College environment and entrepreneurial intention in high school

Purpose – There is no consensus on the most effective way to foster entrepreneurship in educational institutions, and educational policies on entrepreneurial activity differ significantly amidst organizations and contexts. Thus, the objective of this research is to analyze influence of the college environment and entrepreneurial characteristics on the entrepreneurial intention of Brazilian high school/technical students. Design/methodology/approach – The empirical research used partial least squares structural equation modeling (PLS-SEM) and a sample of 384 students of technical courses, such as Administration, Systems Development, Chemistry, Secretariat, among others. Findings – The proposed model was validated, and the hypotheses were confirmed, proving suitable for high school/technical education. Assessing the high school environment with this model can help determine each organization ’ s strengths and weaknesses and, indeed, the impacts on the ecosystems in which it operates. The results of the multi-group analysis indicate differences concerning the courses as well. Research limitations/implications – The main limitations involve non-probabilistic sampling procedures and the collection having been carried out with a single cross-section. Practical implications – For managers and teachers, this article presents indicators to qualify the activities of the educational environment, considering teaching activities, extracurricular activities, fairs, actions of teachers and students, among other initiatives. Social implications – The article contributes to high school managers, particularly in technical schools, so that they understand the factors that influence the profile and entrepreneurial intention of students; in other words, something that can impact the lives of thousands of students, teachers and the community itself. Originality/value – This research presents a novel analysis of the antecedents that drive student entrepreneurship in an underexplored educational context in a developing country. The results show the necessary conditions for technical schools to foster entrepreneurial activity, feeding innovation ecosystems with entrepreneurial talent.


Introduction
According to the academic literature, it is possible to identify studies that report education systems around the world applying the teaching of entrepreneurship at the secondary level (Pommer Barbosa, Silva, Gonçalves, & Morais, 2020;Tentama & Abdussalam, 2020).
When dealing with this subject, it is possible to find two strands: the author Hickie (2011) affirm that the adolescent who has contact with entrepreneurship in their high school education thinks about becoming the owner of their own business; as for Paço et al. (2011), they argue that this experience helps young people to develop specific characteristics by providing skills they will need when reaching adulthood regardless of whether they will continue their entrepreneurial career or not.
Small and medium enterprises (SMEs) account for 62% of total employment and 50% of national added value in the Brazilian context. These values are slightly below the OECD averages of 70 and 55%, respectively. Brazil's GDP growth rate in 2017 was 1.32%, and in 2018 there was a drop of 0.01%, while in 2019, Brazil's GDP growth rate decreased by 0.18p.p. vs 2017, showing a rate of 1.14% (World Bank, 2020). This slight reduction in indicators may point to a reduced capacity of the economy to generate sufficient jobs, improve the quality of employment in current jobs, and share the benefits of economic growth (Fragoso, Rocha-Junior, & Xavier, 2020). Brazil still registers high unemployment rates in the adult population (11.9% in 2019) and particularly among young people (27.8% in 2019) (ILO, 2020). In this context, entrepreneurship plays a key role, as it supports innovation and establishes small and medium-sized companies (Coulibaly, Erbao, & Mekongcho, 2018) as being a possible economic solution, and as it has the potential to allow individualsespecially youngto acquire skills and establish their jobs, contributing to the reduction of the unemployment rate in the country (Coulibaly et al., 2018). For this to happen, a fundamental step is developing entrepreneurial intentions and behaviors, as this process represents the first phase in the effective creation of a business, even in long-term perspectives (Fayolle & Liñ an, 2014).
Based on these arguments, more sophisticated investigations into the effectiveness of educational actions in high schools for entrepreneurial activity are necessary (Pommer Barbosa et al., 2020;Tentama & Abdussalam, 2020). Therefore, the contribution of this research is to fill this gap on the subject at the high school level and measure the influence that entrepreneurial characteristics and the school environment have on the entrepreneurial intention of students. The practical contribution is to bring managers' and teachers' information about the students' level of interest, their profiles and how the environment influences them towards entrepreneurship.
This motivates the following question: to what extent can entrepreneurial characteristics and school environment influence the entrepreneurial intention of high school/technical students?
To carry out this research on the influence of entrepreneurial characteristics and school environment in entrepreneurial intention of high school/technical students, we study two State technical schools (Etecs): Lauro Gomes and Juscelino Kubitschek de Oliveira, from Centro Paula Souza (CPS), an autarchy of the government of the State of São Paulo composed of Etecs and Technology Colleges (Fatecs) created in 1969 with the objective of "promoting professional and technological public education within benchmarks of excellence, aiming at the technological, economic and social development of the State of São Paulo" (Souza, 2015).
In addition to bringing training aimed at the job market, CPS promotes entrepreneurship, taking it as one of its strategic objectives (Souza, 2015(Souza, , 2017. Through fairs, events, projects, actions, covenants and partnerships, CPS introduces the culture of entrepreneurship into its institutions.
The article is structured as follows. After these introductory arguments, Section 2 articulates the literature review. From this literature review, hypotheses were formulated. Section 3 presents the research methodology and data collection procedures. Section 4 reports the description and analysis of empirical results. Section 5 presents the discussion and conclusions with final remarks, implications and avenues for future research.
2. Literature review 2.1 Middle-level entrepreneurship and the school environment The entrepreneurship literature has focused on higher education institutions, and research on entrepreneurship in the high school environment is still embryonic. This is an essential gap in the entrepreneurship literature (Peterman & Kennedy, 2003;Pommer Barbosa et al., 2020;Tentama & Abdussalam, 2020). This level of education is the ideal place to foster entrepreneurship, and this lack of research should not exist (Xu, Ni, & Ye, 2016).
In the national context, some surveys were found: one of them aims to investigate the competencies that make up Entrepreneurial Education. However, it addresses the three levels of education: elementary, secondary and higher (Zambon, 2014); the second study elaborates a proposal for the insertion of an entrepreneurship discipline in a high school institution (Mayer, 2001); another one used the same research object and set out to investigate whether professional education develops entrepreneurial skillsthe analysis was restricted to alumni (Pereira, 2018); and the last, Pommer Barbosa et al. (2020), identified that entrepreneurial education acts on personality traits and positively impacts the entrepreneurial intention of young elementary and high school students, especially in the stages of entrepreneurial development and training.
On international level, research by Paço et al. (2011), Peterman andKennedy (2003) and Xu et al. (2016) state the relevance of studying this topic in high school. The authors Filion (1994); Kourilsky and Walstad (1998) declare that the phases of childhood and adolescence are often identified as the preferred periods for the development of positive attitudes towards entrepreneurship and the acquisition of basic knowledge on the topic. Tentama and Abdussalam (2020) proved the relationship between the internal locus of control and entrepreneurial intention in vocational high school students, demonstrating the importance of understanding the entrepreneurial behavior of this student population.
National and international academic literature on entrepreneurship predominantly focuses on undergraduate and graduate courses. Content and activities are worked so that the student links entrepreneurship with the creation of new businesses (Dana et al., 2021;Fayolle & Liñ an, 2014;Kuratko, 2005;Moraes, Iizuka et al., 2021;Pelegrini & Moraes, 2022;Rocha & Freitas, 2014).
Research on entrepreneurship in high school addresses the issue through two strands: one of them in a playful, balanced and even subjective way so that the experience with entrepreneurship contributes positively, developing characteristics and skills that will help one in any context of life (Paço et al., 2011); the other strand works on entrepreneurship in a more traditional and applied way, in which the environment will provide technical knowledge through workshops, projects and entrepreneurial activities, not with the same intensity performed at the higher level, the approach at the medium/technical level has the purpose of preparing young people for the professional market (Hickie, 2011). To exemplify this reality, Etec Antonio Devisate in the City of Mar ılia created the Business Plan in Practice project in 2011, whose purpose was to improve student learning through the creation of a fictitious company. The challenge was to carry out a business plan for a donut company and present it to the entire school community. In some classes, students chose their theme and often linked it to personal goals. In testimonials, students told how the experience was, raising some aspects such as a much more comprehensive notion of starting a business, something that was considered complex to make easy, and a vision of how to create and run a business (Iizuka, 2015).
In the particular case of the studied environment, Etecs propose teaching and learning related to the second strand of entrepreneurship: applied and classic, thus meeting the objective of this research, which is to contribute to the advancement of knowledge concerning the entrepreneurial intention of high school/technical students.

Teaching and learning measured by entrepreneurial characteristics
The way we teach and learn is undergoing substantial changes. Covid-19 pandemic accelerated these changes, with the migration of the educational environment to collaborative learning environments and high technology use (Hodges, Moore, Lockee, Trust, & Bond, 2020). This abrupt change in how education is delivered demonstrated how much the learning experience needed to be enhanced for a new context (Garcia-Morales, Garrido-Moreno, & Martin-Rojas, 2021). Online emergency remote education has been an experience in all areas (Zimmerman, 2020).
In this context, the relevance of exploring all educational contexts grows. Entrepreneurship education in high school receives little credit. And considering the national and international literature surveyed, we did not find similar research on the high school educational environment. A lack of research makes it difficult to identify measurement mechanisms to assess which skills and competencies this teaching can promote. We adopted a model tested and validated in higher education to investigate high school context, developed by Moraes et al. (2018). The model presents validated indicators for entrepreneurial characteristics and educational environment constructs. Model adaptations have already been used in several recent studies (e.g. Campos et al., 2021;Rocha et al., 2021;Moraes, Fischer et al., 2021;Moraes, Iizuka et al., 2021). In addition, the model proposed by the authors is the closest to the reality studied by this research.
The entrepreneurial characteristics used in the structural model to measure entrepreneurial intention were: self-efficacy, sociability, planning, leadership, innovation and taking calculated risks (Table 1).
Therefore, this research proposed to test whether entrepreneurial characteristics such as sociability, planning, leadership, innovation and taking calculated risks, positively influence self-efficacy; and if self-efficacy and the school environment positively influence the intention of entrepreneurship of high school/technical students, following the structural model of the authors Moraes et al. (2018) (see Table 1). In this case, the research was also willing to test whether the variable course arises any difference between the relationships of all latent variables. The tested hypotheses are shown in Table 2. Figure 1 presents the research conceptual model.

Methodology
The quantitative method will develop the research using multivariate data analysis. The analysis was performed using Partial Least Square Structural Equation Modeling (PLS-SEM). The reasons for using PLS-SEM are: the research is concerned with testing a theoretical framework from a forecasting perspective; the survey features multi-group analysis; the model is complex (Hair, Hult, Ringle, & Sarstedt, 2022).

Construct Description Reference
Self-efficacy People who are able to achieve the desired results. In the context of entrepreneurship, self-efficacy is regarded as the strength that a person has in believing that they are able to successfully perform various roles and tasks to open and manage their own business Bandura (1986) (2014) Source(s): Adapted by the authors from Moraes et al. (2018) Hypotheses description

H1
Sociability positively influences self-efficacy H2 Planning ability positively influences self-efficacy H3 Leadership ability positively influences self-efficacy H4 The capacity for innovation positively influences self-efficacy H5 Taking calculated risks positively influences self-efficacy H6 Self-efficacy positively influences entrepreneurial intention H7 The school environment positively influences entrepreneurial intention H8a There is a difference in the relationship between sociability and self-efficacy with regards to the course H8b There is a difference in the relationship between planning and self-efficacy with regards to the course H8c There is a difference in the relationship of leadership and self-efficacy with regards to the course H8d There is a difference in the relationship between innovation and self-efficacy with regards to the course H8e There is a difference in the relationship between take risks and self-efficacy with regards to the course H8f There is a difference in the relationship between self-efficacy and entrepreneurial intention with regards to the course H8g There is a difference in the relationship between school environment and entrepreneurial intention with regards to the course The questionnaire and the model were created for research involving higher education students; however, there was a need for adaptation to the target audience of this research. We consulted two Etec teachers who teach entrepreneurship classes to verify the instrument. The teachers stated that the questions were in a structure in which students would understand the questionnaire. Therefore, a pre-test was conducted with ten high school students integrated into the technical Administration course. The result was positive; the students had no difficulties understanding the instrument (see Table 3).
To carry out this research, the final sample was 133 students from the Technical Course in Administration and 251 students from other courses, totaling 384 respondents. Data collection took place in February and March, 2019, with the help of teachers and school unit coordinators. The other courses in the sample were: (1) Regular High School; (2) Technician in Logistics, Systems Development and Secretariat and (3) Integrated Technical Education for Mechatronics, Industrial Automation and Chemistry.
More than half of the students in the sample are between 15 and 17 years old (73%), the rest of the sample was made up of students aged between 18 and 21 years old (18%), 22 to 25 years old (5%) and over 26 years old (4%). Regarding gender, students are homogeneously distributed, 51% are men and 49% are women.
The evaluation of the sample and the statistical power of the analysis was performed using the G*Power 3.1 software and recommendations by Hair et al. (2022). To perform the calculation, the largest number of arrows that reach a latent construct is considered. In the case of this study, there are five (greater number of predictors). With a significance level of 5%, the statistical power of 0.8 and average effect size (f 2 5 0.15, which is equivalent to R 2 5 13%), it is known that the minimum sample size is 92. With the analysis of multi-groups double this value is expected. As the sample used was 384 students, it is suitable for estimation by PLS-PM. For the average effect size, the power is 0.99, which is well above the 0.8 recommended by Hair et al. (2022).

Description and analysis of results
The model presents only reflective indicators: Self-efficacy (SE), School Environment ( (LE) and Sociability (SO). Therefore, the criteria used for the analyses were: internal consistency (composite reliability); reliability of the indicator; convergent validity (average variance extracted) and discriminant validity (Hair et al., 2022). The first criterion analyzed was convergent validity, respecting the levels indicated for latent variables, referring that, indicators with loadings greater than 0.7 are significant. In contrast, loadings between 0.4 and 0.7 should be evaluated, as the removal of these indicators can affect the variance of the construct. Finally, loadings lower than 0.4 are recommended to be removed from the research (Hair et al., 2022). Thus, the indicators with loadings between 0.4 and 0.7 were analyzed and some were excluded.
Another indicator necessary for model validation is the Average Variance Extracted (AVE). This indicator represents the average value of the square factor loadings of the indicators associated with the construct, whose value must be greater than 0.5. That is, an AVE of 0.5 or higher indicates that, on average, the construct explains more than half of the variance of its indicators (Hair et al., 2022).
In analyzing the internal consistency of the model, we have two other criteria: Composite Reliability and Cronbach's Alpha. The first depicts the degree to which the indicators reflect the latent construct in common; the second provides an estimate of reliability based on the correlations of the observed indicator variables, assuming that all indicators are equally reliable. Both indicators work at the same level for validation, ranging from 0 to 1, in which the values closer to 1 show that the construct will be more internally consistent. Values above 0.7 are the most satisfactory; however, for exploratory research values between 0.6 and 0.7 are considered acceptable (Hair et al., 2022).
The purpose of discriminant validity is to show that the measure of one construct is different from the other. That is, each construct is unique and captures the phenomenon not represented by other constructs in the model (Hair et al., 2022). The indicators mentioned are presented in Table 4, and all are within the parameters established by Hair et al. (2022).
Bootstrapping technique was used for the evaluation of the structural model. This technique aims to analyze the significance of the indicators by means of the factor loadings, estimating the model by means of parameter estimates and their confidence intervals (Hair et al., 2022). With this information, Student's t-test is calculated to assess the significance of each indicator weight, starting from the assumption that the correlation coefficients are equal to zero if the test results are greater than 1.96, indicating that the correlation is significant (Hair et al., 2022).  (2014)  EN8 The school environment taught me to make decisions when I face a problem, improving my ability to take calculated risks Note(s): Likert scale responses from 1 (totally disagree) to 5 (totally agree). The students responded how much they agreed with the statements Table 3. Table 5 shows that all relationships obtained Student's t-test values above 1.96, with a significance level of 5%.

REGE
Other measures used to evaluate the structural model are the determination coefficient (R 2 ) and the Q 2 . R 2 measures how much the model has predictive accuracy, with values ranging from 0 to 1. The R 2 values of the self-efficacy (0.391) and entrepreneurial intention (0.364) construct are both considered high, both considered high values.
The Q 2 indicator represents the predictive relevance of the model, values of Q 2 greater than 0 indicate the predictive relevance of the model, the Q 2 values of self-efficacy construct (0.183) and entrepreneurial intention (0.260) are within their respective parameters.
In the multi-group analysis we observe the relationship of the course variable in relation to the other latent variables, and we verify whether there are differences in the technical course in Administration in relation to the other courses. Values below 0.05 and above 0.95 show that differences are significant (Hair et al., 2022). The results in Table 6 show that there are significant differences between leadership and self-efficacy and between sociability and self-  Table 6. Multi-group analysis influencing high school efficacy relationships. In this case, leadership and sociability in relation to the self-efficacy of students in the technical course in Administration are more intense than that of students of other courses. Thus, Table 7 shows the results of the hypotheses tests of this study. With this, it is possible to identify that the school environment is the construct that most influences entrepreneurial intention. This demonstrates that the high school student/ technician, independent of the course in which they are inserted, develops an entrepreneurial behavior through the school environment. This result confirms what the authors Kourilsky and Walstad (1998) say, that entrepreneurship education in high school conducted in a supportive environment will increase students' interest in the subject as well as their preparation.
The self-efficacy construct was also validated. That is, the more the student believes that he/she is able to successfully perform various roles such as finding creative solutions, working productively and seeing new market opportunities, the greater their entrepreneurial intention.
Regarding the constructs that influence self-efficacy, what stood out the most was innovation. The more students feel creative and in search of news, the greater their self-efficacy. The second construct that most influenced self-efficacy was sociability. The larger the network of contacts and relationships, the more self-effective that student will be.
The leadership construct was also validated: the greater the student's influence on people, so that they can overcome conflicts and work as a team, the greater their self-efficacy. Then follows risk-taking construct, according to which the greater the willingness to take risks, leave the comfort zone, believe that this will bring advantages and opportunities, the greater the student's self-efficacy.
Finally, the planning construct, which refers to being able to anticipate events and have a future vision: in this case the greater the preparation of that student for the future, the more self-effective they will be.
When performing the analysis of the result of multi-groups, it was evident that there is a difference when it comes to the technical course in Administration compared to other groups.

H1
Sociability positively influences self-efficacy Confirmed H2 Planning ability positively influences self-efficacy Confirmed H3 Leadership ability positively influences self-efficacy Confirmed H4 The capacity for innovation positively influences self-efficacy Confirmed H5 Taking calculated risks positively influences self-efficacy Confirmed H6 Self-efficacy positively influences entrepreneurial intention Confirmed H7 The school environment positively influences entrepreneurial intention Confirmed H8a There is a difference in the relationship between sociability and self-efficacy with regards to the course Confirmed H8b There is a difference in the relationship between planning and self-efficacy with regards to the course Not confirmed H8c There is a difference in the relationship of leadership and self-efficacy with regards to the course Confirmed H8d There is a difference in the relationship between innovation and self-efficacy with regards to the course Not confirmed H8e There is a difference in the relationship between take risks and self-efficacy with regards to the course Not confirmed H8f There is a difference in the relationship between self-efficacy and entrepreneurial intention with regards to the course Not confirmed H8g There is a difference in the relationship between school environment and entrepreneurial intention with regards to the course Not confirmed Table 7.
Hypothesis test result

REGE
The contrast lays on relationships between leadership and self-efficacy and sociability and self-efficacy. The most significant difference was in the leadership indicator: students in the Administration course perceive themselves to have this more prominent feature than students in other courses. This can happen because one of the characteristics of the Administration professional is leadership, and the course also works through its basis and activities so that it can be awakened in students, or the student looking for this course already has predisposition or profile directed to this characteristic.
Regarding sociability, what may have happened is that the sample had courses such as chemistry, mechatronics, industrial automation and systems development, in which the student profile is of low social interaction.

Discussion
When teaching entrepreneurship in high school, one realizes that there are two approaches: the first is more playfully and subjectively, and the second is a more applied way, aimed at more traditional entrepreneurship. Secondary/technical education is more characterized by the second approach, as its objective is to train students for the professional market.
Through a model that proved to be appropriate for higher education and after an adaptation for high/technical school, it was possible to present the efficient results of this research, showing that the model was adequate for the study since national and international academic literature is limited and is at an early stage of research. We believe that the choice of Etecs and technical education students favored this result since such educational institutions, like colleges and universities, seek to prepare students for the world of labor.
The model's hypotheses regarding entrepreneurial characteristics have all been confirmed, and for sample quantity are consistent, reliable and statistically valid results.
In the multi-group analysis, we observed that the relationship of the constructs of leadership and sociability concerning self-efficacy indicated a difference in terms of courses. This result opens a space for research and analysis, as it is indicated that a more in-depth study should be carried out to understand; what are the factors that led this differentiation of the technical course in Administration compared to other courses, questioning the reason why these students had these characteristics more evident. For this, the sample should be expanded and tested in other technical courses and other Etecs.
The hypotheses that were not confirmed in the multi-group analysis indicate that the other entrepreneurial characteristics do not present differences. That is, regardless of the course, the students have the same perception; this must have probably occurred due to the environment being the same, the student's profile having socio-economic similarities, among other factors, something that can be tested with different student samples.
Regarding the practical contribution of the research, it can be certain that the two Etecs environment has influenced their students' entrepreneurial intention. For the managers and teachers of these institutions, this research is an indication that teaching, extracurricular activities, fairs, actions by teachers and students, among other initiatives, have had a positive effect (Paço et al., 2011).
Thus, the assessment of the high school environment with this model can help determine the strengths and weaknesses of individual organizations and, in fact, their impacts on the ecosystems in which they are inserted. The relational skills developed by high schools to connect their student bodies to the broader context of ecosystems stand out as a key factor in entrepreneurial behavior . Consequently, our results shed light on the mechanisms through which such connections can feed the productive structure with entrepreneurial talent. Thus, the model can be replicated in other Etecs to verify how much their institutions may be influencing entrepreneurial intention, as well as verifying whether the result achieved in these two schools is confirmed or not.

Conclusion and final remarks
Forming entrepreneurial behavior and encouraging entrepreneurship is essential for effective business creation. Our research helps fill a research gap on the relationships and effectiveness of educational actions for entrepreneurship in high school. This research presents a new analysis of the antecedents that drive student entrepreneurship in an underexplored educational context in a developing country. The results show the necessary conditions for technical schools to foster entrepreneurial activity, feeding innovation ecosystems with entrepreneurial talent. For managers and teachers, this article presents indicators to qualify the activities of the educational environment, considering teaching activities, extracurricular activities, fairs, actions of teachers and students, among other initiatives.
The research has limitations to be highlighted: the sample used was non-probabilistic, with the perception of only one Brazilian high school context; the collection was carried out in a single cross-section; the use of a model developed for the university context.
Thus, some suggestions for future research can be presented. At the national level, it is possible to reproduce it in other technical courses from the same education network and technical courses from other institutions, such as federal educational institutes (IFES); in regions and states, such as rural technical schools, analyzing whether the environment also impacts entrepreneurial intention. It is possible to test it in several countries where high school/ technical education is part of youth training at the international level. Another possibility is to conduct longitudinal research to see how students' behavior changes over time.