The Influence of Emotional Competencies on the Entrepreneurship Intentions of University Students in Colombia

Abstract

The main objective of this article is to analyze the effect of a group of predictors on entrepreneurial intention from the perspective of emotional competencies. To achieve this aim, a sample of 996 students belonging to ten public and seven private universities was selected, with each university having been granted high-quality accreditation by the Colombian Ministry of Education. The theoretical approach was based on Ajzen’s Theory of Planned Behavior (TPB), and the empirical analysis was based on structural equation models. Eight hypotheses were tested, with entrepreneurial intention as the dependent variable and emotional competencies, subjective norms, entrepreneurial self-efficacy, and entrepreneurial attitude as independent variables. The information was obtained by applying a questionnaire with a Likert scale to students of subjects related to entrepreneurship. From the findings of the study, it is concluded that in the two ecosystems analyzed (public and private), emotional competencies (EC) have neither a direct influence on entrepreneurial intention (EI) nor an indirect one since no significant influence is observed between emotional competencies and entrepreneurial attitude; however, a direct and positive effect was recorded between the self-efficacy and entrepreneurial attitude constructs on entrepreneurial intention. In addition, it is observed that subjective norms (SN) do not directly affect EI, but they do influence it indirectly, being mediated by entrepreneurial attitude and self-efficacy. The contribution of this study is focused on obtaining a better understanding of the entrepreneurial intentions of university students in Colombia, which will make it possible to foster strategies for the generation of youth employment and public policies to promote various entrepreneurial initiatives. This could be based on government regulations adopted in the last decade, which are still under development, and the broad participation of university students and research groups of higher education institutions. Furthermore, given the dearth of research examining the impact of emotional competencies on the entrepreneurial intentions of young Colombian university students, this study aims to bridge the existing knowledge gap, thereby contributing to the development of a more robust body of literature that can inform the design and implementation of educational strategies and public policies aimed at fostering entrepreneurship within the university ecosystems of this country.

1. Introduction

University students are one of the main drivers of entrepreneurship in the contemporary economy, and their participation contributes to a better economic and social performance of the countries where initiatives around entrepreneurship have spread, as stated by Aftab [1], Huezo et al. [2], Lihua [3], and Nayak et al. [4]. This trend is not unrelated to the commitments of governments to get involved in the fulfillment of the 2030 Agenda, which frames the Sustainable Development Goals (SDGs) proposed by the United Nations, particularly with regard to SDG 8 (sustainable economic growth and decent work for all). In the field of corporate social responsibility, the aforementioned concern has given rise to research that proposes green business strategies to ensure a sustainable future [5].

Hence, a manifestation of the importance the university population has is the fact that, according to the Gran Encuesta Integrada de Hogares (Large-scale Integrated Household Survey) released by the National Statistics Department (DANE), the percentage of the employed population with a university education in Colombia increased from 7.0% in 2010 to 12.0% in 2023. In addition, the percentage of Colombian university graduates who are self-employed represented, on average, 27.0% of this occupational category during the period of 2010–2023. The risks and the failure rate of entrepreneurs under these categories are not few, as suggested by Chien Chi et al. [6].

It is precisely entrepreneurship that is perceived as an attractive alternative for young university students to obtain employment and income, given the high levels of unemployment in the population between 15 and 28 years of age, which, according to DANE, was 16.5% in the October–December quarter of 2023. Furthermore, unemployment among people with a university education was 11.6% in 2023, which contrasts with the indicator recorded in 2010: 6.1%. At the same time, a worrying phenomenon that has emerged in recent years is the number of young people who neither study nor work, which represented 23.6% in the October–December 2023 period (according to DANE 2024). Therefore, entrepreneurship has become a possible alternative for young university students as a way of facilitating greater labor inclusion and, consequently, providing better income levels while at the same time becoming an important source of human capital and thus contributing to the country’s development.

Despite the existing interest, as evidenced by the issuance of a national entrepreneurship policy accompanied by a regulatory framework in Colombia, studies on emotional competencies are scarce in the country. Therefore, it is becoming increasingly necessary to conduct research that explores the capabilities and motivations of university students to undertake the creation of companies or businesses.

For these reasons, as of 2020, the National Council for Economic and Social Policy (CONPES) published Document 4011 (the National Entrepreneurship Policy), which outlines important guidelines that should guide entrepreneurial initiatives in Colombia. In turn, it enacted laws 2069 of 2020 and 2125 of 2021, which highlight the expediency of establishing alliances as well as support mechanisms, with cooperation from the Ministry of Science, Technology and Innovation, in order to strengthen the entrepreneurship and innovation systems already in place in higher education institutions.

With respect to the above, the problem to be elucidated is as follows: How do emotional competencies influence the entrepreneurial intention of university students in Colombia? In order to answer this question, several hypotheses were formulated with the aim of identifying the existence of direct or indirect relationships between the variables defined in Ajzen’s planned model [7], which will serve as a theoretical reference for this analysis, which is extended to compare the Colombian experience with that of Spain and Mexico. It should be noted that Spain and Mexico are benchmarks of comparison for Colombia, given that there are several coincidences in the entrepreneurial university ecosystems. In addition to cultural links (language, customs, and religion, among others) that date back to the Spanish colonial period, there is also an inherited institutional framework that has left its mark in the two Latin American countries.

Thus, the present study seeks to determine the influence of some variables or constructs, such as emotional competencies (EC), subjective norms (SN), self-efficacy (SE), and entrepreneurial attitude (EA) on the entrepreneurial intention (EI) of university students, as suggested by Schalegel and Koenig [8] and Karimi et al. [9], among others. However, as noted by Fernández-Pérez et al. [10], emotional competencies alone do not lead university students to undertake an entrepreneurial project or business.

Two university entrepreneurship ecosystems were used to analyze these relationships: seven private universities and ten public universities in Colombia. Structural equation modeling (SEM) was used to validate the eight proposed hypotheses.

Of the total of 996 students from the university ecosystem surveyed, 660 (66.3%) represent students from the public university ecosystem. The remaining proportion (33.7%) corresponds to the private ecosystem. It should be noted that the educational institutions to which the surveyed students belong are located in eight cities in the country with significant population and economic activity: Bogota, Medellin, Cali, Cartagena, Barranquilla, Pereira, Tunja, and Leticia.

This article includes the theoretical foundations of emotional intelligence and its construct, emotional competence, and the formulation of the working hypotheses. The methodology used and the sample selected are described below. The results obtained in the validation of the eight working hypotheses included on the basis of the proposed model are also analyzed. A comparison is made with the results obtained in other studies, such as the cases of Mexico and Spain, and the conclusions are presented, as well as some practical implications and limitations of the study.

According to Tingting et al. [11], in a bibliometric study on college students’ entrepreneurial intention covering the period of 2000–2020, of the 454 articles reviewed, about 41% (184) are based on TPB, stating “TPB is also the main theory presented in the literature representative of the first category of cluster labels. Therefore, it occupies an important position in research on college students’ EI and is the main theoretical basis for this research” (p. 05).

It is hoped that the results obtained in this study can identify the motivations present in university students with entrepreneurial propensity so that they can be channeled both by higher education institutions and through governmental strategies and policies, such as those set forth in the new regulatory framework adopted in Colombia since 2020. Furthermore, it is hoped that this work may inspire new lines of research that contribute to the reinforcement of entrepreneurship as a source of economic growth and gainful employment.

2. Theoretical Basis and Hypothesis

2.1. Relevance of Emotional Intelligence

The concept of emotional intelligence is preceded by the various meanings of intelligence, with multiple intelligences as a reference, which emerged in the 20th century [12,13]. Salovey and Mayer [14] coined the expression emotional intelligence, thus ushering in a new paradigm not so distant from Gardner’s multiple intelligence contributions (including intrapersonal and interpersonal intelligence) [15] in the 1980s, a notion that, in the end, was disseminated by Goleman [16] in the context of professional performance in the corporate world.

The conceptual outline of the emotional intelligence construct and the different measurement models and tools derived from it have evolved since its formulation by Salovey and Mayer [14] and its refinement by Goleman [17]. These authors have substantiated the driving forces that impel individuals to become entrepreneurs [18]. They outline the entrepreneurial emotional competencies that can be developed [19] as well as the teaching and training processes that foster said competencies [20]. This, in turn, points to entrepreneurial intention as being a precursor to entrepreneurial behavior. Thus, there is an evident relationship between people with higher emotional intelligence who are oriented toward fostering an entrepreneurial mentality [21,22].

The study and measurement of intelligence date back to the first half of the 19th century [23]. At the beginning of the last century, more elaborate instruments for evaluating the intelligence construct emerged, such as the intelligence quotient (IQ) and “g factor” intellectual ability, leading to the concept of “social intelligence” proposed by Thorndike [24], who also formulated the concepts of “abstract intelligence” and “mechanical intelligence”. Thurstone [25] gave a twist to the detection of human abilities, emphasizing the multiplicity of determining factors. In addition, Vygotsky [26] identified two fields of interaction in which individuals relate: interpersonal and intrapersonal. Later, Sternberg [27] found that intelligence is a multifactorial category that is influenced by various contexts, such as socio-cultural ones, among others.

Finally, the concept of “multiple intelligences” was proposed by Gardner [28,29], who stated that human beings are characterized by the presence of seven types of intelligence that shape their behavior and allow them to adapt to their environment through various strategies. At the beginning of this century, Gardner expanded this spectrum of intelligences to nine, based on clinical studies [13]. Despite their conceptual closeness, the connotations of the constructs of “intelligence” and “emotion” are different [30].

Regarding emotions that are manifested through brain functions, these bring together a subjective (cognitive) component and a component associated with corporal and motor responses. It was Ramón and Cajal [31] who posited that emotions are based on the brain. This led to subsequent theoretical and experimental developments in the field of neuroscience and psychology. These developments have confirmed that emotions are a multidimensional concept since there is no limit [32] or single definition to demarcate the processes that constitute them, nor the stimuli that motivate them, stimuli that evoke different conscious reactions (feelings) that then lead to different individual or social–emotional states.

The concept of emotional intelligence was pioneered by Salovey and Mayer [14], and its current meaning dates back to the 1980s and 1990s. Some of the most prominent definitions are the following:

According to Salovey and Mayer [14] (p. 189), emotional intelligence is “the ability to monitor one’s own and others’ feelings and emotions, to discriminate between them, and to use this information to guide one’s action and thought”.

The notion of emotional intelligence has undergone a remarkable evolution throughout the 20th century. Three stages stand out and are identified in Figure 1.

The literature on emotional intelligence and collateral topics has been expanding, encompassing more research areas and countries, albeit in a “dispersed, fragmented” way [37], with its appearance in universities still incipient [38]. In Spain, for example, there has been an abuse of tliterature since the beginning of this century [37,39,40,41,42,43,44]. In Latin America, according to Tarapuez [45], various studies have been undertaken on emotional intelligence and entrepreneurship with university students in Argentina [46], Venezuela [47,48,49], and Peru [50,51], among other countries. In Colombia, these studies have been carried out with teachers [52] and university students [53,54,55,56].

2.2. Scope of Entrepreneurship

One of the relevant topics in business behavior research is the investigation of the motivations that guide entrepreneurs despite the risks involved in deciding to create a company or a business. The question immediately arises as to whether specific skills are required that can result from individual intuition or from emotional skills that can be taught and learned and that the courses taught in university classrooms can positively influence, although not directly, as suggested by Fernandez-Perez et al. [10], and Huezo-Ponce et al. [2].

That is why entrepreneurship has been gaining recognition in the university research agenda since the mid-20th century. It is linked to strategies for economic growth, technological innovation, and social development [57,58], as was anticipated in the 1940s by the economist Joseph Schumpeter [59] when referring to the innovative entrepreneur. Subsequently, Audretsch and Keilbach [60] highlighted the importance of entrepreneurship from the perspective of endogenous economic growth. Hence, entrepreneurship is a source of economic growth and, therefore, job creation, even more so in developing countries such as those in Latin America. This means that emotional competencies and their effects on entrepreneurial intention impact both social and economic sustainability. This is important because, as Bigos and Michalik [61] (p. 1) point out, “entrepreneurship is one of the critical determinants of the economic growth of countries”. For this reason, entrepreneurship is part of the public agenda [62], and universities have become suitable settings for promoting entrepreneurship [2,63].

However, entrepreneurship involves intention and subsequent entrepreneurial behavior, where the emotional component is key, framed in emotional intelligence. This emotional component, the concept of which peaked in the 1990s [14], is linked to individual knowledge, abilities, skills, and abilities on which business success depends. But, since emotional competencies are not inherited but taught and learned, entrepreneurship education [37] emerges as a key source for improving the skills that support entrepreneurial intention and action.

2.3. Models of Emotional Intelligence

Emotional intelligence has gained increasing academic relevance not only in theoretical studies but also in empirical studies, both in the United States and in Europe and, recently, in Latin America. Since the 1990s, two main types of emotional intelligence models have been popularized: mixed models, based on a combination of personal dimensions, and skill or emotional information processing models, proposed by Mayer et al. [64]. These authors understand emotional intelligence as the mix of attributes that characterize an individual (assertiveness, extroversion, impulsiveness, optimism, motivation, happiness, etc.), together with the abilities that they possess (self-regulation, self-knowledge, etc.). The latter alludes to the skills of adaptation to emotions, using available information in favor of cognitive processes.

In addition, emotional competencies within the study of entrepreneurial orientation are considered a practical application, a behavioral approach to emotional intelligence, and, in turn, competencies, the source of entrepreneurship. These competencies are understood as capabilities that can be trainable (taught and learned); thus, education [65] plays a key role.

Goleman [16] differentiates between personal and social competencies and argues that emotional competencies based on emotional intelligence allow for more effective job performance. His competency model consists of five clusters. On the one hand, some are classified into personal competencies, such as self-awareness and those associated with self-awareness, such as self-regulation or self-management. These are also related to emotional control and motivation to achieve objectives and goals. On the other hand, there are social empathy-based competencies that are focused on the emotions of others and the social skills needed to relate to others.

Since the 1980s, various theories about entrepreneurial intention have emerged. These include the Theory of Reasoned Action (TRA), proposed by Ajzen and Fishbein [66]; the Theory of Planned Behavior (TPB), formulated by Ajzen [7,67]; and the Intent Theory, elaborated by Begozzi and Warshaw [68]. Other specific contributions linked to entrepreneurship exist, such as those by Shapero and Sokol [69].

The Theory of Planned Behavior (TPB), proposed by Ajzen in 1991, is a reformulation of the Theory of Reasoned Action, formulated by Ajzen and Fishbein in 1980.

According to Ajzen [7], “The Theory of Planned Behavior (TPB) postulates that human behavior is the result of the intention to perform a specific action, which is influenced by three key factors: the attitude toward the behavior, the subjective norm, and the perceived control over the behavior.” (p. 188).

Lihua [3] argues that TP “establishes that an individual’s behavior is directly influenced by behavioral intention and perceived behavioral control. Behavioral intention is determined by attitude, perceived behavioral control, and subjective norms” (P. 2). This statement is in alignment with what was expressed by Fernandez et al. [10], who postulate that “according to the theory of Planned Behavior (TPB), EI relies on rational factors such as subjective norms (SN), entrepreneurial attitudes (EA), and self-efficacy (SE) as proximal predictors of EI”.

The contrast offered by these theories is that the first one is based on the possibility of predicting (in accordance with a probabilistic model) human behavior from the accumulated beliefs about a given object resulting from experiences. The second, which represents a modified version of the previous one, involves the individual’s ability to influence the outcome of his or her attitude toward behavior (from where control over perceived behavior originates), influenced by beliefs. It also incorporates subjective norms that emphasize the control individuals exercise over their will and that highlight intention as being a precursor of behavior in the face of possible opportunities and available resources.

This theory is superior to the previous one since it establishes that the intention and the resulting behavior occur in a deliberate and controlled way [70]. In the context of the Theory of Planned Behavior, the factors that induce the appearance of intention and behavior are attitude, business self-efficacy, and subjective norms, among the most important determinants.

Although Intent Theory (IT) recognizes the importance of attitudes and subjective norms, it states that their influence is limited as predictors of intentions; therefore, both TRA and TPB are insufficient to predict actual behavior [71]. Its differences from the previous theories are related to its attitude around a central object, be it the attitude toward success, failure, or the process involving value judgments and expectations that will influence the entrepreneurial intention, taking into account past attempts.

The Theory of Planned Behavior, which revaluates the Theory of Reasoned Action, has contributed to a better understanding of long-term, results-oriented business intention [72,73,74]. This approach offers a suitable framework to explain the influence of cognitive factors associated with rational variables in entrepreneurial intention and behavior without discarding the incidence of the emotional component [75], even more applicable when decisions are made in an environment of uncertainty [76]. From these two initial theories, different models have been derived, such as the Entrepreneurial Event Model [69,75], the Entrepreneurial Potential Model [75], the Entrepreneurial Attitude Orientation Model [77], and the Entrepreneurial Intent Determinants Model (Context-Specific Entrepreneurial Intention Model) [78].

The Theory of Planned Behavior (TPB) relates individual intention to intended entrepreneurial behavior as well as entrepreneurial action and expected outcomes to achieve its objectives. In this theory, Ajzen [7] incorporates individual perception and its ability to influence outcomes as a conditional factor for deliberate or unintentional entrepreneurial intention as well as control over perceived behavior.

There is a close relationship between intention and entrepreneurial action, with the former being a precursor to the entrepreneurial behavior from which companies or businesses will be derived [79,80]. However, this would not be possible without considering that the personality of the entrepreneur is essential for emotional intelligence, from which arises emotional competencies that foster entrepreneurial intention and in which learning is essential for significant performance [81].

2.4. Emotional Competencies and Predictors of Entrepreneurial

The influence of emotional competencies—as a behavioral construct of emotional, social, and cognitive intelligence [82]—on entrepreneurial intention is derived from their role as a configuring variable of the latter, which, in turn, is linked to the entrepreneurial orientation and, specifically, to the birth, survival, and expansion of a business [19].

Emotional competencies have become an important research topic [83,84] due to their close links with the phenomenon of entrepreneurship and, therefore, with entrepreneurial intention [85] and behavior [18]. Accordingly, Boyatzis et al. [86] argue that skills can be improved by strengthening emotional competencies. As has been stated in other sections, these competencies can be measured, taught, and learned and can trigger individual business intentions [87].

Emotional competencies (ECs) are derived from emotional intelligence, and the conceptual approaches are diverse because they are associated with multiple individual capacities encompassing both personal and social knowledge and skills that can be taught and learned. In the words of Huezo-Ponce et al. [2] (p. 8), emotional competencies “are the (practical) application of emotional intelligence” in a particular context. Hence, its conceptualization allows various approaches (either as skills, traits, or a combination thereof) that are assimilated to “a learned capacity based on emotional intelligence” [2] (p. 7), from which personal skills (e.g., self-management) and social skills (e.g., empathy) emerge. According to these authors, emotional competencies can be taught and trained through university education.

Gerli et al. [88] recognize the incidence of emotional (and social) competencies in individual propensities toward performance as well as organizational and inter-organizational professional mobility. They propose the notion of “a career without borders” related to global flexibility due to changing work, social, and macroeconomic environments. The authors point to individual professional competencies (why, how, and who) in their professional choice, leaving aside individual behavior oriented toward entrepreneurial intent.

2.4.1. The Influence of Subjective Norms on Self-Efficacy, Entrepreneurial Attitude, and Entrepreneurial Intention

The theoretical model of planned behavior (TPB) [7]—one of the existing approaches, together with the Theory of Reasoned Action, among others—proposes entrepreneurial intention (EI) as a dependent variable and the following as independent variables: emotional competencies (ECs), subjective norms (SNs), entrepreneurial self-efficacy (SE), and entrepreneurial attitude (EA).

It should be noted that the four independent variables are relevant predictors of entrepreneurial intention and behavior in certain controlled circumstances. The influence of the environment on individual actions is unquestionable. Even though a decision is the product of personal criteria, the approval or disapproval of the perceptions of others about said actions is relevant [89]. Positive perceptions and emotional support act as an incentive or a brake to continue or desist from entrepreneurial-oriented activities [90].

However, attention has also been drawn to the prevailing individualism in Western societies [8], when there are, for example, ambivalent relationships between the entrepreneur and his or her family environment, friends, and colleagues, grouped into the so-called “reference or comparison persons or groups”. However, in SNs, internal stimuli predominate, based on the individual’s conviction regarding the opinion of the most influential people in his environment—influencing the decision to start a business—and not external incentives, such as rewards and punishments, as suggested by Fini [89].

Likewise, authors such as Engle et al. [91], as well as Kolvereid and Isaksen [92], have revealed a positive and direct relationship between SN and EI and, in turn, between SN and EA and SE, as revealed by Carr and Sequeira [93]. On the other hand, Fu et al. [94] point out that in the case of vocational school students in a Chinese province, there is no evidence of a positive effect of subjective norms on entrepreneurial intention. These findings are consistent with those of Anderson [95] and Talukder et al. [96] yet contrary to what was proposed by Ajzen [7], Cavazos-Arroyo et al. [97], and Shi et al. [98]. However, Doanh and Bernat [85] and Hongdiyanto [99] concur with Fu et al. [94].

Chin et al. [100] showed that subjective norms had a positive but weak influence on entrepreneurial intention in a study based on a sample of students from a private university in Malaysia, adding that self-efficacy mediated the relationship between subjective norms and entrepreneurial intention. This assessment is shared by Anderson [95] in a study of U.S. college students, who posits that subjective norms, depending on the risk or impact of the environment, are the weakest of the three predictors of entrepreneurial intention included in the Theory of Planned Behavior (TPB). For this author, if an individual has a negative or positive attitude toward entrepreneurial intention, this attitude would enhance or frustrate the relationship between subjective norms and entrepreneurial intention. In conclusion, subjective norms and entrepreneurial intention are mediated by attitude and entrepreneurial self-efficacy.

In contemporary society, social networks may play a more important role in shaping rational motives (subjective norms) than the perceptions of family and friends, thereby directly or indirectly influencing entrepreneurial intention to some extent [89], which could be contrasted as follows:

H1. 

Subjective norms [SNs] directly and positively influence entrepreneurial self-efficacy [SE] (H1a), entrepreneurial attitude [EA] (H1b), and entrepreneurial intention [EI] (H1c) of Colombian university students.

2.4.2. The Influence of Entrepreneurial Attitude and Self-Efficacy on Entrepreneurial Intention

Similarly, the close relationship between attitude and entrepreneurial intention has been established [101] and, sequentially, its direct or indirect impact on individual behavior [102] through the detection of opportunities and the willingness to take risks in the projects undertaken by university students. For this, the sense of self-efficacy, self-perception, and confidence based on their abilities, skills, and abilities to obtain the goals and the desired results are essential.

This construct is the most powerful predictor of entrepreneurial intent, leading to successful entrepreneurial action [9,103]. There is a consensus that personal abilities and individual skills (entrepreneurial self-efficacy) oriented to the desired result can be enhanced through the training provided by the educational system [104].

Similarly, other studies such as those published by Usaci [90], Iakovleva et al. [105], Yurtkorua et al. [106], and Zhang et al. [107] have corroborated the direct and positive influence of EA on EI and, consequently, on entrepreneurial behavior. Authors such as Liñán and Chen [108] consider EA to be the strongest predictor of EI, an assessment that is shared by Kickul et al. [109].

On the other hand, Karimi [9] found that entrepreneurial self-efficacy exerts a significant influence on entrepreneurial intention. This assessment is corroborated by Yurtkorua et al. [106], who stated that entrepreneurial attitude and entrepreneurial self-efficacy have a strong predictive capacity on entrepreneurial intention. Fu et al. [94] argue that entrepreneurial attitude and self-efficacy are the two most significant predictors of entrepreneurial intention, in line with the findings of Seng Te et al. [110], Otache et al. [111], and Wjayati et al. [112].

Wardana et al. [103] conducted a study based on university students in a province in Indonesia, where they found that self-efficacy has a direct effect on entrepreneurial intention, given young people’s confidence in their abilities to create and run a company or business.

With the variables listed, the corresponding hypothesis can be formulated as follows:

H2. 

The entrepreneurial attitude [EA] (H2a) and the entrepreneurial self-efficacy [SE] (H2b) directly and positively influence the entrepreneurial intention [EI] of Colombian university students.

2.4.3. The Influence of Emotional Competencies on Entrepreneurial Intention

As stated, entrepreneurial intention is affected by a range of emotional, cognitive, rational, and functional factors or components [10], one of them is the knowledge and ability to convert said knowledge into general and specific skills [83,113], represented in emotional competence and behavioral approach of the emotional intelligence. Its potential can be increased through educational programs, teaching, and learning in entrepreneurship [114,115] among university students.

Emotional competence is based on the identification and recognition of emotions from which skills and abilities emerge to use emotional expressions to improve individual performance aimed at business success. The conversion of a university student into an entrepreneur is mediated by emotional competencies [22]; so, students with a higher level of emotional competency show a greater entrepreneurial intention, which is an adequate predictor of their entrepreneurial behavior [116].

Bigos et al. [61] define emotional competencies as a combination of three key elements such as knowledge, skills, and attitudes, stressing that there is currently “a research gap” in the analysis of the influence of students’ emotional competencies on entrepreneurial intentions. Moreover, they add that the higher the emotional competencies, the stronger the entrepreneurial intentions will be. Likewise, Chien-Chi et al. [6] found that self-efficacy exerts a mediating role between emotional competence and entrepreneurial intention, an interpretation that coincides with that of Wu and Tian [117].

The verification of these assertions can be made through the following hypothesis:

H3. 

Emotional competencies [ECs] directly and positively influence the entrepreneurial intention [EI] of Colombian university students.

2.4.4. The Influence of Emotional Competencies on Entrepreneurial Attitude

Krueger [118] points out the incidence of ECs in the shaping of IE through EA; however, no major implications of ECs on EA are known. Simultaneously, Gray et al. [119] found that those who assume the role of entrepreneur take action motivated by their emotions. On the other hand, Souitaris et al. [120] argue that emotional intelligence contributes to the strengthening of EA and, therefore, promotes greater entrepreneurial intention.

The entrepreneurial attitude is closely linked to the entrepreneurial intention because it denotes the transition between motivation and the decision to start a business project [121], where having certain emotional attributes and certain emotional competencies acts as an articulator between the entrepreneurial attitude and intention and is key to unleashing a more defined propensity toward entrepreneurship in university students, testing their aversion to the risk involved in undertaking this type of initiative. Contrasting these statements can be performed by formulating the following conjecture:

H4. 

Emotional competencies [ECs] directly and positively influence the entrepreneurial attitude [EA] of Colombian university students.

2.4.5. The Influence of Emotional Competencies on Self-Efficacy

Entrepreneurs have some personal emotional and cognitive traits that allow them to detect opportunities and derive from them a reaffirmation of their leadership and social recognition [122]. Emotional competencies are a source of satisfaction, confidence, creativity, and emotional control in interpersonal interaction of college students, and they motivate them to start businesses or companies.

The connection between emotional competencies and self-efficacy is very evident, given that if there is understanding and conviction about individual abilities, this will motivate university students to put their cognitive resources to the benefit of the demands of the environment [123], demonstrating a greater adaptive and proactive ability, tolerance to stress and solution of difficulties [121], and assuming a pro-entrepreneurial attitude. In this sequence, emotional competencies become a great activator of other entrepreneurial competencies. This approach is summarized as follows:

H5. 

Emotional competencies [ECs] have a direct and positive influence on the entrepreneurial self-efficacy [SE] of Colombian university students.

The above hypotheses are depicted in Figure 2.

Bigos and Michalik, from the perspective of Ajzen’s Theory of Planned Behavior (TPB), sought to validate five research hypotheses in which they contrasted students’ entrepreneurial intentions with self-awareness, self-regulation, self-motivation, empathy, and social skills. The results of this exercise tested the positive influence of the independent variables (the five types of emotional competencies proposed by Goleman) as well as entrepreneurial intentions (a dependent variable), confirming two hypotheses that include self-awareness and self-motivation.

3. Methodology

In recent decades, Structural Equation Models (SEMs) have been used more frequently for data analysis in social research in which there are unobservable (latent) variables. This statistical methodology [124] facilitates the analysis of the relationships between latent variables and their indicators and allows the hypotheses under study to be visually represented.

As Barroso et al. [125] state, systematic analysis with SEM is performed through two approaches: methods based on analysis of covariances (Covariance-Based SEM, CBSEM) and methods based on variance (Partial Least Squares). The difference is that the first method makes estimates of a group of model parameters, such that the theoretical covariance matrix is the closest to the empirical matrix of these [126], while the second [127] calculates the parameters by maximizing the explained variance of the dependent variables (latent and observed).

In the present study, the information obtained by completing 996 surveys was collected and applied in the period between February 2021 and April 2022. The survey was completed by students from 17 public and private universities in Colombia, with 33.7% of those students being from private higher education institutions and 66.3% being from public ones.

Table 1. Universities under study.

University	Periods	Type		Total
		Private	Public
Foundation Juan de Castellanos	2	59		59
Pontifical Javierian University	1	37		37
Antonio Nariño University	1	14		14
University of Antioquia	1		26	26
University of Boyacá	2	67		67
University of Cartagena	2		63	63
University of Amazonia	1		16	16
University of Atlántico	1		15	15
University of Magdalena	2		60	60
University of Valle	2		81	81
Externado de Colombia University	1	12		12
Unilibre	1	19		19
Militar University of Nueva Granada	1		13	13
National University of Colombia	1		37	37
Pedagogical and Technological University of Colombia	2		332	332
St. Thomas University	2	128		128
Technological University of Pereira	1		17	17
Total		336	660	996

Source: Author’s own elaboration.

shows a summary of the composition of the analyzed sample. The universities included in the research represent approximately 34% of the population under study.

The surveys were applied to students of the compulsory subjects of the Business Administration and Public Accounting programs, both of which are connected to entrepreneurial fields. These classes correspond to semesters IV and VIII of the curriculum of each of the programs of study. The population from which the sample was drawn is approximately 3100 students (Appendix A).

The criteria to determine the universities and the student population included being registered in the National Higher Education Information System (SNIES, by its acronym in Spanish) of the Ministry of National Education, offering Business Administration and Public Accounting programs, teaching mandatory courses related to entrepreneurship, and having the endorsement of the university.

The number of surveys answered by the students of the two programs was initially determined by the cooperation of the university as well as the willingness of the students to respond within the stipulated period. The number of surveys conducted was not consistent across the board, as access to certain campuses was more difficult due to pandemic restrictions.

Regarding the demographic information of the respondents, 600 students (60.24%) are female, and 396 are male (39.76%). The age distribution of the respondents was as follows: 5.4% were under 20 years old, 53.4% were between 20 and 22 years old, 21.7% were between 23 and 25 years old, 12.4% were between 26 and 30 years old, and the remaining 7.2% were over 30 years old. With regard to the semester in which they were enrolled, 20% were in a semester lower than the fifth, 55% were between the fifth and seventh semester, and 25% were enrolled in a semester higher than the seventh.

4. Results and Discussion

According to the Theory of Planned Behavior (TPB), entrepreneurial attitude (EA), self-efficacy (SE), and subjective norms (SNs), as cognitive factors, are proximal predictors of the configuration of entrepreneurial intention (EI) [115] in which emotional competencies are decisive [128]. These can be improved with entrepreneurial education [23], deriving a direct or indirect relationship. Thus, the more favorable the attitude and subjective the norm, the greater the perceived control will be and, consequently, the stronger the personal intention to achieve the desired behavior will be [70].

Attitude is associated with the individual will to act in certain favorable or unfavorable situations [72], which, in turn, is influenced by some normative beliefs based on personal and/or family environment, or it is created by the group of reference or even through social networks [72,91,92]. This reinforces self-confidence in the decision to undertake a business project or not, to adopt a certain behavior [10], as well as acquire necessary knowledge and skills [129], which are often enhanced by business training [83].

In the words of Montes [39] (p. 175), “According to the TPB in the context of entrepreneurship, the formation of entrepreneurial intention in general depends on the individual’s own perceived ability to execute the expected behavior (which is expressed through entrepreneurial self-efficacy), attitudes toward the convenience/desirability of an entrepreneurial career (which is expressed through perceived role models), and subjective norm (which is formed through interactions in one’s personal networks)”.

Simultaneously, various studies have found a positive relationship, although with a differentiated effect between these variables, as suggested by [9,22,106,116,121,129], among other authors.

Hence, the constructs included in this study are as follows:

• SE: entrepreneurial self-efficacy;
• SNs: subjective norms;
• ECs: emotional competencies, construct of emotional intelligence;
• EA: entrepreneurial attitude;
• EI: entrepreneurial intention.

Likewise, within the postulates of the Theory of Planned Behavior (TPB), the key variables refer to emotional intelligence (EI), to which are added the social and cognitive ones, as well as their conduct or behavioral approach. Emotional Competencies (ECs), personal and social, can be taught and learned as a practical application of EI through entrepreneurship education and can influence professional and work performance.

The constructs are articulated in a sequence that begins with the predictors of EI, such as SN, EA, and SE, according to the references defined by the entrepreneurial individual. It is worth noting that EC is the key articulator. In this way, SN maintains a relationship between individual behavior and family, professional, and/or friendship environments. In turn, at this level, the emotional support of the reference group, even though it varies according to cultural singularities and the influence of social networks, has called into question its predictive capacity in entrepreneurial intention.

In the same way, the SE reinforces this cycle, given that based on her/his abilities, the entrepreneur perceives and trusts that control over her/his thoughts and actions will lead her/him to obtain the expected results. Added to the aforementioned is that the SE can be trained and strengthened to allow greater detection of opportunities and, in this sense, has a decisive predictive capacity, together with the entrepreneurial attitude, mediated by entrepreneurship education, in the configuration of entrepreneurial intention.

4.1. Preliminary Analysis

The structural model proposed in Figure 2 poses two exogenous latent variables (subjective norms and emotional competencies) and three endogenous latent variables (entrepreneurial attitude, entrepreneurial self-efficacy, and entrepreneurial intention).

To collect the information, it was necessary to fill out the survey in a Google form, which was previously structured and verified through a pilot test. Subsequently, the coordinators and professors of the academic programs of the different universities in Colombia who taught subjects related to entrepreneurship were contacted. This procedure was carried out by using emails obtained from the websites of the institutions and by insisting through phone calls. As a result, 1087 students from 17 higher education institutions (see Table 1) agreed to participate in the application of the surveys, of which 996 were considered valid for the study, while the remaining 91 surveys showed inconsistencies in the information provided.

A first analysis was performed, including the data from the 996 surveys. For this analysis, a verification test of the reliability of data obtained through the survey was carried out. For this, the following were calculated: (a) Cronbach’s α, (b) the KMO index (Kaiser, Meyer, and Olkin index), and (c) the Bartlett sphericity test. In addition, a CFA (Confirmatory Factor Analysis) of one factor was performed with the following findings:

(a)

Factor loadings;

(b)

The ${\chi }^2$ for model fit;

(c)

The Comparative Fit Index (CFI);

(d)

The Tucker–Lewis Index (TLI);

(e)

The p-value of the Root Mean Square Error of Approximation (RMSEA);

(f)

The average variance extracted (AVE).

Each one of the estimated statistics was obtained using the lavaan library (Latent Variable Analysis) from R-studio software 4.3. The results are shown in the following table.

The nomenclature used in the row labeled items corresponds to the statements of the survey questions applied to Colombian university students (see Appendix A).

An analysis of the above table allows us to verify that the constructs included meet the theoretical requirements of dimensionality, reliability, and validity of the scales used. The SN construct (a) presents an RMSEA of 0.465 for the RMSEA estimator, which is statistically greater than 0.05; (b) has the lowest values in the TLI and CFI indices; and (c) has three of its items with a factor loading of less than 0.65. For the construction of the structural equations model, the criteria have been set to take those items that have a factor loading greater than 0.65 (they are marked in blue in

Table 2. Results of the verification test of the reliability of the constructs.

Construct	EI	SN	EA	SE	EC
					Self-Awareness	Self-Regulation	Motivation	Empa Thy	Social Skills
Items	EI06, EI07, EI08, EI09	EI01, EI02, EI03, EI04, EI05	EI10, EI11, EI12	SE01, SE02, SE03, SE04, SE05	GE01, GE02, GE03, GE04	GE05, GE06, GE07, GE08	GE09, GE10, GE11, GE12	GE13, GE14, GE15, GE16, GE17	GE18, GE19, GE20, GE21, GE22
α de Cronbach	0.8826	0.7149	0.9021	0.9208	0.7746	0.7736	0.7986	0.8233	0.8533
KMO	0.8142	0.6285	0.7400	0.8941	0.7645	0.7140	0.7771	0.8339	0.8282
(valor p) [gl]	2411.90 (0.000) [6]	1620.60 (0.000) [10]	1964.52 (0.000) [3]	3535.97 (0.000) [10]	1100.41 (0.000) [6]	1156.70 (0.000) [6]	1197.57 (0.000) [6]	1688.96 (0.000) [10]	2135.89 (0.000) [10]
Factor loadings	EI06 = 0.852 EI07 = 0.688 IE08 = 0.947 EI09 = 0.775	EI01 = 0.707 EI02 = 0.736 EI03 = 0.458 EI04 = 0.603 EI05 = 0.467	EI10 = 0.872 IE11 = 0.928 EI12 = 0.813	SE01 = 0.801 SE02 = 0.797 SE03 = 0.855 SE04 = 0.867 SE05 = 0.860	GE01 = 0.711 GE02 = 0.785 GE03 = 0.705 GE04 = 0.546	GE05 = 0.613 GE06 = 0.671 GE07 = 0.726 GE08 = 0.714	GE09 = 0.745 EG10 = 0.672 GE11 = 0.687 EG12 = 0.724	GE13 = 0.676 GE14 = 0.797 EG15 = 0.753 GE16 = 0.686 EG17 = 0.571	GE18 = 0.728 GE19 = 0.648 GE20 = 0.673 GE21 = 0.786 GE22 = 0.832
CFI	0.999	0.866	0.999	0.993	0.998	0.976	0.996	0.996	0.993
TLI	0.998	0.733	0.998	0.986	0.994	0.929	0.989	0.991	0.986
RMSEA	0.089	0.465	0.000	0.083	0.054	0.103	0.045	0.058	0.086
AVE	0.665	0.795	0.765	0.701	0.541	0.516	0.500	0.532	0.584

Source: Authors of this paper. Criteria with factor loadings above 0.65 are shown in blue.

).

One criterion for checking the validity of the construct is the “average variance extracted” (AVE), which, according to Fornell and Larcker [130], must be equal to or greater than 0.5. For the calculation of the AVE reported in Table 2, only the blue items were used (i.e., those with a factor loading greater than 0.65).

According to Table 1, the information was collected from 17 universities located in different cities in Colombia. Furthermore, the survey was applied on different dates, from the first semester of 2021 to the second semester of 2022. This helped to avoid common method bias (CMB). However, to ensure the absence of multicollinearity [131] between the latent variables that make up each of the constructs, the “variance inflation factor” (VIF) was used. For the information collected in the present study, the highest VIF obtained was 3.2341 and was generated between the latent variables that make up the EI construct, where this value is less than 3.3, which guarantees that there is no collinearity between the variables involved [131].

To verify each of the hypotheses proposed in the study, the structural equation model (SEM) corresponding to the diagram in Figure 2 was adopted.

The variables used are not directly observable since they are considered latent variables and, therefore, are inferred from other variables, which provide the information to have an assessment of the chosen category.

Structural equation models (SEMs) allow us to differentiate between observed and latent variables and, therefore, make it possible to build models in which both types of variables are explicitly defined [132].

Hair et al. [132] propose six stages to develop in an SEM process:

• First step: development of a model based on the theory;
• Second step: construction of a diagram of causal relationships;
• Third step: conversion of the relationship diagram into a set of structural equations and specification of the measurement model;
• Fourth step: selection of the type of input matrix and assessment of the proposed model;
• Fifth step: evaluation of the identification of the structural model;
• Sixth step: evaluation of the fit quality criteria.

In developing the six proposed stages, compliance with the assumptions was verified, and the reliability and validity of the scale used in each construct were determined. Also, a working file was made in Rstudio. For the specification of the model, the LAVAAN (Latent Variable Analysis) library was used, which has the SEM (Structural Equation Modeling) function incorporated, which allows obtaining the estimates of the parameters involved in the model to be contrasted. The method used in the estimation of the parameters is the maximum likelihood (ML).

In the verification of the fit of the model, the values of some indices and some statistics are presented below:

${\chi }_{modelo}^2=1681.556$, with 450 degrees of freedom and p-value = 0.000. This allows us to infer that there is a good fit of the model to the data.

Constructs established in the model explain 70.5% of the variance generated in the EI construct.

The goodness-of-fit index (GFI) = 0.901. The comparative fit index (CFI) = 0.937. The Tucker–Lewis index (TLI) = 0.930. The root mean square of approximation, RMSEA = 0.052. According to Bollen [133], Hooper et al. [134], and Hair et al. [135], the above indicates a good fit of the information from the 996 respondents to the proposed model.

Discussion of the Results

According to Figure 3, it was found that the direct effects of subjective norms (SNs) (0.032) on entrepreneurial intention (EI), whose value is 0.032, are not statistically significant. Therefore, the statement “SN has a direct and positive influence on EI (H1C Hypothesis)” is rejected. Nevertheless, the indirect effect of SN on EI, calculated as 0.430, is significant. This indicates that EA and SE have a notorious predictive capacity in EI, as mentioned by Karmiri et al. and Yurtkorua et al. [9,106]. Moreover, the global influence (direct effect and indirect effect) of SN on EI is 0.462 (0.032 + 0.430). This means that the influence of SN (family, friends, and colleagues) by itself does not affect EI; however, the SN via EA and the SE exert an indirect influence on the EI.

In the case of hypotheses H1a (0.220) and H1b (0.515), it is found that the incidence is significant, i.e., there is a positive dependence relationship of SN with SE and EA. However, SN exerts a greater influence on the EA construct, as suggested by [101,102].

On the other hand, emotional competencies (ECs) also have no direct impacton entrepreneurial intention (EI), since its value is 0.027. This reveals a lack of influence between these two constructs, which is in line with what was stated by Hair et al. [132]. Likewise, the indirect effect of ECs on EI via EA and SE, calculated as 0.091, is not significant. This is contrary to the cases of Spain and Mexico, as well as investigations that served as reference studies. Additionally, the global influence of ECs (direct effect and indirect effect) on EI is 0.118 (0.027 + 0.091), which is not statistically significant. Therefore, hypothesis H3 is not validated.

In hypothesis H5, it is observed that the effect exerted by ECs on SE is significant (0.358), which indicates that university students who perceive themselves to have remarkable emotional and cognitive abilities can detect attractive opportunities in their environment where their confidence and leadership are reaffirmed. This is in addition to their interest in social recognition, which tests their adaptive and proactive capacities (among others) by what was proposed by Welpe et al., Sánchez, and Wong [121,122,123]. Regarding hypothesis H4, which seeks to determine the direct influence between EC and EA, an insignificant effect was found (0.059); therefore, the hypothesis was rejected.

Finally, when validating the hypotheses H2a and H2b, it is observed that the results obtained corroborate the direct and positive influence existing between the EA (0.781) and the SE (0.127) in EI, remarkably coinciding with the findings found in the studies of Spain (EA: 0.75 and SE: 0.19) and Mexico (EA: 0.819 and SE: 0.257). This means that Colombian university students enhance their entrepreneurial intention, taking advantage of their propensity to start a business or activity entrepreneurship as well as their leadership abilities, creativity, and confidence. In turn, it confirms the importance of these two constructs as predictors of entrepreneurial intention [9,106].

It should be noted that the results of the research revealed a significant correlation between EA and EI, which indicates that Colombian university students, once they obtain their professional degree, consider that becoming entrepreneurs will provide them with greater personal satisfaction and will strengthen their work performance, which will result in better status and social recognition.

4.2. Comparative Analysis: Colombia, Spain, and Mexico

Results of the Formulated Hypotheses

Table 3. Estimated parameters of the structural equation models in the studies carried out in Spain, Mexico, and Colombia.

Estimate	Spain 2017	Mexico 2020	Colombia 2022
H1a The subjective norms → The entrepreneurial self-efficacy.	0.380	0.284	0.220
H1b The subjective norms → The entrepreneurial attitude.	0.560	0.540	0.515
H1c The subjective norms → The entrepreneurial intention.	−0.060	0.000	0.032
H2a The entrepreneurial attitude → The entrepreneurial intention.	0.750	0.819	0.781
H2b The entrepreneurial self-efficacy → The entrepreneurial intention.	0.190	0.257	0.127
H3 The emotional competencies → The entrepreneurial intention.	−0.010	0.000	0.027
H4 The emotional competencies → The entrepreneurial attitude.	0.320	0.161	0.059
H5 The emotional competencies → The entrepreneurial self-efficacy.	0.620	0.463	0.358
$R_{model}^2$	0.618	0.832	0.705
$R_{EA}^2$	0.571	0.318	0.283
$R_{SE}^2$	0.738	0.295	0.214
${\chi }_{model}^2$ p-value	593.96 0.000	1782 0.00	1681.5 0.00
GFI	0.980	0.937	0.901
CFI	0.980	0.949	0.937
TLI (NNFI)	0.980	0.942	0.930
RMSEA	0.060	0.044	0.052
Data number	751	1690	996

Source: Authors of this paper.

shows the results obtained in the present study and those found in the studies carried out by Huezo-Ponce et al. [2] and Fernández-Pérez et al. [10]. The constructs used in the structural equation model in the three studies are the same; however, the latent variables used in the elaboration of some constructs may vary, especially in terms of the latent variables that are part of the SN construct, such as the opinion of the family (IE04) and the opinion of friends (IE05) on individual decision-making (see Appendix A: IE04 and IE05).

To deepen the analysis, Figure 4 shows the assessed values for each of the three models that are compared.

In general, if ρ is the estimated correlation, the distribution of ρ is non-central, and therefore, it is necessary to use Fisher’s Z-transformation, which has a normal distribution, and with standard deviation ${\displaystyle \frac{1}{\sqrt{n-3}}}$, where n is the data number. The transformation is defined by $z_{\rho }={\displaystyle \frac{1}{2}}Ln{\displaystyle \frac{1+\rho }{1-\rho }}$. Once the transformations of the data in Table 3 (corresponding to the rows of the eight hypotheses) have been carried out, we proceed to the hypothesis $H_0: {\rho }_i={\rho }_j$ and, as an alternative hypothesis, $H_a: {\rho }_i\ne {\rho }_j$, where $i,j\in \left\{Spain, Mexico, Colombia\right\}$ and the corresponding p-value is determined. The data are reported in

Table 4. p-values for Spain, Mexico, and Colombia.

		Mexico 2020	Colombia 2022
H1a The subjective norms → The entrepreneurial self-efficacy	Spain 2017	0.0288 *	0.0009 **
	Mexico 2020		0.1095
H1b The subjective norms → The entrepreneurial attitude.	Spain 2017	0.6489	0.3526
	Mexico 2020		0.5319
H1c The subjective norms → The entrepreneurial intention.	Spain 2017	0.1719	0.0573
	Mexico 2020		0.4236
H2a The entrepreneurial attitude → The entrepreneurial intention.	Spain 2017	0.1162	0.5219
	Mexico 2020		0.3420
H2b The entrepreneurial self-efficacy → The entrepreneurial intention.	Spain 2017	0.1272	0.1931
	Mexico 2020		0.0011 **
H3 The emotional competencies → The entrepreneurial intention.	Spain 2017	0.8199	0.4447
	Mexico 2020		0.4996
H4 The emotional competencies → The entrepreneurial attitude.	Spain 2017	0.0002 **	0 **
	Mexico 2020		0.0107 *
H5 The emotional competencies → The entrepreneurial self-efficacy.	Spain 2017	0.0494 *	0 **
	Mexico 2020		0.0086 **

Source: Authors of this paper. ** p-value < 0.01, * p-value < 0.05

.

The p-values that are reported without * or without ** mean that the correlational values are statistically equal for levels of α ≤ 0.05 or α ≤ 0.01, respectively. For example, in the case of H1a, it is found that the value obtained in Mexico (0.284) and the value obtained in Colombia (0.220), with a p-value of 0.1095, are considered to be without evidence that they are statistically different.

The three reference studies show a very similar pattern of behavior in terms of the standardized estimates obtained. However, for hypotheses H4 and H5 and for the three countries, there are differences between the correlational values estimated by the model. Between Mexico and Colombia, there is a difference between the correlational values estimated for hypothesis H2b. In the case of hypothesis H1a, between Mexico and Spain, there is a difference between the estimated correlated values, and there is also a difference between Spain and Colombia.

On the other hand, all of the estimates of the parameters obtained in the Mexican research are greater than those obtained in Colombia and Spain, which implies that the Mexican model shows the existence of a greater correlation between each of the constructs of the model. The greatest differences occur in the assessments corresponding to hypotheses H1c (SN→EI), H3 (EC→EI), and H2b (SE→EI). Hypotheses H1c and H3 were rejected in all three studies; in contrast, the H2b hypothesis was validated, which reveals that university students with a higher SE degree have a greater propensity for entrepreneurship.

In addition, when comparing the estimates corresponding to each one of the hypotheses included in the studies of Spain and Colombia, it is observed that they are coincident in H1b, H2a, and H2b. Perhaps the greatest differences are presented in H1a, H4, and H5, in which those of Spain exceed those of Colombia.

It is worth noting that the results for Mexico and Spain indicated an indirect relationship between EC and EI, while in the case of Colombia, this influence was not corroborated.

In the testing of hypotheses H1b and H2a, the three studies report a great similarity in the assessments obtained through the proposed SEM. The standardized estimates reported for H2a in the three models are the closest to 1 (values above 0.75), which implies that there is a high correlation between EA and EI. In the case of H1b, the estimates exceed the value of 0.5 in the three studies, indicating that there is a moderate-to-strong influence of SN on EA.

4.3. A Comparative Analysis of Public vs. Private Universities in Mexico and Colombia

In the case of Colombia (reflected in the information in Table 1), a second analysis was carried out, taking into account the variable “type” of the university. It should be noted that for this comparison, the study from Spain was not included because it did not involve students from private universities.

For each of the groups, a CFA (confirmatory factor analysis) was performed, and it was verified that each of the measurement models was a good fit. The factor loadings of the latent variables in each of the constructs were greater than 0.65, except for the variables EI03 and EI04 in the SN construct. In the case of the public university, the factor loadings were 0.38, 035, and 0.19, respectively, whereas in the case of private universities, the factor loadings were 0.43, 0.45, and 0.31, respectively. The other items that had factor loadings below 0.65 were GE04 (self-awareness construct), GE05 (motivation construct), GE17 (empathy construct), and GE19 (social skills construct). These items were once again excluded from the structural equation model proposed for each type of university (see Appendix A: IE03, IE04, GE04, GE05, GE17, and GE19).

When applying the SEM for each subgroup, the standardized estimators included in Figure 5 were obtained. The values of some adjustment indices and some statistics for the SEM are provided below:

(a)

Public university: ${\chi }^2=1274.565$, degrees of freedom = 450, p-value = 0.00, CFI = 0.934, w0o TLI = 0.927, GFI = 0.889, RMSEA = 0.053;

(b)

Private university: ${\chi }^2=1108.675$, degrees of freedom = 450, p-value = 0.00, CFI = 0.910, TLI = 0.901, GFI = 0.833, RMSEA = 0.066.

The goodness-of-fit indices (GFIs) for the public university are slightly better than those for the private university, which is possibly due to the difference in sample size.

The behavior of the values estimated by the SEM, according to Figure 5, is quite similar in both ecosystems.

Table 5. Estimated parameters of the structural equation models in the studies carried out in Mexico and Colombia (public universities–private universities).

	Estimación
	Public Col	Private Col	Public Mex	Private Mex
H1a The subjective norms → The entrepreneurial self-efficacy	0.153	0.344	0.290	0.290
H1b The subjective norms → The entrepreneurial attitude.	0.458	0.621	0.540	0.540
H1c The subjective norms → The entrepreneurial intention.	0.086	−0.091	0	0
H2a The entrepreneurial attitude → The entrepreneurial intention.	0.793	0.747	0.800	0.800
H2b The entrepreneurial self-efficacy → The entrepreneurial intention.	0.121	0.161	0.230	0.280
H3 The emotional competencies → The entrepreneurial intention.	−0.016	0.049	0	0
H4 The emotional competencies → The entrepreneurial attitude.	0.041	0.011	0.200	0.120
H5 The emotional competencies → The entrepreneurial self-efficacy.	0.457	0.159	0.390	0.530

Source: Authors of this paper.

shows the correlational values, considering the type of university (public or private), obtained in the study conducted by [24] along with those obtained in the present study.

Figure 6 shows the graph of the standardized estimations in Figure 5, together with the estimations obtained by Huezo-Ponce et al. [2].

Taking the estimated correlational values presented in Table 5 and using Fisher’s Z-transformation, we proceed to find the p-value for each null hypothesis and, as an alternative hypothesis, $H_0: {\rho }_i={\rho }_j$, and, as an alternative hypothesis, $H_a: {\rho }_i\ne {\rho }_j$, where $i,j\in \left\{Public Col, Private Col, Public Mex, Private Mex\right\}.$ The values obtained are reported in

Table 6. p-values for Public Col, Private Col, Public Mex, and Private Mex.

		Priv Col	Pub Mex	Priv Mex
H1a The subjective norms → The entrepreneurial self-efficacy	Pub Col	0.0035 **	0.0082 **	0.0068 **
	Priv Col		0.3986	0.3914
	Pub Mex			1
H1b The subjective norms → The entrepreneurial attitude.	Pub Col	0.0127 *	0.114	0.1058
	Priv Col		0.2054	0.1986
	Pub Mex			1
H1c The subjective norms → The entrepreneurial intention.	Pub Col	0.0068 **	0.0974	0.0898
	Priv Col		0.1548	0.1487
	Pub Mex			1
H2a The entrepreneurial attitude → The entrepreneurial intention.	Pub Col	0.4820	0.8926	0.8901
	Priv Col		0.4073	0.4003
	Pub Mex			1
H2b The entrepreneurial self-efficacy → The entrepreneurial intention.	Pub Col	0.5410	0.0356 *	0.0017 **
	Priv Col		0.2807	0.0589
	Pub Mex			0.3055
H3 The emotional competencies → The entrepreneurial intention.	Pub Col	0.6140	0.7578	0.7523
	Priv Col		0.4437	0.4368
	Pub Mex			1
H4 The emotional competencies → The entrepreneurial attitude.	Pub Col	0.0654	0.0518	0.0507
	Priv Col		0.0031 **	0.0836
	Pub Mex			0.1011
H5 The emotional competencies → The entrepreneurial self-efficacy.	Pub Col	0 **	0.1966	0.1499
	Priv Col		0.0003 **	0 **
	Pub Mex			0.0041 **

Source: Authors of this paper. ** p-value < 0.01, * p-value < 0.05.

.

Values that are reported without * or ** indicate that the correlational values are statistically different at the significance levels of α ≥ 0.05 or α ≥ 0.01, respectively. For example, in the case of H1b, the value obtained for Colombian public universities is (${\rho }_{pubCol}=0.458$), and the value obtained for private universities in Colombia is (${\rho }_{privCol}=0.621$), having a p-value of 0.0127. This means that there is no evidence that they are statistically different for significance levels of less than 0.0127 (i.e., statistically ${\rho }_{pubCol}={\rho }_{privCol}$ for $\alpha =0.01$ but statistically ${\rho }_{pubCol}\ne {\rho }_{privCol}$ for $\alpha =0.05$).

Figure 6 and Table 6 reveal that the behavior of the estimates made through the SEM in Mexico is very similar in both public and private universities, and it is in H5 where the greatest discrepancy is found. The results obtained for these ecosystems are presented and interpreted in Huezo-Ponce et al. [9]. Whereas in Colombia, if one considers α = 0.05, in hypotheses H1a, H1b, H1c, and H5, it is observed that statistically, there is a difference between the estimates given for the private and public universities. In hypotheses H1a and H1b, the estimate is higher for the private university, while in hypotheses H1c and H5, the estimate is higher for the public university.

Considering the type of university, the estimates obtained for hypothesis H2a are similar (all of them above 0.74 and all significant for α = 0.01) both in Mexico and Colombia. It can be affirmed that there is a direct effect of EA on EI in the students surveyed, regardless of the country or the type of university, whether public or private. In the case of the Colombian private university, the value of the estimation of the correlation between EC and EA (H4) is not significant.

5. Conclusions

The constructs considered latent exogenous independent variables were the SN and the EC, while the EA and the SE acted as endogenous independent variables; hence, the search was made to establish the relationship of these four variables with the dependent variable EI; therefore, the hypotheses H1a, H1b, and H1c, correlate the SN with the SE, the EA, and the IE, respectively. On the other hand, hypotheses H2a and H2b sought to validate the direct influence exerted by the predictor constructs EA and SE on IE, while hypotheses H3, H4, and H5 captured the influence of CE on IE.

In relation to hypothesis H1c, it was found that SN does not have a direct effect on EI, given that the coefficient is not statistically significant (0.032); therefore, this hypothesis was rejected. This reveals that despite the importance of family, friends, and coworkers, this is not a determining factor in the intention to start a business or company for the students surveyed; in fact, it is the weakest construct in the configuration of EI (as indicated by Fini et al. [89]) and varies among different countries (as indicated by Moriano [102]).

The study also revealed that there is a direct relationship between the SN and EA (0.515), which shows that family and close individuals strongly influence the motivation of students to become entrepreneurs once they complete their university studies. This allows them to acquire a better social status, deepen their work experience, and increase their personal satisfaction, thus validating hypothesis H1b.

On the other hand, there is a moderate relationship between SN and SE (0.22), which reveals that the family and the individuals closest to the student have a lesser influence on their perception of their ability to carry out entrepreneurial processes such as defining a business idea, developing a plan, and establishing relationships with investors, among others, which would confirm hypothesis H1a. The validation of the aforementioned hypotheses (H1a and H1b) is convergent with the findings of Huezo-Ponce, et al. [2], Férnandez-Pérez et al. [10], Engle et al. [91], Kolvereid and Isaksen [92], and Carr and Sequeira [93].

Regarding hypothesis H2a, the study showed that the coefficient estimated by the SEM is the highest (0.781), thus confirming that EA directly influences EI to a greater extent, coinciding with the findings of Usaci [90], Engle et al. [91], Moriano [102], Iakovleva et al. [105], Yurtkorua et al. [106], and Zhang et al. [107]. This result is reinforced by the fact that nearly 80% of the students surveyed stated that their main professional goal is to become an entrepreneur once they finish their university studies. However, entrepreneurial attitude exhibits a weak influence on the preparatory actions for creating their own business, such as obtaining information, conducting preliminary diagnostics, and creating a business plan, among others.

With regard to hypothesis H2b, the obtained value of 0.127 confirms that there is a direct influence of SE on EI, as suggested by Yurtkorua et al. [106]; however, this influence of SE on EI is relatively weak, contrary to what Karimi et al. [9] suggest, likely due to the low capacity of the students studied to prepare a business plan as well as to carry out the pertinent actions of seeking support and advice from lenders and investors, for example. This may be due to the fact that the respondents still maintain their role as students.

In addition, the study revealed that EC does not have a direct impact on EI (0.027), which leads to the rejection of hypothesis H3, i.e., factors such as self-awareness, self-regulation, motivation, empathy, and social skills do not have a direct effect on the entrepreneurial intention of the respondents. These results do not coincide with the findings of Padilla-Meléndez et al. [22] and Lackéus [116]. Likewise, this study particularly highlights that assuming responsibility for personal actions, being receptive to new ideas, striving to achieve the best results, committing to the objectives of a group of interest, being able to understand people’s feelings, and reaching agreements and/or overcoming disagreements with those around them are not related to the propensity to be entrepreneurial or not.

On the other hand, it was observed that the direct effect of EC on the SE construct was significant (0.358), which validates hypothesis H5 and shows that individuals with greater strength in emotional competencies tend to exhibit greater self-confidence and greater control of their environment, as stated by Wong and Law [123]; therefore, university students with a defined emotional and rational perspective will experience an increase in their personal satisfaction and self-confidence, and thus, a greater propensity toward efficiency and productivity, as pointed out by Goleman [81] and Padilla-Meléndez [22]. As such, the study revealed that social skills and empathy have the greatest impact on self-efficacy in the surveyed student population, which is reflected in work performance success, better status, and, consequently, higher recognition and social visibility.

When validating hypothesis H4, a non-significant effect (0.059) is observed between EC and EA, a situation that leads to rejecting this hypothesis. Individual performance oriented to entrepreneurship is characterized by EA, which is derived from emotions and motivations, as pointed out by Gray et al. [119]; therefore, it could be considered that students with higher emotional capacities are more prone to entrepreneurship, which agrees with what was expressed by Souitaris et al. [120]. However, this study indicates that there is a weak dependency relationship between aspects such as the commitment of the students surveyed to the objectives of a group or organization, the interest in the development of others’ personal abilities, the ability to inspire and guide groups as well as to promote and channel the necessary changes, and to their connection with an entrepreneurial attitude.

On the other hand, it was also found that the indirect effect of the SN on EI is significant, with a value of 0.430 (≅0.515 × 0.781 + 0.220 × 0.127). This suggests that the family and social environment of the students influence the decision to undertake, but through EA and SE. In fact, the greatest effect is through EA, which corroborates the findings found in the studies on Spain (0.488) and Mexico (0.515). Likewise, the indirect effect of EC on EI was not very significant since the value obtained was 0.091 (≅0.059 × 0.781 + 0.358 × 0.127), which would indicate that emotional competencies do not influence the decision to undertake, through EA and SE, contrary to what was found in the studies referred to Spain (0.357) and Mexico (0.251).

The measure of the global effects (direct effect + indirect effect) of the SEM for the students surveyed in Colombia in relation to the SN construct on EI is 0.462 (≅0.032 + 0.43), and the value of the EC construct on EI is 0.118 (≅0.027 + 0.091). Regarding the Spanish study, the overall effect of SN on EI was 0.431, while that of Mexico was 0.515, which shows a notable coincidence between the three studies. Likewise, the overall effect of EC on EI was 0.343 and 0.251 for Spain and Mexico, respectively, given that in these two studies (unlike the Colombian case), a direct relationship between EC and EA was evidenced.

From the above, it can be concluded that emotional competencies (ECs), in their different components, indirectly influence entrepreneurial intention (EI) through the mediating role of self-efficacy (SE), as demonstrated by Huezo-Ponce et al. [2], Fernández-Pérez et al. [10], and Chien-Chi et al. [6]; however, the present study did not find a direct and positive influence of entrepreneurial attitude (EA) on entrepreneurial intention (EI).

When contrasting the results of the reference studies (Spain, Mexico, and Colombia), the similarities are evident regarding the relationship between the predictor constructs of entrepreneurial intention, particularly hypotheses H1c (SN→EI) and H3 (EC→EI), the results of which, in the three cases, were not statistically significant. This led to the rejection of these hypotheses, especially in the Spanish study, in which the estimates were negative (−0.060 and −0.010, respectively). Likewise, these studies are similar regarding hypotheses H1b (SN→EA), H2a (EA→EI), and H2b (SE→EI), which showed a high incidence of EA in EI (Spain: 0.75, Mexico: 0.819, and Colombia: 0.781), evidencing a close relationship.

Similarly, discrepancies were found in hypotheses H1a (SN→SE), H4 (EC→EA), and H5 (EC→SE). With respect to hypothesis H1a, the greatest difference is found between the values obtained for Spain (0.38) and Colombia (0.219). On the other hand, in reference to hypothesis H4, the value obtained for Colombia was not significant (0.059), so this hypothesis was not validated, while for Spain, it was significant (0.32). In the case of hypothesis H5, the values were relevant in all three studies, but there were differences among them: Spain (0.62), Mexico (0.463), and Colombia (0.358).

Finally, when taking subpopulations of the public or private segments of Colombian university students, it is observed that SN does not exert a direct influence on EI, but it does indirectly. This is evidenced by the following values: 0.382 (≅0.153 × 0.121 + 0.458 × 0.793) for the public university, 0.519 (≅0.344 ×0.161 + 0.621 × 0.747) for the private university, with 0.430 (≅0.515 × 0.781 + 0.220 × 0.127) being the overall result. Furthermore, for each of the populations, it can be stated that there is a moderate indirect influence of NS on EI through EA and SE.

Entrepreneurship contributes to economic growth and social development [1,2,3,4], so public incentives, such as those announced by the Colombian government in CONPES Document 4011 of 2020 (National Entrepreneurship Policy) and Law 2069 of 2020, are key, especially those aimed at promoting a favorable environment that strengthens the entrepreneurial attitude and intention of university students since entrepreneurship is a source of self-employment and income for young people pursuing higher education.

5.1. Theoretical Contributions

This study broadened the scope of analysis and understanding of the interrelationships between emotional competencies (ECs), entrepreneurial attitude (EA), and self-efficacy (SE) in the context of university entrepreneurship. Moreover, as there is a paucity of research on the significance and influence of emotional competencies in countries like Colombia, the study makes a noteworthy contribution to the extant literature. It endeavors to narrow the knowledge gap by examining the entrepreneurial dynamics of public and private higher education institutions, which exhibit a convergence of behaviors between these two sectors.

In contrast with the findings of previous studies [2,10], the present study revealed that emotional competencies (ECs) do not exert a direct and positive influence on the entrepreneurial attitude (EA) of university students. This suggests that emotional competencies do not influence either directly or indirectly the entrepreneurial intention (EI). Similarly, the study demonstrated that the weakest construct in the configuration of entrepreneurial intention (EI) is the subjective norm (SN), which aligns with findings from prior studies, including those by Fernández-Pérez et al. [10] and Fini et al. [89]. Conversely, other researchers, such as Huezo et al. [2], have proposed that self-efficacy (SE) is the weakest predictor of entrepreneurial intention (EI). It would, therefore, be beneficial to continue to examine the particularities of the university ecosystems in each country in order to identify new theoretical approaches to the application of the proposed model.

5.2. Practical Contributions

The increasing prevalence of entrepreneurial intentions among young Colombian university students is a notable phenomenon, particularly in light of the elevated rates of unemployment experienced by this demographic. It serves as a conduit for enhancing employability and income prospects. Therefore, research such as this makes a valuable contribution to the search for educational strategies that should be integrated into the university curriculum and focused on programs and courses on entrepreneurship. This will facilitate the stimulation of emotional skills, reinforcement of self-efficacy, and strengthening of the entrepreneurial attitude, which, in turn, enhances entrepreneurial intentions and benefits the most vulnerable sectors of a developing economy, consequently reducing poverty and inequality. To this end, it is imperative to define and implement joint initiatives between higher education institutions, backed by academic communities, in collaboration with their respective entrepreneurship networks and the relevant governmental bodies responsible for fostering young entrepreneurship, whose regulatory framework was adopted in Colombia at the beginning of this decade.

5.3. Limitations and Research Possibilities

This study has some restrictions that must be overcome, such as (1) the selection of a larger sample, which includes a sizable number of public and private universities, as well as students from academic programs other than economics; (2) the inclusion, for international comparison, of investigations carried out in other countries representative of the research in the constructs related to the entire thematic field of emotional intelligence; and (3) the contrasting and use of the conclusions obtained in this research to actively promote entrepreneurship in the universities where the surveys were applied.

The possibilities in the field of research on emotional competencies are enormous, particularly in the impact of entrepreneurship education and the convenience of strengthening emotional competencies to reinforce self-efficacy and an entrepreneurial attitude among university students in such a way that the intention to launch business projects could be consolidated through the adoption of training strategies (innovative teaching methodologies and reforms to the study plans) that improve individual cognitive processes and stimulate the appearance of successful entrepreneurs.

In addition, it is necessary to involve many more entrepreneurial colleagues to explore new lines of research that address the influence of emotional competencies from different theoretical perspectives and in various contexts, which benefit social development and sustainability in an increasingly globalized and uncertain world.