them. The main focus of this technique is on the ability to be able to analyze the complexity of a system, based on a set of latent concepts and indicators, given by the Latent Variables and Manifest Variables, respectively [
94]. In this manner, PLS is a structural variance-based model, used to develop theories in the context of exploratory research. Its objective is to maximize the explained variance between the dependent variables of the model, that is, the R Square value. Thus, it allows for the testing of complex theoretical relationships defined by the supporting literature and enhances the probability of identifying significant relationships between variables when in fact these relationships exist in the sample [
95,
96].
In the estimated model, eight latent variables were created: three endogenous latent variables (Entrepreneurial attitude, Entrepreneurial Intention, and Being an Entrepreneur) and five exogenous latent variables (Business Strategy, Business Growth, Business Success, Business Skills, and Perceived Behavior), as previously mentioned, and a total of 34 indicators were used. The use of the PLS model requires the validation of the sample size which, according to Hair, et al. [
97], must be at least equal to one of the following conditions: (1) ten times the number of indicators, or (2) ten times the number of structural paths directed to a latent variable in the structural model. The sample has 1114 observations and, thus, is ten times greater than the number of indicators, fulfilling the first condition. On the other hand, there are seven structural paths between the latent variables, also fulfilling the second condition. We conclude that the sample used is suitable for the application of the PLS method. On the other hand, the application of this method does not require normal data, the measurement scale used is generally metric, and the ordinal scale is also accepted [
97,
98].
To measure the quality of the model obtained by the PLS method, it is necessary to analyze the discrepancy between the values of the dependent variables, whether they are observed (in the case of manifest variables) or approximate (in the case of latent variables), and the value predicted by the model. Consequently, the overall quality of the model is given by its predictive capacity, and the measurement and structural models that compose it must be validated. One of the measures used is reliability and validity, that is, instruments for mediating the relationship between the latent and observed variables of the model, which implies an analysis of the reliability of each latent variable at the indicator level and the convergent and discriminant validity. For this, the indicator factor loadings, the reliability, and average variance extracted (AVE) of each indicator used should be analyzed, as shown in
Table 3.
The reference value for reliability coefficients of latent variables is 0.70 (Hair et al., 2019). All the estimated reliability coefficients of the model are higher than the reference value as shown in
Table 3 and as such, they are “satisfactory to good”, meaning that all latent variables are above the acceptable values for the outer loadings, reliability, and validity of the model estimated.
Latent variables have high indicator loads (greater than 0.555) and acceptable validity and convergence measured by Cronbach’s Alpha (all results of this indicator are greater than 0.700—reference value). Cronbach’s Alpha gives us an estimate of the reliability between the observed indicators and the corresponding latent variable. The Average Variance Extracted (AVE) of the latent variables indicates the estimate of how much the variation of the indicators is due to the inherent latent variable, having a reference value of 0.50. In the estimated model, for all latent variables, a stroke greater than 0.50 was obtained.
Finally, the Fornell–Larcker criterion was used as a measure of Discriminant Validity since this criterion allows the analysis of cross-loadings that are indicators of the discriminant validity of latent variables.
Table 4 shows the results of applying the Fornell–Larcker criterion.
Each AVE of the latent variables (elements in the main diagonal that are in bold) is superior to all the square correlations of the latent variables (elements outside the diagonal), thus establishing the discriminant validity of each of the eight latent variables.
The model has been validated by the measures of reliability and validity and discriminant validity, and the PLS algorithm in Smart PLS 3.0 was applied to this model, resulting in the PLS Path Model shown in
Figure 2. In this model, the algorithm converged the parameter of the PLS-SEM algorithm (out of 300 iterations) after the 7th iteration.
The PLS Path Model contains the 34 collected indicators (represented in the rectangles) and the eight latent variables created—Entrepreneurial Intention, Perceived Behavior, Entrepreneurial Attitude, Business Strategy, Business Skills, Business Success, Business Growth, and Being an Entrepreneur. The structural relationships established between the latent variables are in accordance with the structural model shown in
Figure 1.
The predictive precision of the PLS Path Model is assessed by analyzing the R Square (R
2) values of the endogenous (dependent) latent variables, that is, Being an Entrepreneur, Entrepreneurial Intention, and Entrepreneurial Attitude. Several authors report reference values for R
2 to be different according to the areas of study. Ritchey [
99], states that 0.02 represents a “small” effect, 0.15 represents a “medium” effect, and 0.35 represents a “high” effect. Hair, Sarstedt, Matthews and Ringle [
97] considered that, in general, the reference R
2 values are 0.75, 0.50, and 0.25 and, consequently, the endogenous latent variables are, respectively, described as substantial, moderate, and weak. Höck and Ringle [
100] state that R
2 of 0.67, 0.33, and 0.19 are “substantial”, “moderate”, and “weak”, respectively.
In this model, the R
2 values of the endogenous latent variables are: Being an Entrepreneur > 0.510, Entrepreneurial Intention > 0.705 and Entrepreneurial Attitude > 0.102. Following the most current criteria by Hair, Sarstedt, Matthews and Ringle [
97], we can refer that the R
2 of the latent variables Being an Entrepreneur and Entrepreneurial Intention is “moderate” and the latent variable Entrepreneurial Attitude is “weak”.
The significant relationships between the latent variables are measured by their path coefficients. As shown in
Figure 2, there are significant relationships between the latent variables, the most significant being between Entrepreneurial Attitude > Entrepreneurial Intention (0.599, that is, a variation of 1% in the variable Entrepreneurial Attitude has a positive impact of 59.9% on the Entrepreneurial Intention variable) and between Entrepreneurial Intention > Being an Entrepreneur (0.588, that is, a 1% variation on the Entrepreneurial Intention variable has a positive impact of 58.8% in the Being an Entrepreneur variable).
Once the path coefficients were calculated, the bootstrap method was applied, with a significance level of 95%, to test the hypotheses formulated in this study. This non-parametric technique allows analyzing the accuracy of the estimates of the PLS parameters. Its procedure is to obtain a specific number of sub-samples of the same size as the original sample. The selection of observations is made through sampling with replacement; to obtain more reasonable estimates, a high number of sub-samples is advised (in this study, in Smart PLS, 500 sub-samples were used).
Table 5 shows the results of applying the bootstrap method. We conclude that all latent variables are very significant at
p < 0.05, with 95% confidence.
In such a way, and following the conclusions by Birley, Cromie and Myers [
79] and Donckels and Lambrecht [
80], the results obtained reinforce the importance attributed by the undertaking of continuous development and business growth (Business Growth) has a positive impact (β = 0.0756) on the desire to be an entrepreneur, confirming Hypothesis 1.
Greater Business Skills such as the recognition of opportunities, creativity, problem-solving skills, leadership and communication skills, ability to develop new products or services, and to form networks and professional contacts have a positive impact (β = 0.0796) relationship previously mentioned by [
83,
84].
As defended by Pobee and Mphela [
75], Damke, Gimenez and Damke [
81] and Hashimoto and Nassif [
82], the results attained reveal that the use of business strategies to expand a business (Business Strategy) such as the export of a significant part of the production, the regular introduction of new products/services and production processes, the development of research projects and development, detailed planning of the different areas of the company, the establishment of agreements or partnerships with other companies, the provision of specialized training for employees, and the willingness to grow the business (in personnel and facilities) has a positive impact (β = 0.3188) on Entrepreneurial Attitude, thus confirming Hypotheses 3.
The valuation of Business Success factors such as effective competition in world markets, achieving a high level of income, doing work that is appreciated, being socially recognized, helping to solve the surrounding problems, keeping the business alive, and maintaining a path of positive growth has a positive impact (β = 0.1086) on the desire to be an entrepreneur, confirming Hypothesis 4. This significance confirms the studies developed by Mitchelmore and Rowley [
85], Rasul, Bekun and Akadiri [
86], Kummerow, Wilson, Ramayah and Hazlina Ahmad [
87], Boyd and Vozikis [
34], Anastasia [
77], and Fillis and Rentschler [
78].
The results also demonstrate that Entrepreneurial Attitude is measured by will, if there was an opportunity to start a new business, by the satisfaction of being an entrepreneur, and by the recognized advantages of being an entrepreneur has a positive impact (β = 0.5986) in Entrepreneurial Intention, confirming Hypothesis 5. This relationship has already been evidenced by several authors, including [
38,
42,
72].
In turn, Entrepreneurial Intention, measured by the desire to be an entrepreneur at any cost, by the efforts committed to start or manage a business of its own, by the determination to create a business in the future, and by the professional objective of being an entrepreneur has a positive impact (β = 0.5877) on the desire to be an entrepreneur, confirming Hypothesis 6, according to what was advocated by Kristiansen and Indarti [
76]. Finally, Perceived Behavior, measured by the ease of starting a company and keeping it viable, the ability to control the process of creating a new business, the likelihood of being successful in creating a business, and the knowledge of all the practical details needed to start a new business have a positive impact (β = 0.3348) on Entrepreneurial Intention, confirming Hypothesis 7. These results are in line with what has been reported in several studies, such as Linan, Urbano and Guerrero [
37], Kolvereid [
43], Krueger, Reilly and Carsrud [
45], Fayolle and Gailly [
47], and Vamvaka, Stoforos, Palaskas and Botsaris [
72].
In addition to the direct path coefficients, the model also allows estimating four path coefficients of indirect effects (
Table 6). Values are obtained, for example, for Entrepreneurial Attitude > Entrepreneurial Intention > Being an Entrepreneur (0.3518), making the product between the Entrepreneurial Attitude > Entrepreneurial Intention (0.5986) influence, and the Entrepreneurial Intention > Being an Entrepreneur (0.5877). It means that a 1% variation in Entrepreneurial Attitude has an indirect impact of 35.18% on the Being an Entrepreneur latent variable. As we can see in
Table 6, all indirect effects are statistically significant with
p = 0.000, for 95% bootstrap.
Also in
Table 6, the total effects of exogenous latent variables on endogenous latent variables are represented, with the most expressive total effects being verified in the influence of Entrepreneurial Attitude > Entrepreneurial Intention (0.5986) and Entrepreneurial Intention > Being an Entrepreneur (0.5877). All total effects are statistically significant at
p < 0.05 to 95% bootstrap.
We can also conclude that the indirect effects and the total effects confirm the results obtained in the significance test (
Table 5), that is, there is a direct, indirect, and total positive influence of Business Growth, Business Skills, Business Success, and Entrepreneurial Intention in the will to be an entrepreneur (Being an Entrepreneur) and there is also a positive influence of Business Strategy in Entrepreneurial Attitude and Perceived Behavior in Entrepreneurial Intention, confirming again all the hypotheses formulated in this study (
Table 7).
5. Conclusions
In the context of cooperation with and for the community, through the integration of business groups and social partners, universities develop activities that go beyond training by transferring knowledge, which has a fundamental importance in the development of the region and country where they operate. Inside the university boundaries, academic training may potentiate a wish within the students to follow entrepreneurial activities.
Within this framework emerges the main objective of this research, which aims to evaluate the determinants that influence the interest of Portuguese higher education students to become entrepreneurs. Based on the development of a structural model, and grounded by the relevant literature, which studies the direct and indirect relationships between the variables Entrepreneurial Attitude, Entrepreneurial Intention, Being an Entrepreneur, Business Strategy, Business Growth, Business Success, Business Skills, and Perceived behavior, a total of 34 indicators were accounted.
The results allowed us to conclude that several factors influence the entrepreneurial intentions of higher education students. Specifically, in line with the relevant literature, we found that the will to be an entrepreneur, and therefore the entrepreneurial intention, is directly affected by behavioral variables (such as entrepreneurial attitude and perceived behavior), as well as by the perception of young people in the workplace, which refers to aspects related to the business (including business growth, business strategy, business success) and business skills.
Being so, we confirm the vital role of universities in the training of future entrepreneurs, not only by providing the development of behavioral, social, and technical skills for students to develop their business in the future but also by allowing them to integrate a set of networks and projects, which allow them to identify/perceive the value of projects in terms of their business growth, business strategy, and business success.
This research contributed to the theory by adding new knowledge to the literature on the perception of the HEI’s students to become entrepreneurs, specifically the students of Portuguese universities. In practical terms, the contributions offered within this research are based on suggestions on the third mission of universities and knowledge transfer to the community, business groups, and policy makers, as well as the creation of the essentials within university limits to promote entrepreneurship amongst its students.
Attracting more students for entrepreneurship, innovation, and knowledge transfer to companies and communities will be increased, thus allowing economic growth and development and job creation.
This research is original and innovative, as no research on this field and with all the aggregated elements under study was performed in Portugal. Moreover, this research provides universities and other local entities with the knowledge of drivers that capture the interest of students, and thus creates projects and training that go according to their motivations.
As for clues for future investigations, it would be significant to integrate into this study other variables, which according to the relevant literature, may influence the entrepreneurial intention of higher education students in Portugal, namely, the students’ perception of the norms and institutions in which they are inserted, for example, regulations, financing, advice, and instruments to support entrepreneurship. It is also important to compare the results obtained in Portugal with results from other countries in which entrepreneurship is more interesting and captivating, and where the rates of creation of companies by recent graduates are higher. In order to assess the more active role of universities as drivers of entrepreneurial activity, we must also assess the existence of entrepreneurship promotion centers provided by universities, the existence of incumbents of new companies and the contents of the entrepreneurship subjects that make up the plans evaluated of higher education students. From this analysis, lessons can be learned about the policies and strategies to be adopted in promoting entrepreneurship in Portugal and in creating jobs through nascent entrepreneurship.
Another possibility for future research could be the production of a longitudinal study, which has the advantage of showing the evolution of the indicators over time and making temporal comparisons. Thus, it would be pertinent, for example, to compare student behavior in an environment of economic growth and recession.