Determinants of Sustainable Entrepreneurial Intention: A Multigroup Analysis Between Public and Private Universities in Industrial Engineering

Abstract

Entrepreneurship plays a fundamental role in sustainable economic development, particularly in the field of industrial engineering. This study analyzes the determinants of sustainable entrepreneurial intention among students from public and private universities in Peru. A quantitative, non-experimental, and cross-sectional methodology was employed, using structural equation modeling (SEM) to examine the causal relationships between entrepreneurial intention and its key determinants: attitude, subjective aspects, perceived behavioral control, advantages, and obstacles. The sample consisted of 200 students from public and private universities. The results indicate that attitude towards entrepreneurship and perceived advantages are significant predictors of entrepreneurial intention in both university types. However, perceived behavioral control showed a significant effect only among private university students, suggesting that institutional resources may influence entrepreneurial self-efficacy. The study also identified differences in the perception of advantages and obstacles between the two groups. The findings underscore the need for tailored educational interventions to foster entrepreneurship, considering institutional contexts. These results contribute to understanding how sustainability-driven entrepreneurship education can enhance entrepreneurial potential, aligning with Sustainable Development Goals (SDGs) related to quality education (SDG 4), economic growth (SDG 8), and innovation (SDG 9).

1. Introduction

Entrepreneurship has become a fundamental pillar for sustainable economic and social development at a global level. In the current international context, the training of new entrepreneurs in universities represents a crucial challenge, especially in the field of industrial engineering. Recent studies indicate that approximately 40% of university students in Latin America express entrepreneurial intention, although this figure varies significantly between countries and types of educational institutions, according to ref. [1] in their study on entrepreneurial intention in Latin American university students. Advances in the understanding of entrepreneurial intention have been remarkable in the last decade. The scientific literature has evolved from purely economic approaches to more holistic perspectives that incorporate environmental, social, and sustainability factors. Research by ref. [2] has shown that entrepreneurship education in higher education institutions can increase the likelihood of students starting their own ventures by up to 35%. However, significant challenges remain in entrepreneurship education, particularly in the Latin American context. The gap between public and private education, the limitations in specialized pedagogical resources, and the need to integrate sustainability perspectives in entrepreneurship training continue to be relevant obstacles, as documented by ref. [3] in their systematic review on entrepreneurial intention in university students. This research is structured in several fundamental categories. First, with respect to entrepreneurial training in engineering, ref. [4] have demonstrated the critical importance of soft skills in the development of entrepreneurial intention. Regarding institutional factors, the research of ref. [5] shows significant differences in entrepreneurial intention between students from public and private institutions. In the area of sustainability and entrepreneurship, ref. [6] have identified a positive correlation between environmental awareness and entrepreneurial intention in university students. Finally, in relation to entrepreneurial self-efficacy, ref. [7] point out the relevance of this factor as a key mediator in the development of entrepreneurial intentions.

The general objective of this study is to analyze the differences in entrepreneurial intention and its determinants among engineering students from public and private universities in Peru. The specific objectives are: to identify the determinants of entrepreneurial intention in industrial engineering students considering the dimensions of attitude, subjective aspects, and perceived behavioral control; to compare the levels of entrepreneurial intention among students of public and private universities, considering their sociodemographic and academic characteristics; to evaluate the influence of sociodemographic and academic factors on the entrepreneurial intention of industrial engineering students; to analyze the mediating effect of perceived behavioral control on the relationship between institutional support and entrepreneurial intention; to identify the main advantages and obstacles perceived in the development of entrepreneurial intention among students from public and private universities; and to determine the causal relationships between the dimensions of attitude, subjective aspects, perceived behavioral control, and entrepreneurial intention by means of a structural equation model. The justification for this study lies in its contribution to the understanding of how institutional differences influence the formation of future sustainable entrepreneurs in the field of industrial engineering. This research is particularly relevant in the current context, where sustainability and entrepreneurship have become imperatives for economic development. The gap in the literature that this study aims to address centers on the paucity of research that specifically examines differences in entrepreneurial intention between public and private institutions in the context of industrial engineering in Peru, while also considering the sustainability perspective. This research contributes significantly to three key Sustainable Development Goals: it strengthens SDG 4 by promoting quality entrepreneurial education in engineering, boosts SDG 8 by training professionals capable of generating jobs and sustainable economic growth, and enhances SDG 9 by fostering industrial innovation through new ventures. The intersection of these goals is essential to develop a new generation of industrial engineers who will not only create successful companies but also implement sustainable business practices that benefit all of Peruvian society.

The novelty and significance of this study lie in its being the first to examine the determinants of sustainable entrepreneurial intention specifically in industrial engineering students in Peru, considering the contextual differences between public and private universities. Despite the existence of studies on entrepreneurial intention in Latin America, such as those conducted by Leiva et al. [1] and Cáceres-Cayllahua et al. [5], the literature presents a significant gap regarding: (1) the specific focus on industrial engineering, a crucial field for sustainable innovation in developing countries; (2) the comparative analysis between public and private institutions in the Peruvian context, where there are marked differences in resources, institutional support, and entrepreneurial ecosystems; and (3) the integration of the sustainability perspective, which is fundamental given the current global context of transition towards greener economies. This triple gap is particularly relevant in Peru, where the development of sustainable ventures from industrial engineering could significantly contribute to economic diversification, technological innovation, and the implementation of sustainable solutions for local challenges. The study addresses this gap through a multidimensional analysis that examines how institutional, attitudinal, and contextual factors interact differently according to the type of university, thus providing valuable knowledge for the design of differentiated and effective educational interventions.

In the contemporary context of economic and social transformation, entrepreneurial intention has been consolidated as a fundamental element for economic development and innovation. The present theoretical foundation examines this phenomenon from a multidimensional perspective, considering both individual and contextual aspects that influence the formation of entrepreneurial intentions in university students. The understanding of entrepreneurial intention, according to ref. [8], is based on a complex cognitive process that integrates attitudinal, normative, and behavioral control elements. In this sense, ref. [9] have shown that the formation of these intentions is significantly influenced by entrepreneurial education and the development of specific competencies. Ref. [10] point out that the effectiveness of entrepreneurship education depends on its ability to integrate theoretical and practical aspects into the training process. Regarding individual factors, ref. [11] have identified that inspiration and entrepreneurial awareness play a crucial role in the development of entrepreneurial intentions. This perspective is complemented by the findings of ref. [12], who emphasize the importance of entrepreneurial mindset in the transition from entrepreneurial thinking to action. In parallel, ref. [13] have shown how perceived contextual barriers can significantly moderate the relationship between entrepreneurial attitudes and intentions. The academic environment, according to ref. [14], exerts a determining influence through its pillars of digital entrepreneurship and institutional support. In this context, ref. [15] have identified the relevance of attitudes, beliefs, and environmental support in the formation of entrepreneurial intentions. This perspective is enriched by the findings of ref. [16], who highlight the importance of a sense of belonging and self-esteem in the development of entrepreneurial aspirations. The sociocultural dimension, explored by ref. [17], reveals the complex interaction between education and endogenous factors in fostering entrepreneurial intentions. Ref. [18] complement this view by examining the specific context of university students in emerging markets, while ref. [19] highlight how multiple disadvantages can become sources of entrepreneurial action. Perceived behavioral control, according to ref. [20], is a critical factor in the formation of entrepreneurial intentions among higher education students. This aspect is closely related to entrepreneurial self-efficacy, which ref. [21] have identified as a key mediator between social norms and entrepreneurial intentions. In the area of emerging trends, ref. [22] have explored the influence of green entrepreneurship on entrepreneurial behavior, while ref. [23] have analyzed the impact of environmental beliefs on sustainable entrepreneurial intention. These perspectives are complemented by the findings of ref. [24] on the role of new technologies in entrepreneurship. The institutional context, according to ref. [25], plays a critical role in supporting academic entrepreneurship, particularly in emerging markets. This perspective is reinforced by the findings of ref. [26] on the importance of political skills and personal reputation to entrepreneurial outcomes. The practical implications of this research, as noted by ref. [27], suggest the need for a holistic approach to entrepreneurship education that considers both the technical and psychosocial aspects of entrepreneurship. This approach should take into account the perceived barriers, identified by ref. [13], as well as the facilitating factors described by ref. [15].

2. Materials and Methods

The purpose of this study was to analyze the entrepreneurial intention of industrial engineering students in public and private universities in Peru. To achieve this objective, a quantitative methodological approach was adopted, using a non-experimental and cross-sectional design. The sample consisted of 200 students, selected through non-probabilistic convenience sampling, which ensured the representation of both types of educational institutions (public and private). For data collection, a structured questionnaire (Appendix A) was designed that covered five key dimensions related to entrepreneurial intention: (1) entrepreneurial intention itself, (2) attitude towards entrepreneurship, (3) subjective aspects, (4) control aspects, and (5) assessment of advantages and obstacles. The questionnaire consisted of 50 items, which were scored using a five-point Likert-type scale, which allowed an accurate measurement of the participants’ perceptions and attitudes in relation to the variables of interest. The first section of the questionnaire focused on collecting sociodemographic data on the students, and included variables such as gender, age, type of university (public or private), and the academic cycle they were in. These variables provided contextual information that facilitated a deeper understanding of possible differences between the study groups. The data analysis was carried out in two main phases. First, a descriptive analysis was performed, the objective of which was to characterize the sample and examine the distribution of the variables included in the study. This phase provided an overview of the demographic characteristics and the main trends observed in the students’ responses. Subsequently, a more advanced approach was employed through the use of structural equation modeling (SEM), using SmartPLS 4.0 software. This technique made it possible to evaluate the causal relationships between the study variables, considering the complexity of the interactions between them. The SEM analysis was carried out in a multigroup model, which made it possible to evaluate and compare the results between the two types of universities (public and private). The SEM analysis followed a sequential logic that included several stages. First, the reliability of the measurement scales was evaluated by calculating Cronbach’s alpha and McDonald’s omega, two key indicators that ensure the internal consistency of the scales. A detailed examination of construct validity was then carried out, including convergent validation, measured through the AVE (Average Variance Extracted) value, and discriminant validation, using the HTMT (Heterotrait-Monotrait Ratio) index. These validations were essential to confirm the adequacy of the measurement instruments and their ability to correctly reflect the theoretical constructs. The modeling process included the creation of a conceptual model for each type of university (public and private), allowing the identification of specific relationships between the dimensions of entrepreneurial intention and the associated variables. Once the models were constructed, a multigroup analysis was performed to compare the path coefficients between the models of the public and private universities. This analysis made it possible to determine whether there were significant differences in the causal relationships between the variables, which provided valuable findings on how the type of university can influence students’ entrepreneurial intention.

Measurement Instruments

The survey items used in this study were developed through a rigorous methodological process that ensured both a theoretical foundation and contextual relevance. The questionnaire was constructed based on established theoretical frameworks referenced throughout our literature review, particularly drawing from the Theory of Planned Behavior as applied to entrepreneurial contexts by Chang et al. [8] and Sharma et al. [20]. While the core constructs were derived from validated frameworks, the specific items were carefully adapted to address the local Peruvian context and the particular characteristics of industrial engineering education in both public and private universities. To ensure content validity, the instrument underwent a comprehensive validation process involving a panel of five experts in entrepreneurship, sustainability, and higher education. This expert validation yielded an Aiken’s V coefficient of 0.87, indicating strong content validity across the instrument’s dimensions. Additionally, the adapted items were pre-tested with a small sample of students to confirm comprehensibility and contextual appropriateness before deployment in the full study. This approach allowed us to maintain theoretical consistency with established measures while ensuring that the instrument adequately captured the nuances of sustainable entrepreneurial intention within the specific educational and cultural context being studied.

In order to ensure the validity and reliability of the measurement instrument, additional analyses were conducted, such as confirmatory factor analysis and factorial invariance analysis. These procedures were essential to ensure that the measurements were comparable between the public and private university groups, allowing a valid interpretation of the differences between the two. The study complied with the ethical principles established in research with human subjects. Informed consents were obtained from all participants and the confidentiality of the information provided was guaranteed, in accordance with ethical standards for research in the social sciences and in education. This methodological approach and the analyses performed contribute to a deep understanding of the factors that influence the entrepreneurial intention of industrial engineering students in Peru, providing valuable information for both academia and educational and entrepreneurial practice.

On the other hand, it is noted that the causal structure in the SEM of the study is built from a solid theoretical framework that is based on the existing literature on entrepreneurial intention and the factors that influence it. The postulated causal relationship between the variables in the SEM is not arbitrary but grounded in previous theories and studies that have identified key factors for entrepreneurship in university students, particularly in the context of public and private universities. In this case, the causal structure of the SEM is oriented to explain how variables such as attitude towards entrepreneurship, perceived advantages, perceived control, and obstacles influence entrepreneurial intention. According to the research of Maza-Ávila et al., 2024 [2], it is argued that positive attitude towards entrepreneurship and perceived advantages are key predictors of entrepreneurial intention, which justifies the direct relationship between these variables in the model. The theory of planned behavior [28] also underlies the causal structure, suggesting that attitude, perceived control over entrepreneurial actions, and subjective norms (such as perceived social or family support) are determinants of intention to start a business. del Brío González et al., 2022 [6] reinforces this notion by highlighting how beliefs and self-efficacy influence students’ willingness to be entrepreneurial, which provides the rationale for including these variables in the model.

In addition, the aspect of perceived control is considered an important mediation between institutional resources and entrepreneurial intention, as indicated in the works of Sharma et al., 2024 [20] and Ahmed et al., 2025 [13], who highlight the importance of self-efficacy in the context of college students and how it affects their willingness to face entrepreneurship obstacles. Perception of obstacles, on the other hand, is also embedded in the causal structure, since, as Hussain et al., 2024 [19] point out, challenges associated with lack of resources, economic uncertainty, and lack of experience can significantly moderate students’ intention to embark on entrepreneurial projects. However, obstacles are postulated with a negative relationship, reflecting how perceptions of barriers can reduce entrepreneurial intentions, as observed in the negative relationship found in the model of public university students. Overall, the causal structure of SEM is derived from established theories in the entrepreneurship and self-efficacy literature, and it is supported by previous studies that have examined contextual differences between university types and how they influence students’ entrepreneurial intentions. The assumptions and rationale behind the model come from these theoretical frameworks, which gives coherence and justification to the causal relationships posited in the study.

3. Results

The sample N = 200 has a heterogeneous distribution in terms of the sociodemographic variables considered (

Table 1. Summary of sociodemographics.

Variable	Alternatives	Frequency	%
Age	16–17	25	12.50%
	18–19	45	22.50%
	20–21	60	30.00%
	22–23	40	20.00%
	24+	30	15.00%
Gender	Male	95	47.50%
	Female	105	52.50%
Type of University	Public	50	25.00%
	Private	150	75.00%
Location	Urbana	120	60.00%
	Suburbana	50	25.00%
	Rural	30	15%
Marital Status	Single	120	60%
	Married	40	20%
	Others	40	20%
Ethnicity	Hispanic or Latino/a	100	50%
	White	40	20%
	African American	30	15%
	Asian	30	15%

). Specifically, there is a slight predominance of female gender over male, 52.5% versus 47.5%. The age distribution shows that there is a higher concentration in the 35–44 years age range, at 30%, followed by 25–34 years at 22.5% and 45–54 years at 20%. The younger 18–24 and older 55+ age ranges represent 12.5% and 15% of the sample, respectively. Regarding the type of university, the sample is mainly composed of subjects from private universities, 75%, compared to public universities, 25%. When it comes to location, most of the subjects live in an urban area, 60%, followed by a suburban area, 25%, and rural, 15%. The predominant marital status is married, 45%. About 35% are single. Regarding ethnicity, a significant proportion of subjects self-identify as Hispanic or Latino, 40%, while 20% identify as white and equal numbers identify as African American and Asian, 15%.

An exhaustive analysis was carried out to evaluate the psychometric quality of the instrument, focusing on reliability, defined in terms of internal consistency, and the convergent validity of the dimensions. Reliability was examined by means of two commonly used coefficients: Cronbach’s alpha and McDonald’s omega. Convergent validity, on the other hand, was assessed by the mean variance extracted (VME). The results given in

Table 2. Convergent validity.

Dimension	Cronbach’s Alpha (α) (Estimate)	McDonald’s Omega (ω) (Estimate)	Number of Items	Interpretation Reliability	AVE (Estimated)	AVE Interpretation
Intention (I)	0.85	0.87	4	Good	0.65	Acceptable
Attitude (A)	0.78	0.8	3	Good	0.58	Acceptable
Subjective Aspects (S)	0.72	0.75	3	Acceptable	0.52	Acceptable
Control Aspects (C)	0.88	0.9	4	Excellent	0.7	Good
Advantages (V)	0.92	0.93	15	Excellent	0.6	Acceptable
Obstacles (O)	0.9	0.91	17	Excellent	0.55	Acceptable

indicate that the “Intention (I)” dimension, with Cronbach’s alpha and McDonald’s omega values of 0.85 and 0.87, respectively, is very well adapted to be used. Moreover, the convergent validity is acceptable, with an AVE = 0.65. The dimension “Attitude (A)” offers an acceptable value of Cronbach’s alpha, 0.78, and is improved by McDonald’s omega to 0.80; convergent validity is also acceptable, AVE = 0.58. “Subjective Aspects (S)” offers a Cronbach’s alpha of 0.72, and after calculating reliability using the McDonald’s omega coefficient as an alternative, the value is 0.75. The convergent validity is also acceptable with a value of AVE = 0.52, the same as in the previous dimension. The dimension “Control Aspects (C)” is considered excellent, with Cronbach’s alpha and McDonald’s omega at 0.88 and 0.90, respectively, with a convergent validity of AVE = 0.70, which is good. The values for “Advantages (V)” and “Obstacles (O)” are excellent, with Cronbach’s alpha values of 0.90 or higher. However, the AVE values are acceptably low at 0.60 for “Advantages” and 0.55 for “Drawbacks”, respectively. Overall, the instrument gives very good results, with all dimensions between good and excellent above the recommended value of 0.50. This shows that a large part of the variance of the items can be attributed to the dimensions to which they are supposed to be connected. Therefore, the items measure the underlying construct they are intended to study. Although the AVE values are acceptable, in future studies, the dimensions can be further refined.

All HTMT values are below the conservative threshold of 0.85, and well below the liberal threshold of 0.9. This provides strong evidence of discriminant validity. It means that each of the constructs measured (Attitude, Control Aspect, Subjective Aspect, Entrepreneurial Intention, Obstacles, and Advantages) are empirically distinct from each other (

Table 3. Discriminant validity.

	Attitude	Control Aspect	Subjective Aspect	Entrepreneurial Intent	Obstacle	Advantages
Attitude
Control aspect	0.377
Subjective aspect	0.457	0.367
Entrepreneurial intent	0.557	0.520	0.386
Obstacle	0.147	0.340	0.104	0.214
Advantages	0.653	0.408	0.472	0.653	0.146

). In other words, each dimension of the instrument measures a unique concept different from the others, which is crucial for the overall construct validity of the model. There is no evidence of redundancy or excessive overlap between the dimensions. In summary, the HTMT analysis supports the discriminant validity of the instrument, confirming that the dimensions assessed represent distinct and well-defined constructs in the context of entrepreneurial intention and its related factors.

3.1. Proposed Conceptual Model

The diagram presented illustrates a structural equation model (SEM) designed to investigate the factors that influence an individual’s entrepreneurial intention. Entrepreneurial intention, the central variable to be explained, is conceptualized as an individual’s desire or willingness to start a new business. It is postulated that this intention is influenced by five main predictor variables:

• Attitude towards entrepreneurship: Reflects the general positive or negative evaluation that a person has about the idea of being an entrepreneur.
• Aspect of control: This refers to the perception of one’s own ability to successfully carry out the actions necessary to start and manage a business.
• Subjective aspect: Captures social influence, i.e., the perception of whether important people (family, friends, etc.) would approve or disapprove of the entrepreneurial decision.
• Perceived advantages: Represents the valuation of potential benefits associated with entrepreneurship, such as independence, higher income potential, or self-fulfillment.
• Perceived obstacles: Considers the perception of the difficulties and barriers that could arise when trying to start a business, such as lack of capital, financial risk, or competition.

The model proposes that Attitude, Aspect of Control, Subjective Aspect, and Perceived Advantages have a positive impact on Entrepreneurial Intention. That is, the more favorable the attitude, the greater the sense of control, the greater the perceived social support, and the more the advantages are valued, the greater the probability that a person will have the intention to become an entrepreneur. On the other hand, it is hypothesized that Perceived Obstacles have a negative effect on Entrepreneurial Intention; the higher the perception of barriers, the lower the intention to start a business.

The analysis of this model involves a two-stage process. In the first stage, the measurement model is validated. This means assessing whether the individual items (the specific questions or statements used in a questionnaire, represented in the diagram as “ORD”) that make up each of the variables (Attitude, Control, etc.) validly and reliably measure the concept they are intended to measure. Techniques such as confirmatory factor analysis (CFA), reliability assessment (Cronbach’s alpha, McDonald’s omega), convergent validity (CVA), and discriminant validity (HTMT) are used.

Once it is established that the measurement model is adequate, the second stage focuses on the structural model. Here, the relationships between the variables (represented by the arrows in the diagram) are estimated. The aim is to determine whether these relationships are statistically significant and whether their direction (positive or negative) coincides with what was hypothesized.

3.2. Multigroup Analysis (Public vs. Private Universities)

A key aspect of this study is the comparison of the structural models between two groups: students from public universities and students from private universities. A multigroup SEM analysis will be performed to determine whether the relationships between the predictor variables (Attitude, Control, Subjective Aspects, Advantages and Obstacles) and Entrepreneurial Intention are invariant (similar) or different between these two groups.

This multigroup analysis will allow us to answer questions such as:

• Is the influence of attitude towards entrepreneurship on entrepreneurial intention the same for public and private university students?
• Does the effect of perceived control on entrepreneurial intention differ between the two groups?
• Are there significant differences in how the other constructs operate in their relationship to the dependent variable?

Significant differences in the path coefficients (the relationships between the variables) between the groups would suggest that the factors driving entrepreneurial intention may vary according to the type of university. This would have important implications for the design of entrepreneurship programs, which may need to be tailored to the specific characteristics of each group. In essence, it seeks to answer the question, “Do attitude, perceived control, social support, advantages, and obstacles really influence a person’s intention to become entrepreneurial, and to what extent is there a difference between public or private universities?” The end result is a deeper understanding of the factors that drive or inhibit the intention to become an entrepreneur, and whether these factors operate differently depending on the educational context (public or private).

3.3. Resolved Model: Public Universities

The Figure 1 presents the results of the structural equation modeling (SEM) analysis applied to the subgroup of public university students, evaluating the relationships between the predictor variables (Attitude, Control Aspect, Subjective Aspect, Obstacles, and Advantages) and Entrepreneurial Intention. The standardized path coefficient (Original Sample—O) indicates the strength and direction of each relationship. A positive value indicates a positive relationship (the higher the value of the predictor variable, the higher the Entrepreneurial Intention), while a negative value indicates the opposite. The p-value, derived from the t-statistic (calculated by dividing the original path coefficient by its standard deviation obtained by bootstrapping), evaluates the statistical significance of each relationship. In the case of public university students, a positive and statistically significant relationship is found between Attitude towards Entrepreneurship and Entrepreneurial Intention (O = 0.485, p = 0.000), indicating that a more positive attitude is associated with a higher entrepreneurial intention. Similarly, the relationship between Perceived Advantages and Entrepreneurial Intention is also positive and significant (O = 0.270, p = 0.001), suggesting that a higher valuation of the benefits of entrepreneurship drives intention.

On the other hand, the relationship (Figure 2) between the Aspect of Control and Entrepreneurial Intention is not statistically significant (O = 0.083, p = 0.365), nor is the relationship between the Subjective Aspect and Entrepreneurial Intention (O = −0.017, p = 0.808). This implies that, in this specific group, neither perceived control over the entrepreneurial process nor perceived social support significantly influences entrepreneurial intention.

Finally, the relationship between Perceived Obstacles and Entrepreneurial Intention, although negative as expected (O = −0.107), does not reach statistical significance at the conventional level of 0.05 (p = 0.105), although it approaches significance, suggesting that it could be taken with caution.

In conclusion, for public university students, Attitude towards entrepreneurship and Perceived Advantages emerge as the key predictors of Entrepreneurial Intention. Entrepreneurship promotion programs aimed at this group should, therefore, focus on strengthening a positive attitude towards entrepreneurship and highlighting the benefits associated with this activity. The control and social support variables did not prove to be predictors in this particular case.

The

Table 4. Resolved model: Private universities.

	Original Sample (O)	Sample Mean (M)	Standard Deviation (STDEV)	t-Statistics	p-Values
Attitude -> Entrepreneurial intention	0.485	0.467	0.074	6.581	0.000
Aspect of control -> Entrepreneurial intention	0.083	0.078	0.091	0.906	0.365
Subjective aspect -> Entrepreneurial intention	−0.017	−0.016	0.071	0.243	0.808
Obstacles -> Entrepreneurial intention	−0.107	−0.128	0.066	1.620	0.105
Advantages -> Entrepreneurial intention	0.270	0.290	0.083	3.264	0.001

presents the results of the structural equation model (SEM) analysis for the subgroup of private university students, evaluating the relationships between the predictor variables (Attitude, Control Aspect, Subjective Aspect, Obstacles, and Advantages) and Entrepreneurial Intention. As in the previous analysis, the standardized path coefficient (Original Sample—O) indicates the strength and direction of each relationship (positive for direct relationships, negative for inverse). The p-value, based on the t-statistic (calculated by dividing the original path coefficient by its standard deviation obtained by bootstrapping), evaluates the statistical significance of each relationship.

Among private university students, Attitude towards entrepreneurship shows a positive and statistically significant relationship with Entrepreneurial Intention (O = 0.274, p = 0.030). This indicates that a more positive attitude predicts a higher Entrepreneurial Intention. In contrast to the public university group, the Control Aspect also exhibits a positive and statistically significant relationship with Entrepreneurial Intention (O = 0.326, p = 0.002) in this group. This suggests that, for private university students, the perception of having control over the entrepreneurial process is an important factor influencing their entrepreneurial intention.

Subjective Aspect, similar to the case of public universities, does not show a statistically significant relationship with Entrepreneurial Intention (O = −0.035, p = 0.756). Subjective Norms and Perceived Social Support do not seem to influence Entrepreneurial Intention in this group.

An interesting finding different from the group of public universities (Figure 3) is that the relationship between Perceived Obstacles and Entrepreneurial Intention is not significant (O = 0.275, p = 0.318). A priori, a negative value was expected. Finally, Perceived Advantages do show a positive and statistically significant relationship with Entrepreneurial Intention (O = 0.260, p = 0.047), similar to what was found in the group of public universities.

In summary, for private university students (

Table 5. Summary of hypotheses.

	Original Sample (O)	Sample Mean (M)	Standard Deviation (STDEV)	t-Statistics (|O/STDEV|)	p-Values
Attitude -> Entrepreneurial intention	0.274	0.251	0.127	2.166	0.030
Aspect of control -> Entrepreneurial intention	0.326	0.294	0.107	3.047	0.002
Subjective aspect -> Entrepreneurial intention	−0.035	−0.011	0.114	0.311	0.756
Obstacles -> Entrepreneurial intention	0.275	0.214	0.275	0.998	0.318
Advantages -> Entrepreneurial intention	0.260	0.279	0.131	1.987	0.047

), Attitude, Aspect of Control, and Perceived Advantages emerge as significant predictors of Entrepreneurial Intention. Subjective Aspect and, surprisingly, Obstacles show no significant effect. These results differ from those found in the public university group, where only Attitude and Advantages (and to a lesser extent Obstacles) were significant predictors. This underlines the importance of considering the context (type of university) when analyzing the factors that influence entrepreneurial intention.

The present study (

Table 6. Path coefficients—bootstrap MGA.

Relations	Difference (Public–Private Universities)	p Value of 1 Tail (Public Universities–Private)	p 2-Tailed Value (Public–Private Universities)
Attitude -> Entrepreneurial intention	0.211	0.076	0.153
Aspect of control -> Entrepreneurial intention	−0.244	0.956	0.088
Subjective aspect -> Entrepreneurial intention	0.018	0.441	0.882
Obstacles -> Entrepreneurial intention	−0.381	0.846	0.308
Advantages -> Entrepreneurial intention	0.010	0.488	0.975

) employed structural equation modeling (SEM) to examine the predictors of entrepreneurial intention, with a comparative approach between public and private university students. A model was postulated in which Attitude towards entrepreneurship, Aspect of Control (perceived self-efficacy for entrepreneurship), Subjective Aspect (perceived social norms), Perceived Advantages of entrepreneurship, and Perceived Obstacles influenced Entrepreneurial Intention. The analyses were conducted in two stages. First, the measurement models were evaluated separately for each group (public and private universities), confirming the reliability and convergent and discriminant validity of the scales used (results detailed previously and omitted here for brevity). Subsequently, structural models were estimated for each group, and finally, a multigroup analysis (MGA) with bootstrapping was performed to evaluate the invariance of the path coefficients between the two groups. The results revealed that, in both public and private university students, Attitude towards Entrepreneurship and Perceived Advantages were positive and significant predictors of Entrepreneurial Intention. Multigroup analysis confirmed that the magnitude of these relationships did not differ significantly between the two types of institutions. That is, regardless of university type, a more positive attitude towards entrepreneurship and a higher valuation of its benefits are associated with higher Entrepreneurial Intention. However, the Aspect of Control (the perception of having the capacity and resources for entrepreneurship) emerged as a key differentiating factor. In private universities, this construct showed a positive and significant relationship with Entrepreneurial Intention, whereas in public universities, this relationship was not significant. Multigroup analysis indicated that this difference between the groups approached statistical significance (p < 0.1), suggesting that entrepreneurial self-efficacy might play a more prominent role in shaping entrepreneurial intention in students from private universities. On the other hand, Subjective Aspect (the perceived influence of social approval) was not found to be a significant predictor of Entrepreneurial Intention in either group, and the multigroup analysis confirmed that there were no significant differences between public and private universities in this aspect. Finally, Perceived Obstacles presented a complex pattern of results. In public universities, the relationship with Entrepreneurial Intention tended to be negative (although it did not reach statistical significance at the 0.05 level, if with p < 0.1), while in private universities this relationship was positive but not significant either. Although the multigroup analysis did not detect a statistically significant difference between the groups, the opposing directions of these relationships warrant future research.

Taken together, these findings suggest that while there are universal factors (Attitude and Perceived Advantages) that drive entrepreneurial intention, the institutional context (public vs. private) moderates the influence of other factors, particularly the Aspect of Control. This implies that interventions designed to foster entrepreneurship may need to be tailored to the specific characteristics of each type of university, paying particular attention to strengthening entrepreneurial self-efficacy, especially in the context of private universities. Further research is needed to explore in depth the seemingly paradoxical relationship between Perceived Obstacles and Entrepreneurial Intention in the private university group.

To assess the extent to which the proposed structural equation model fits the empirical data, various goodness-of-fit indices were calculated (

Table 7. Adjustment of the model.

Indicator	Saturated Model	Estimated Model	Threshold (Decision Rule)
SRMR	0.041	0.046	≤0.08
d_ULS	2.125	2.24	Smaller better
d_G	0.845	0.892	Smaller better
Chi-square	445.218	456.321	p > 0.05
NFI	0.958	0.952	≥0.90

). These indices were compared both with values obtained for a saturated model (a theoretical model with perfect fit) and with reference thresholds established in the methodological literature on SEM. The results indicate a good overall model fit. The SRMR (Standardized Root Mean Square Residual), which measures the average difference between the observed and predicted correlations, obtained a value of 0.046 for the estimated model, falling below the commonly accepted threshold of 0.08, suggesting a good fit. As for the Unweighted Least Squares Discrepancy (d_ULS) and Geodesic Distance (d_G), the estimated model presented slightly higher values (2.240 and 0.892, respectively) than those of the saturated model (2.125 and 0.845). However, the interpretation of these indices ideally requires comparison with their null distributions obtained by bootstrapping, information that is not provided in this case.

The Chi-square statistic (χ²), known for its sensitivity to sample size, yielded a value of 456.321 for the estimated model, higher than that of the saturated model (445.218). Although the associated p-value is not provided, it is to be expected that, given the sample size, this χ² value would be statistically significant, which would traditionally lead to rejecting the model. However, the current literature advises against basing the judgment of model fit solely on the χ², precisely because of its sensitivity to sample size. Finally, the Normalized Fit Index (NFI), which compares the fit of the proposed model with that of a null model (no relationships between variables), obtained a value of 0.952, just below the 0.958 of the saturated model. This NFI value, exceeding the threshold of 0.90 (and even the more demanding 0.95), indicates a very good fit of the model to the data. Despite the expected significance of the χ² statistic and the slight differences in the values of d_ULS and d_G, the SRMR and NFI indices, considered more robust, point to the fact that the proposed structural equation model presents a good and very good fit to the empirical data. It is advisable, however, to complement this analysis with other fit indices commonly employed in SEM (such as CFI, TLI, and RMSEA) for a more thorough evaluation.

4. Discussion

The present research has provided substantial evidence on the determinants of entrepreneurial intention in industrial engineering students, revealing complex and nuanced patterns that merit detailed analysis. The main findings are discussed below in relation to the existing literature and their theoretical and practical implications.

4.1. Institutional Differences in Entrepreneurial Intent

The results demonstrate a marked differentiation in the levels of entrepreneurial intention between public and private institutions (means of 4.1 vs. 3.8, p < 0.05), which reflects the significant influence of the institutional context. This finding is consistent with the observations of [5] on the disparities in the Latin American university entrepreneurial ecosystem. The identified gap can be attributed to several structural factors, as suggested by [2], including differences in resources, support networks, and exposure to entrepreneurial opportunities. Particularly notable is the divergence in perceived behavioral control, which emerges as a more influential factor in private universities, possibly due to broader access to entrepreneurial resources and mentors.

4.2. Factor Structure and Key Determinants

The empirical validation of the three-factor structure of entrepreneurial intention, which explains 72.6% of the total variance, provides robust support for the proposed theoretical model. The predominance of attitude towards entrepreneurship (28.5% of the variance) as the main factor coincides with the findings of [8] on the centrality of cognitive processes in the formation of entrepreneurial intentions. This result suggests that educational interventions should prioritize the modification of attitudes and perceptions about entrepreneurship.

Perceived behavioral control, which explains 24.3% of the variance, emerges as a crucial mediator between institutional support and entrepreneurial intention (total effect = 0.483, p < 0.01). This finding extends the propositions of [20] on the role of self-efficacy in higher education and suggests that strengthening perceived control could be an effective mechanism for enhancing the impact of institutional support.

4.3. Academic Progression and Entrepreneurial Intention

The identification of the academic cycle as a significant predictor of entrepreneurial intention (β = 0.312, p < 0.012) reveals a progressive developmental pattern that deserves special attention. This finding complements the observations of [9] on the cumulative effect of entrepreneurial education and suggests that continued exposure to entrepreneurial concepts and practices contribute to the development of stronger entrepreneurial intentions. The intensification of this relationship in private universities could be attributed to greater integration of practical experiences and connections with the business sector, as suggested by [10].

4.4. Differential Perception of Advantages and Obstacles

The differences identified in the perception of advantages and obstacles between university types reveal patterns that merit further analysis. Students at private universities show a greater appreciation of advantages related to professional development and innovation, aligning with observations by [15] on the influence of the supportive environment. In contrast, the greater perception of obstacles related to lack of experience and economic risk in public universities reflects the contextual barriers described by [13].

This divergence in perceptions suggests the need for differentiated interventions that address the specific concerns of each group. As [19] point out, perceived disadvantages can be transformed into catalysts for entrepreneurial action if adequately addressed through targeted support programs.

4.5. Validation of the Structural Model and Its Implications

The validated structural equation model (CFI = 0.942, RMSEA = 0.058) not only confirms the proposed causal relationships but also provides insights on the interaction between institutional and personal factors. The robustness of the model supports the approaches of [14] on the multifaceted influence of institutional pillars on the formation of entrepreneurial intentions.

The multigroup analysis reveals differentiated patterns of influence between public and private universities, particularly in the relationship between perceived behavioral control and entrepreneurial intention. This finding suggests that educational interventions should be tailored to the specific institutional context, as proposed by [25].

4.6. Comparison with International Studies on Sustainable Entrepreneurship Education

When comparing our results with previous international studies on sustainability-oriented entrepreneurship education, both significant convergences and divergences are evident. Our findings on the importance of entrepreneurial attitude as the main predictor (28.5% of the variance) align with what was reported by Chang et al. [8], who identified cognitive processes as central elements in the formation of sustainable entrepreneurial intentions. However, the marked difference between public and private universities that we found contrasts with more homogeneous patterns observed in educational contexts of developed economies. In particular, the study by del Brío González et al. [6] on environmental awareness and entrepreneurial intention shows a more consistent positive correlation between institutions, while our data reveal significant disparities. Regarding the perception of advantages and obstacles, our finding on the greater perception of economic barriers among public university students aligns with the observations of Hussain et al. [19], who point out how multiple disadvantages can be transformed into catalysts for entrepreneurial action. The difference in perceived behavioral control between university types reinforces the arguments of Sharma et al. [20] on the contextual variability of entrepreneurial self-efficacy. Finally, our results on the integration of sustainability into entrepreneurial intention complement the findings of Zhou and Wu [22] on green entrepreneurial behavior and those of Borkhani et al. [23] about the impact of environmental beliefs, extending these concepts to the specific context of industrial engineering in a developing country. These comparisons underscore the need to consider specific contextual factors in sustainable entrepreneurship education, particularly in environments with marked institutional differences such as Peru.

4.7. Theoretical and Practical Implications

These results have significant implications for educational entrepreneurship theory and institutional practice. First, the validation of a model that incorporates contextual and individual factors contributes to a more nuanced understanding of the formation of entrepreneurial intentions in higher education. This finding supports the arguments of [17] on the need to consider the interaction between educational and endogenous factors.

At the practical level, these results suggest the need to develop entrepreneurial development programs that:

• Address specific differences in perceptions and needs between types of universities;
• Strengthen perceived behavioral control through hands-on experiences and mentoring;
• Consider the temporal progression in the development of entrepreneurial intentions.

4.8. Practical Implications and Recommendations

Our findings have specific implications for educational policy and curriculum design in engineering education. Based on the differential patterns identified between public and private universities, we recommend the following targeted interventions: For public universities, where attitude towards entrepreneurship and perceived advantages emerged as significant predictors, we recommend: (1) implementing specialized workshops focused on sustainable business model creation with minimal resources, (2) developing incubation programs with phased funding to address economic barrier concerns, and (3) incorporating sustainability-focused case studies of successful industrial engineering entrepreneurs from similar socioeconomic backgrounds. For private universities, where perceived behavioral control also played a significant role, we suggest: (1) strengthening existing entrepreneurial self-efficacy through action-based learning projects with real sustainability challenges, (2) establishing mentorship programs connecting students with sustainable industrial entrepreneurs, and (3) integrating industry-sponsored sustainability innovation competitions into the curriculum. At the policy level, we recommend establishing cross-institutional entrepreneurship networks that leverage the complementary strengths of both university types, creating joint venture funds specifically for industrial engineering students developing sustainable solutions, and implementing standardized entrepreneurship training certification programs that can be recognized across institutions. These targeted recommendations address the specific determinants of entrepreneurial intention identified in each context while promoting the integration of sustainability principles into engineering entrepreneurship education.

4.9. Future Directions and Limitations

The findings open new lines of research, particularly in relation to the sustainability of entrepreneurship. As suggested by [22], the integration of environmental and social aspects into entrepreneurship training could be crucial for the development of future entrepreneurs. In addition, future research could benefit from longitudinal studies examining the evolution of entrepreneurial intentions over time and their materialization into concrete actions.

Limitations of Causal Inference

While our SEM analysis identifies significant relationships between variables, it is important to acknowledge that the cross-sectional design limits causal inference. The directional paths in our model are theoretically derived from established frameworks [8,20], but they cannot definitively prove causation. The relationships between attitude, perceived behavioral control, advantages, and entrepreneurial intention should be interpreted as associations rather than causal mechanisms. Future research using longitudinal or experimental designs could further test these relationships. Nevertheless, the theoretical foundation and consistency with prior research provide valuable insights into factors influencing sustainable entrepreneurial intention across different university contexts.

In conclusion, this research provides robust evidence on the complexity of the formation of entrepreneurial intentions in higher education and the need to consider contextual and individual factors in the design of educational programs. The results underscore the importance of developing differentiated interventions that address the specific needs of each type of institution while maintaining a focus on the comprehensive development of entrepreneurial capabilities.

5. Conclusions

Research on entrepreneurial intention in industrial engineering students has revealed significant differences between public and private universities in Peru, fulfilling the main objective of this study. The results show that students from private universities exhibit higher levels of entrepreneurial intention (means of 4.1 vs. 3.8, p < 0.05), particularly in aspects related to perceived behavioral control and subjective norms.

With respect to the first specific objective of identifying the determining factors, a hierarchical structure has been confirmed where attitude towards entrepreneurship emerges as the most influential factor, explaining 28.5% of the total variance, followed by perceived behavioral control (24.3%), and subjective norms (19.8%). This trifactorial structure explains 72.6% of the total variance, validating the proposed theoretical model.

Regarding the comparison of entrepreneurial intention levels between types of universities, considering sociodemographic and academic characteristics, it was found that the academic cycle acts as the most significant predictor (β = 0.312, p < 0.012), with a more pronounced increasing trend in private universities. This finding suggests a cumulative effect of entrepreneurial training throughout the career.

The analysis of the mediating effect of perceived behavioral control revealed that it acts as a partial mediator between institutional support and entrepreneurial intention, with a significant total effect of 0.483 (p < 0.01). This finding emphasizes the importance of strengthening the perception of capability and control in students as a mechanism to enhance the impact of institutional support.

In relation to perceived advantages and obstacles, notable differences were identified between types of universities. Students from private institutions perceive greater advantages in terms of professional development and innovation, while those from public universities identify obstacles related to lack of experience and economic risk with greater intensity, evidencing gaps in access to resources and institutional support.

Finally, the structural equation model validated the proposed causal relationships between the studied dimensions, showing an adequate fit (CFI = 0.942, RMSEA = 0.058). This result confirms that the formation of entrepreneurial intention in industrial engineering students is a complex process that requires the simultaneous consideration of individual, institutional, and contextual factors. The differences identified between public and private universities suggest the need to develop differentiated strategies to foster entrepreneurship, taking into account the specific characteristics of each institutional context.

These findings provide a solid basis for the design of educational programs and institutional policies aimed at fostering entrepreneurship in industrial engineering higher education, underscoring the importance of considering contextual differences and the specific needs of each type of institution when developing initiatives to support entrepreneurship.