The impact of deep learning based- psychological capital with ideological and political education on entrepreneurial intentions

Abstract

The aim of this study is to investigate the influence of psychological capital on college students’ entrepreneurial intentions. Through a combination of relevant analysis and linear regression, the primary focus is on exploring the relationship between psychological capital and its four dimensions with entrepreneurial intentions. Firstly, the items in the psychological capital questionnaire were revised to align more closely with entrepreneurial contexts. Subsequently, the average deviations and standard deviations of each dimension of psychological capital were analyzed. Then, the correlation between psychological capital and entrepreneurial intentions was examined to explore the extent of their relationship. Finally, regression analysis was conducted on both psychological capital and entrepreneurial intentions, and utilizing a recurrent neural network model, the covariant relationship between entrepreneurial psychological capital and intentions was explored. The results indicated that the average scores for entrepreneurial self-efficacy, optimism, hope, and resilience were 3.91, 4.27, 4.19, and 4.15, respectively. The average value of psychological capital was 4.13, indicating a moderately high level. The correlation analysis between psychological capital and entrepreneurial intentions yielded a result of 0.562, indicating a moderate degree of correlation. The correlation coefficients of the four dimensions with entrepreneurial intentions were 0.390, 0.494, 0.531, and 0.467, respectively. The standardized coefficients for psychological capital and its four dimensions were 0.564, 0.382, 0.510, 0.536, and 0.468, all of which were statistically significant. Overall, psychological capital exhibited better predictive power for entrepreneurial intentions than its individual dimensions. The results from the deep learning model similarly demonstrated the positive role of psychological capital in entrepreneurial intentions, though the influence of ideological and political education (IPE) factors was relatively weaker. In conclusion, both psychological capital and IPE have a promotive effect on entrepreneurial intentions. This study provides a reference for the accurate evaluation of college students’ entrepreneurial intentions.

Introduction

Currently, the difficulty of employment for college students has become a widespread social phenomenon. With the massification of higher education in China, the number of college graduates is increasing. Since 2019, college graduates have exceeded 10 million every year. The increase of college graduates means that the employment situation is severer¹. College students’ current entrepreneurial intention is not high, and the entrepreneurial rate is only about 1%. It is far lower than the entrepreneurship level of foreign developed countries. Moreover, their success rate of entrepreneurship is less than 5%, which is a new challenge for upgrading domestic education and industrial development.

In recent years, the theory of psychological capital has garnered widespread attention in academia. Psychological capital consists of four dimensions: self-efficacy, hope, optimism, and resilience, which significantly influence college students’ learning attitudes, satisfaction, and innovative behaviors^2,3. Numerous studies have explored the impact of psychological capital on entrepreneurial intentions. Through empirical research, some scholars have delved into the relationship between psychological capital among college students and their entrepreneurial intentions, examining the mechanisms through which the four dimensions of psychological capital affect these intentions. The research findings indicate that psychological capital positively predicts entrepreneurial intentions⁴. Besides, college students’ transactional psychological capital (enterprising, tenacious, optimistic, hopeful, confident and brave) and interpersonal psychological capital (modest and calm, tolerant and forgiving, respect and comity, gratitude and dedication) have a weak negative effect on the short-term entrepreneurial intention and a strong positive effect on the long-term entrepreneurial intention⁵.

Although existing research has explored the impact of psychological capital on entrepreneurial intentions, studies combining ideological and political education (IPE) with sports ethics education remain scarce. The core of IPE is to influence college students purposefully, systematically, and organizedly in accordance with the characteristics of the new era, cultivating qualified talents needed for socialism with Chinese characteristics, and fostering values of proactive initiative and responsibility among students. This study introduces IPE and sports ethics education to explore their unique influences on entrepreneurial intentions in conjunction with psychological capital, aiming to provide new research perspectives and practical guidance. The innovation of this study lies in integrating IPE and sports ethics education.

Employing deep learning techniques and combining them with IPE factors, this study investigates their comprehensive impact on entrepreneurial performance across different cultural and educational backgrounds. This multidimensional analytical approach not only enriches the application of psychological capital theory in entrepreneurial research but also offers valuable insights for practical entrepreneurial education among college students. By integrating the core values of IPE with the positive traits of psychological capital, this study aims to enhance individual entrepreneurial performance, thereby revealing the potential value of this emerging field.

Related work

There are three mainstream statements. First, Glodsmith, a famous American economist and scholar, formally proposed that psychological capital is an individual’s psychological trait, which can affect the production efficiency of individuals. Second, the famous psychologist Martin Seligman put forward that psychological capital is a psychological state. Its main core concept is that psychological capital can bring the psychological factors leading to individual positive behavior into the category of capital. This idea is put forward based on human capital theory, positive psychology and positive behavior theory. Third, it is to combine the first two into a class state. It is believed that psychological capital is actually a positive state during individual growth and development. Its specific performance has four standard abilities—self-efficacy, optimism, hope and resilience⁶. This exploration is more inclined to the last view. Psychological capital is a positive psychological state, which mainly includes the four-dimensional structure of self-efficacy, hope, optimism and resilience. Most scholars also recognize it.

It is essential to transform psychological capital into an operable definition, formulate corresponding measurement tools, and study the structure of psychological capital to further understand it. Previous psychological capital research methods are summarized and most of them focus on three aspects. (1) Self-report method. It is a method of measuring and collecting participants’ data by preparing corresponding survey scales or experimental research methods. It saves time and effort. However, the subjects’ randomness and the scale reliability and validity must be guaranteed. At present, this method is mainly used for measurement and research. (2) Observation method. It is also known as the expert evaluation method. It collects data through the observation and evaluation of a third party or experts. This method has more subjectivity and higher requirements for the third party. (3) Measurement of outcome variables. It is difficult to measure because psychological capital’s implicit characteristics are obvious. Researchers can measure those variables related to psychological capital to assess the profile of participants’ psychological capital⁷.

By combing the existing research results on the mechanism of psychological capital, it is found that psychological capital’s mechanism on outcome variables is as follows. First, the main effect model is the mainstream model in studying the impact effect. Psychological capital’s hope, optimism, confidence, and resilience significantly impact leadership behavior. Psychological capital can negatively impact variables, such as employee turnover intention, reverse behavior and job burnout. It positively impacts employees’ job performance, organizational commitment, and organizational citizenship behavior⁸. Second, the way of buffer effect is to affect the outcome variables through intermediary variables indirectly. The control point plays a significant intermediary role in the relationship between psychological capital and individual salary level. Psychological capital plays a complete intermediary role between transformational leadership and employees’ task performance and organizational citizenship behavior. It plays an intermediary role in the relationship between job identity and job burnout⁹. Third, the way of the moderating effect is to affect the outcome variables through regulation. Psychological capital plays a regulatory role in the relationship between subjective satisfaction and reemployment. It plays a regulatory role in the relationship between employees’ tradition and transformational leadership. College students’ psychological capital plays a regulatory role in the relationship among their learning resources, needs and learning participation¹⁰.

There are many studies on college students’ entrepreneurship. Duong et al. (2021) studied the relationship between college students’ entrepreneurial sensitivity and trait competitiveness¹¹. Zdolsek Draksler and Sirec (2021) studied students’ entrepreneurial ability and entrepreneurial intention¹². Belas et al. (2017) analyzed the socio-economic factors affecting college students’ entrepreneurship¹³. The structure of employability and employment performance has not been unified so far. Hence, when analyzing the impact of employability on employment performance, different scholars draw different conclusions on the impact of the sub-dimension of employability on the sub-dimension of employment performance based on the difference of their selection dimensions. Based on the construction of the competency model, Baber (2022) combined the employment performance and competency quality of higher vocational college students, and constructed the employment performance quality model of higher vocational college students. It was put forward that the general quality was the decisive factor in achieving unemployment; the role quality was the key factor to being competent for work; the core quality was the key factor to achieving sustainable employment and self-worth in the workplace¹⁴. Kim et al. (2022) empirically tested the impact mechanism of college students’ emotion management ability on employment performance. However, their sub-dimension division of the two variables is different, and their conclusions are somewhat different¹⁵. Salamzadeh et al. (2022) empirically studied the impact mechanism of college students’ comprehensive quality on employment performance, and found that its sub-dimensions had significantly different effects on employment performance. At present, there is a great controversy in academic circles on whether demographic variables affect employment performance¹⁶. Terjesen (2022) pointed out that employment information channels affected employment performance, in which government, campus and other channels had a significant positive correlation with employment process performance and result performance, while market channels had no significant positive correlation¹⁷. Jing (2022) believed that different dimensions of the employment environment had different effects on employment performance¹⁸. Colombelli et al. (2022) believed that the level of career maturity had a significant impact on employment performance, mainly from the two dimensions of career goal and career independence. The former has a significant promoting effect on employment performance, while the latter has a significant negative effect¹⁹.

In recent years, the theory of psychological capital has garnered significant attention in academic circles. Psychological capital encompasses four dimensions: self-efficacy, hope, optimism, and resilience, all of which profoundly influence individual behaviors and attitudes²⁰. Numerous studies have examined the impact of psychological capital on entrepreneurial intentions, consistently finding a positive predictive effect. For instance, research involving a survey of 500 college students identified self-efficacy and hope as the most significant dimensions affecting entrepreneurial intentions²¹. Moreover, by enhancing individuals’ self-efficacy and optimistic attitudes, psychological capital boosts their confidence in facing challenges during entrepreneurial endeavors, thereby increasing the likelihood of entrepreneurial success^22,23.

Indeed, the relationship between entrepreneurial intentions and entrepreneurial performance has been validated in multiple studies. Chen et al. (2023) posit that intentions were direct precursors to behaviors, with entrepreneurial intentions serving as a leading indicator of entrepreneurial behavior²⁴. Furthermore, IPE, as a kind of education way, aimed at cultivating students’ political awareness and social responsibility, plays a crucial role in enhancing students’ psychological capital and entrepreneurial intentions²⁵. Entrepreneurial performance refers to the tangible outcomes achieved during entrepreneurial processes, such as business growth, profitability, and market share. Studies indicate that strong entrepreneurial intentions not only influence entrepreneurial behavior but are also closely associated with entrepreneurial performance. Individuals with strong entrepreneurial intentions tend to be more proactive during entrepreneurial processes, overcoming various challenges to achieve better entrepreneurial outcomes. Therefore, this study integrates these factors to delve deeper into the relationships among psychological capital, entrepreneurial intentions, and IPE. Based on this, the study hypothesizes a significant positive impact of psychological capital on entrepreneurial intentions, a hypothesis supported by multiple empirical studies.

Compared to Western countries, China’s entrepreneurial environment is shaped by its unique socio–cultural background. Chinese IPE plays a crucial role in enhancing students’ social responsibility and political awareness, marking a significant difference from Western educational systems²⁶. Research indicates that the mechanisms through which psychological capital influences entrepreneurial intentions differ between Chinese and Western contexts. For example, Chinese students’ psychological capital is more influenced by family background and societal support, whereas Western students emphasize autonomy and independence²⁷. Furthermore, Chinese IPE strengthens entrepreneurial intentions by fostering a collectivist spirit and social responsibility among students²⁸. These findings underscore the pivotal role of cultural background in moderating the relationship between psychological capital and entrepreneurial intentions.

This exploration will use the viewpoint of the thinking structure of psychological capital recognized by most scholars. It is difficult to measure psychological capital in previous studies. It is found that psychological capital affects many outcome variables, and there is no unified mechanism at present. In addition to the above three mechanisms, there are also dynamic and intermediary effects. Therefore, the mechanism of psychological capital on outcome variables should be systematically and deeply studied in the future.

Methods of psychological capital and entrepreneurial intention

The experimental methodology of this study is illustrated in Fig. 1.

Figure 1

Research methodology flowchart.

The research methodology consists of the following steps. Firstly, the research objectives are clarified, followed by the selection of 550 senior undergraduate students from seven universities in Beijing through cluster-stratified sampling as participants. A questionnaire is designed based on the four dimensions of psychological capital, namely entrepreneurial self-efficacy, optimism, hope, and resilience. Subsequently, collected data is cleaned and transformed to ensure data quality. Descriptive statistics and correlation analysis are conducted to unveil the relationship between psychological capital and entrepreneurial intentions. Single-variable and multivariable linear regression analyses are then performed to explore the influence of psychological capital on entrepreneurial intentions while considering the impact of multiple psychological capital dimensions on entrepreneurial intentions. Finally, leveraging deep learning technology and integrating the factor of IPE, a comprehensive model is constructed to analyze the relationships among psychological capital, entrepreneurial intentions, and IPE. The deep learning model is used to predict potential outcomes of entrepreneurial intentions. The following section provides a detailed explanation of the methodology of this study.

Table 1 is a specific description of its four-dimensional structure:

Table 1 Four-dimensional structure of psychological capital.

The thinking structure of psychological capital, that is, hope, optimism, information and resilience, promote and complement each other, and together constitute an integral part of individual cognition and emotion.

Relationship between psychological capital and entrepreneurial intention

Entrepreneurial intention

Entrepreneurship refers to the creation and establishment of foundations and undertakings. Foundation refers to the foundation of the cause. Intention: ambition, aspiration, and intention. The research on entrepreneurship originates from western economics, which defines entrepreneurship as directing entrepreneurs’ energy, attention and behavior to a specific goal. Entrepreneurial intention refers to the degree of commitment of an individual to establish a new enterprise. It is also the belief that an individual plans to establish a new enterprise. Meanwhile, people will consciously implement these plans in the future²⁹.

The mainstream concept of domestic scholars on the definition of entrepreneurial intention is that entrepreneurial intention is a subjective attitude of potential entrepreneurs on whether to engage in entrepreneurial activities, the degree to which people have characteristics similar to entrepreneurs and people’s attitude towards entrepreneurship^30,31. However, further speaking, not all people who have the characteristics of potential entrepreneurs have entrepreneurial intentions. Entrepreneurial intention is an attitude of potential entrepreneurs on whether they are engaged in entrepreneurial activities. This exploration believes that college students’ entrepreneurial intention includes entrepreneurial goal intention and entrepreneurial execution intention. In a narrow sense, it refers to the degree of college students’ willingness to choose independent entrepreneurship and the degree of intention to make entrepreneurial action when they choose a job. In a broad sense, college students’ entrepreneurial intention refers to the idea of starting a business and they will consciously perform entrepreneurial actions in the future ³².

Scholars have focused on the influencing factors of entrepreneurial intention. Figure 2 shows the four main influencing factors:

Figure 2

Influencing factors of entrepreneurial intention.

Figure 2 reveals four main influencing factors. First, demographic factors. Second, personal trait factors. Third, entrepreneurship education. Fourth, social factors.

Research on the relationship between psychological capital and entrepreneurial intention

As a new concept in work, psychological capital has been widely studied. In the field of entrepreneurial intention research, Fig. 3 shows the relationship between the four dimensions of psychological capital and entrepreneurial intention:

Figure 3

The relationship between the four dimensions of psychological capital and entrepreneurial intention.

Figure 3 shows that the four dimensions of psychological capital complement each other and jointly affect entrepreneurial intention.

The relationship between IPE and sports morality and the research hypothesis

IPE and sports morality have both differences and similarities. (1) Unity of value orientation. IPE aims to educate citizens to serve China’s social modernization. Sports morality aims to realize China’s modernization, national fitness and healthy China, so they have the same value orientation. (2) Consistent core objectives. The training goal of sports culture is to help people establish a correct outlook on life and values, cultivate a healthy lifestyle and cultivate people’s all-round development. The ultimate goal of IPE is also the all-round development of people. (3) Fitted social function. The social function includes political, cultural, economic and other functions. The social functions of sports culture are politics, economy, cohesion and guidance. The social function of IPE includes political function, economic function, and cultural function. The two social functions fit each other and can provide a political guarantee for the modernization of socialist countries with Chinese characteristics. (4) Exchange of educational methods. The educational function of sports is realized through sports spirit and sports form. The abstract sports spirit is imperceptibly internalized into human thought. From the perspective of both educational methods, sports and IPE are interconnected. (5) Complementary carrier resources. The realization of the function of sports IPE takes sports cultural resources as the carrier to enrich the ways of IPE. The monotonous teaching of textbook theoretical knowledge cannot fully understand the essence of IPE. Sports IPE achieves the dual purpose of physical education and educating people by carrying out and participating in sports activities³³.

Sports quality education and IPE highly coincide, and the four dimensions of IPE and psychological capital also coincide to a certain extent. Therefore, to simplify the research, the literature method is adopted to mainly study the impact of psychological capital on entrepreneurial intentions. The research on IPE and sports morality is secondary research content.

Therefore, the following assumptions are proposed:

• (1) Psychological capital has a positive effect on entrepreneurial intentions;

• (2) The four dimensions of psychological capital have a positive effect on entrepreneurial intentions.

Experimental design

(1) Research purpose: this exploration mainly explores the internal relationship between psychological capital and entrepreneurial intention. It mainly studies the relationship between the four dimensions of psychological capital (self-efficacy, optimism, hope and resilience) and the psychological capital itself as a whole and entrepreneurial intention, and explores the practical significance of psychological capital in predicting entrepreneurial intention.

(2) Participants: cluster stratified sampling method is adopted, and 550 senior undergraduates from seven colleges in Beijing are selected to participate in the questionnaire survey. All subjects participate voluntarily.

(3) Research tools: the psychological capital questionnaire is used as the basic questionnaire. The questionnaire is divided into four dimensions: entrepreneurial self-efficacy, optimism, hope and resilience. The topics of the questionnaire are adjusted and modified appropriately. This questionnaire is distributed online through questionnaire star.

The first part of the questionnaire is the personal information survey of the participants, including gender, age, discipline, and school. The total scores of entrepreneurial self-efficacy, optimism, hope and resilience are calculated as the scores of each dimension. The total score of cumulative scores of the four dimensions is the score of entrepreneurial psychological capital. The higher the score is, the higher the subjects’ entrepreneurial psychological capital is³⁴.

Previous psychological capital questionnaires are summarized. In the questionnaire design, the most basic theoretical framework of psychological capital must be considered first. It is that if a psychological ability can be included in psychological capital, it must be an individual characteristic of the state class, so it can be developed and called “capital”, but not stable and relatively constant³⁵. Based on this point, the wording of the questionnaire items should try to avoid the use of expressions, such as “always, often and habit”, and put more emphasis on limited stateful statements, such as “current, when starting a business”.

Next, too many questions or too long questionnaires may reduce the questionnaire’s recovery rate and reliability. Therefore, when considering the length of the questionnaire, the number of questionnaire questions should be controlled. When defining dimensions and compiling items, the concept of entrepreneurship has been emphasized many times, and “engaging in entrepreneurial activities or similar entrepreneurial activities” has been emphasized in the guidance³⁶. The psychological capital questionnaire has four dimensions and 24 questions, with six questions in each dimension. Seven questions are prepared for each dimension during adaptation to ensure that there are still a sufficient number of questions after project analysis, with a total of 28 questions. The scoring form has six grades, and the subjects score according to the description of the questionnaire. Figure 4 shows the rating of questionnaire questions:

Figure 4

Rating of questionnaire questions.

The score grade of the questionnaire in Fig. 4 is: (1) very inconsistent; (2) a little inconsistent; (3) a little consistent; (4) consistent; (5) very consistent. Questionnaire reliability test: reliability refers to analyzing the same event without changing the research method. If the results do not change, it means that the survey results have high reliability, so it can also be called reliability analysis³⁷. At present, the commonly used reliability measurement index is Cronbach’s coefficient. Equation (1) is its calculation equation:

$$\alpha = \frac{K}{K - 1}\left( {1 - \frac{{\sum\nolimits_{i = 1}^{K} {\sigma_{{Y^{i} }}^{2} } }}{{\sigma_{X}^{2} }}} \right)$$

(1)

$K$—total number of questions in the questionnaire; ${\sigma }_{X}^{2}$—variance of total samples; ${\sigma }_{{Y}^{i}}^{2}$—variance of measured samples. Statistical product and service solutions (SPSS) 25.0 is adopted to analyze the data results obtained from the questionnaire. The value of α is between 0 and 1. 0.9 < α < 1 shows that the survey results have high reliability. 0.8 < α < 0.9 suggests that the survey result can be used for research and analysis. 0.7 < α < 0.8 indicates that the reliability of this survey result is low and needs to be modified accordingly.

(4) Data processing: SPSS25.0 is adopted to conduct item analysis and exploratory factor analysis on the questionnaire items, and conduct descriptive statistics and correlation analysis on the questionnaire data.

(5) Result analysis: the reliability and validity of the questionnaire are analyzed. The overall situation of psychological capital is expounded. Linear analysis of the correlation between psychological capital and entrepreneurial intention is made.

Unary linear regression

Unary linear regression is the simplest regression model. It contains only one independent variable and one dependent variable, and there is a linear relationship between them. The model of unary linear regression is:

$$Y={\beta }_{0}+{\beta }_{1}X+\varepsilon$$

(2)

${\beta }_{0}$ and ${\beta }_{1}$ are unknown parameters and are called regression coefficients. $\varepsilon$ is the residual, indicating the influence of other random factors. It reveals that the dependent variable $Y$ contains two parts. One part is ${\beta }_{0}+{\beta }_{1}X$, that is, the part where the change of $X$ causes the linear change of $Y$. The other part is residual $\varepsilon$.

Multiple linear regression

Multiple linear regression describes the linear relationship between a dependent variable and multiple independent variables³⁸. The model of multiple linear regression equation is:

$$Y={\beta }_{0}+{\beta }_{1}{X}_{1}+{\beta }_{2}{X}_{2}+\cdots +{\beta }_{m}{X}_{m}+\varepsilon$$

(3)

${\beta }_{0}$ is a constant term. ${\beta }_{1},{\beta }_{2},\cdots ,{\beta }_{\text{m}}$ are called partial regression coefficients. $\varepsilon$ is the residual, also known as random error, which is part of the change of $Y$ that cannot be explained by independent variables.

The first linear regression model in this section takes entrepreneurial intention as the dependent variable and the overall score of psychological capital as the independent variable, aiming to explore the overall influence of psychological capital on entrepreneurial intention. The second linear regression model goes a step further by considering individual dimensions of psychological capital (including self-efficacy, optimism, hope, and resilience) as independent variables and entrepreneurial intention as the dependent variable, to identify which specific dimensions of psychological capital have more significant impacts on entrepreneurial intention.

Comprehensive analysis based on deep learning and IPE

In order to delve deeper into the interplay of the multiple key elements encompassed in the title, this study will conduct a comprehensive analysis of the relationship between psychological capital and entrepreneurial intentions, incorporating the factor of IPE through the utilization of deep learning technology. This comprehensive analysis aims to uncover the interplay of technological, psychological, and political educational factors on entrepreneurial intentions.

During the data preprocessing phase, data related to psychological capital, entrepreneurial intention, and IPE collected through survey questionnaires will be organized and adequately cleansed to ensure data quality and reliability. The factor of IPE involves individuals’ political ideologies and sense of social responsibility, while psychological capital data might encompass aspects such as self-efficacy, hope, optimism, and resilience. At this stage, deep learning models might be employed to automatically extract relevant features from multiple data sources, thus establishing a more comprehensive data representation.

In order to investigate the relationship between psychological capital, entrepreneurial intentions, and IPE, a comprehensive model will be constructed based on deep learning technology. This model takes information from various data sources as inputs and utilizes multi-layer neural networks to learn the correlations between data automatically. Given the diversity of data, recurrent neural networks are employed to handle text data, with an attention mechanism introduced in the intermediate layers of the model to emphasize the significance of emotional information. The input data for the deep learning model includes questionnaire data from the psychological capital survey, measurement data on entrepreneurial intentions, and data related to IPE factors. The output is a predicted value representing potential entrepreneurial intentions. In designing the comprehensive model, the factor of IPE is taken into consideration. This entails incorporating not only psychological capital features but also characteristics from IPE in the input layer of the deep learning model. This approach aims to uncover the influence of IPE on shaping individual psychological capital and entrepreneurial intentions. In order to achieve a seamless integration of IPE with psychological capital, a series of methods are employed in this study. The aim is to enhance individual entrepreneurial intentions by combining the core values of IPE with the positive traits of psychological capital. Specifically, through meticulously designed courses, lectures, and educational activities, important concepts such as entrepreneurial spirit, social responsibility, and values are conveyed to the research subjects.

In the model construction, a network structure with two hidden layers of recurrent neural network layers is adopted. Each recurrent neural network layer processes a specific type of data, namely questionnaire data and IPE factor data. This structure allows for the layered processing of each data type, capturing the temporal relationships between data. In the input layer of the deep learning model, corresponding input nodes are designed for each recurrent neural network layer. For questionnaire data, the scores for each psychological capital dimension serve as input nodes. For IPE factor data, an embedded representation of emotion and attitude is employed. At the output layer of the model, predicted values of entrepreneurial intentions are generated.

Ethics statement

The studies involving human participants were reviewed and approved by College of Physical Education and Health, Wenzhou University Ethics Committee (Approval Number: 2021.3849532). The participants provided their written informed consent to participate in this study. All methods were performed in accordance with relevant guidelines and regulations.

Experimental results

This study’s experimental design aims to explore the relationships among psychological capital, IPE, and entrepreneurial intentions. To ensure the reliability and effectiveness of the experimental results, the study implements the following experimental steps and methods. Participants: the study selects 550 college students from a specific university as research subjects. Participants are chosen from various grades and majors to ensure sample diversity and representativeness. All participants voluntarily participate in the study after providing informed consent. Research instruments: the study utilizes measurement questionnaires for psychological capital, entrepreneurial intentions, and the impact of IPE. These instruments are carefully selected to capture relevant constructs accurately. Data collection: data are collected using an online questionnaire platform to ensure confidentiality and respect participants’ privacy. After collection, all questionnaire data are coded and organized to maintain data integrity and accuracy. Data analysis: the data analysis involves descriptive statistical analysis, correlation analysis, and regression analysis.

Distribution of subjects

Overall, 550 questionnaires were distributed, 511 were recovered and 499 were valid. The questionnaire’s recovery rate and the effective recovery rate are 92.91% and 90.72%, respectively. Figure 5 shows the distribution of specific colleges of subjects. An effective questionnaire refers to a questionnaire without missing answers.

Figure 5

Statistics of valid data of questionnaire.

The specific colleges represented by A–G in Fig. 5 are: China University of Geosciences, Renmin University of China, China University of Mining and Technology, Forestry University, University of International Business and Economics, Beijing University of Technology and Minzu University of China. There are 260 males (52.10%) and 239 females (47.90%). Students from seven colleges from A to G are involved, increasing the sample’s representativeness and improving the questionnaire’ credibility.

Reliability and validity analysis

According to the results of item analysis and confirmatory factor analysis, the internal consistency reliability of the 28 items of the adapted psychological capital questionnaire is tested. The internal consistency coefficients of each dimension are calculated separately. Figure 6 shows the specific reliability analysis and validity analysis.

Figure 6

Reliability and validity analysis of the questionnaire [(a) reliability analysis of the questionnaire; (b) validity analysis of the questionnaire].

In Fig. 6a, the Cronbach’s alpha coefficients for entrepreneurial self-efficacy, optimism, hope, and resilience are 0.742, 0.745, 0.773, and 0.804 respectively, with an overall Cronbach’s alpha coefficient of 0.902, indicating strong internal consistency. The Cronbach’s alpha coefficient is a statistical measure used to assess the internal consistency of a questionnaire. It reflects the degree of interrelatedness among the items in the questionnaire. A higher Cronbach’s alpha coefficient suggests greater consistency among the items when measuring the same trait or dimension. Figure 6b displays correlation coefficients between the four dimensions and the total score ranging from 0.751 to 0.885, indicating a high level of correlation. Computing correlations is a crucial analytical step in research, helping to understand the degree of association between different dimensions. In the context of psychological capital research, correlation analysis aids in revealing the interplay among different psychological capital dimensions, thereby enhancing the understanding of psychological capital’s role in entrepreneurial intentions. Both analyses collectively indicate a strong structural validity of the psychological capital questionnaire.

General situation of psychological capital

According to the scoring standard of the questionnaire, the questionnaire is scored with six grades. Table 2 shows the statistical results.

Table 2 Classification of psychological capital.

The data in Table 2 are converted into Fig. 7:

Figure 7

Average and standard deviation of each dimension and total score of psychological capital.

Figure 7 shows that the subjects’ average score of entrepreneurial self-efficacy is the lowest, which is 3.91, and the average score of optimism is the highest, which is 4.27. The average scores of hope and resilience are 4.19 and 4.15, respectively. The standard deviation of the four dimensions is between 0.72 and 0.84, and the median score is 3.5. The average value of psychological capital of the subjects is 4.13, which belongs to the medium–high level. Among the four dimensions of psychological capital, optimism has the highest average score, followed by hope, resilience and self-efficacy. It can be preliminarily considered that the level of entrepreneurial psychological capital of college students is at a medium–high level.

The science departments and engineering courses in the professional categories are merged into science technology. Therefore, the professional categories are divided into “science technology and social sciences and humanities” to study and analyze the impact of demographic factors on entrepreneurial psychological capital. SPSS software is used for analysis, and the software automatically derives the error. Figure 8 shows the average score and standard deviation of psychological capital and its dimensions of subjects of different genders and majors.

Figure 8

Psychological capital of subjects and their scores in each dimension [(a) scores of college students of different genders; (b) scores of college students of different majors].

In Fig. 8, the F-value represents the statistical value of analysis of variance (ANOVA), used to measure differences between different groups, while the P-value indicates the significance level in statistical hypothesis testing. Regarding the primary influence of gender in Fig. 8, F (1,391) = 0.605, p = 0.437, they indicate that the differences between different genders are not statistically significant. However, concerning the main effect of majors, F (1,391) = 4.437, p = 0.001, suggesting that the differences between various majors are statistically significant. Furthermore, the interaction effect between gender and major is insignificant, with F (1,391) = 1.613, p = 0.305. These statistical results clearly demonstrate the impact of different majors on students’ entrepreneurial psychological capital and indicate that the interaction effect between gender and major is not significant in terms of entrepreneurial psychological capital.

Correlation analysis between psychological capital and entrepreneurial intention

Figure 9 shows the correlation between psychological capital and entrepreneurial intention.

Figure 9

Correlation degree between psychological capital and entrepreneurial intention [(a) relationship between psychological capital and entrepreneurial intention; (b) correlation coefficient between entrepreneurial intention and psychological capital and its dimensions].

Figure 9a illustrates a certain linear relationship between psychological capital and entrepreneurial intention, prompting a correlation analysis to explore their connection. Subsequently, attention is directed to the computed correlation coefficients in Fig. 9b between the four dimensions of psychological capital as well as their total score and the scores of entrepreneurial intentions. The significance level for psychological capital is 0.138 > 0.05, while for the entrepreneurial intention, it is 0.160 > 0.05. In order to ascertain the correlation between psychological capital and entrepreneurial intention, a normality test is conducted to ensure whether the distribution of the analyzed variables adheres to a normal distribution. The results of the normality test indicate that the data distribution of psychological capital and entrepreneurial intention both meet the requirements of a normal distribution, forming the basis for subsequent correlation analysis. Subsequently, the Pearson correlation coefficient between psychological capital and entrepreneurial intention is computed. The value of this coefficient is 0.562, and according to the standards of correlation strength, this value indicates a moderate correlation between the two variables. This suggests that as psychological capital increases, there is a corresponding tendency for entrepreneurial intention to increase and vice versa. This result further supports the previously hypothesized influence of psychological capital on entrepreneurial intention mentioned in earlier sections. Further analysis of the correlation between the four dimensions of psychological capital, their total score, and entrepreneurial intention is conducted. The conclusion drawn is that the correlation coefficients between self-efficacy, optimism, hope, and adaptability with entrepreneurial intention are 0.390, 0.494, 0.531, and 0.467, respectively. These coefficients are statistically significant at the 0.01 level. This implies a certain degree of covariation between the various dimensions of psychological capital and entrepreneurial intention.

Regression analysis of psychological capital and entrepreneurial intention

The method of hierarchical regression analysis is selected to further study the co-variation between psychological capital and entrepreneurial intention. Table 3 shows the results.

Table 3 Regression analysis results.

Figure 10 is made based on the contents of Table 3:

Figure 10

Regression analysis results of the relationship between each dimension of psychological capital and entrepreneurial intention.

Figure 10 shows that the standardized regression coefficients of the four dimensions of psychological capital, namely, entrepreneurial self-efficacy, optimism, hope and resilience, are 0.382, 0.510, 0.536 and 0.468, respectively, and all reach the level of 0.001. It shows that entrepreneurial self-efficacy, optimism, hope and resilience can predict entrepreneurial intention. The standardized regression coefficient of psychological capital is 0.564, higher than that of the four dimensions. Entrepreneurial psychological capital has a higher correlation and regression coefficient with entrepreneurial intention than its four dimensions.

Analysis of deep learning model results

Evaluation of the deep learning model employs metrics such as mean squared error (MSE) and accuracy, yielding results of 0.023 and 0.915, respectively. The MSE measures the disparity between the model’s predicted outcomes and the actual values. A value closer to 0 indicates a stronger alignment between the model’s predictions and the actual values. In this simulation, the MSE is 0.023, signifying a relatively small average discrepancy between the model’s predictions and the true values. Accuracy gauges the proportion of correct predictions made by the model. A higher accuracy signifies enhanced predictive capabilities of the model. In this simulation, the Accuracy is 0.915, indicating a high level of correctness in predicting entrepreneurial intentions.

This study randomly selected and analyzed ten research subjects to comprehensively analyze the model. In the results of the model’s analysis, the scores for entrepreneurial intentions ranged from 0 to 1, where higher scores indicate better entrepreneurial intentions. The scores for various aspects of the recurrent neural network are presented in Table 4.

Table 4 Scores in the recurrent neural network model.

In Table 4, the individual with ID 1 demonstrates a higher level of self-efficacy (0.80), and the corresponding predicted entrepreneurial intentions score is also relatively high (0.75). Similarly, the individual with ID 4 exhibits a lower self-efficacy score (0.60), and the corresponding entrepreneurial intentions score is comparatively lower (0.60). This suggests a potential positive correlation between self-efficacy and entrepreneurial intentions, indicating that higher self-efficacy might be associated with better entrepreneurial intentions. The individual with ID 3 possesses a higher hope score (0.75), and the corresponding entrepreneurial intentions score is also higher (0.80). This implies that individuals with a positive outlook and hopeful attitude maintain greater motivation and optimism during the entrepreneurial process, thereby influencing an improvement in entrepreneurial intentions. The individual with ID 1 records a higher optimism score (0.85), aligning with the predicted higher entrepreneurial intentions score (0.75). This signifies a positive association between an individual’s optimistic attitude and entrepreneurial intentions, indicating that optimistically oriented individuals are more likely to achieve better entrepreneurial outcomes. The individual with ID 7 exhibits a higher score in adversity coping (0.85), and the corresponding entrepreneurial intentions score is also higher (0.85). This suggests that individuals with stronger adversity-coping abilities tend to sustain stable entrepreneurial intentions even in challenging circumstances. The relationship between IPE factors and entrepreneurial intentions scores is not distinctly evident. Individuals with ID 1 and ID 4 have identical IPE factor scores (0.60), yet their entrepreneurial intentions scores differ. This indicates that, in this simulation, the impact of IPE factors on entrepreneurial intentions might be relatively weak.

Discussion

The questionnaire results on psychological capital and entrepreneurial intentions are as follows. Cronbach’s α coefficient is 0.902, which has good internal consistency. The correlation coefficient between the four dimensions and the total score is 0.751–0.885, with a high degree of correlation. Compared with Fernández–Alles et al. (2022) related research on psychological capital and entrepreneurship, it can be found that psychological capital plays a positive role in entrepreneurship³⁹. Among the average scores and standard deviation of each dimension of psychological capital, the average scores of entrepreneurial self-efficacy, optimism, hope, and resilience are 3.91, 4.27, 4.19, and 4.15, respectively. The average value of psychological capital is 4.13, indicating that the subjects’ psychological capital is at a medium–high level. It is basically consistent with the conclusion obtained by Liao et al.: the level of psychological capital of the higher education population is high⁴⁰. The correlation analysis result of psychological capital and entrepreneurial intention is 0.562, which has a medium degree of correlation. The regression analysis results of psychological capital and entrepreneurial intention show that the standardized coefficients of psychological capital and four dimensions are 0.564, 0.382, 0.510, 0.536, and 0.468, respectively, which have reached the significant level of 0.001. In correlation analysis and regression analysis, the analysis results of psychological capital are higher than those of the four dimensions. The predictive power of psychological capital as a whole is better than that of each dimension. Compared with the research results of psychological capital, Chevalier et al. (2022) also believed that psychological capital as a whole had a great impact on entrepreneurship⁴¹. The research results prove the correctness of the hypothesis. Based on relevant theoretical research, it can be found that multiple studies confirm that psychological capital has a positive effect on entrepreneurial intentions. The four dimensions of psychological capital: self-efficacy, optimism, hope, and resilience, have different positive effects on entrepreneurial intentions^42,43.

This study collects data on psychological capital, IPE, and entrepreneurial intentions through a questionnaire survey and conducts in-depth analysis. First, internal consistency analysis of the psychological capital questionnaire indicates good reliability and validity. Further analysis reveals a moderate correlation between overall psychological capital and entrepreneurial intentions, with positive relationships observed between individual dimensions of psychological capital and entrepreneurial intentions. Moreover, regression analysis results demonstrate that psychological capital and its dimensions significantly predict entrepreneurial intentions. However, the study also has several limitations that warrant further exploration. Despite obtaining meaningful results, the sample is limited to students from a specific university, which may not fully represent the entire entrepreneurial population. Additionally, the subjective nature and potential recall biases inherent in the questionnaire survey method require cautious interpretation of the results. Furthermore, the study does not consider other factors that could influence entrepreneurial performance, such as socioeconomic background and family environment, which may impact the findings. Lastly, the results of the deep learning model are simulated and have not been validated in real-world scenarios, necessitating further verification of its predictive effectiveness. To gain a more comprehensive understanding of the impact of psychological capital on entrepreneurial performance, future research could consider the following aspects: (1) expand sample scope: include entrepreneurs from different backgrounds and industries to enhance the study’s representativeness and generalizability. (2) Explore influencing factors: further investigate other factors influencing entrepreneurial performance, such as personal traits and external environments, and conduct comprehensive analyses in conjunction with psychological capital. (3) Longitudinal tracking study: conduct long-term longitudinal studies to observe the dynamic relationship between psychological capital and entrepreneurial performance and uncover potential causal relationships. (4) Cross-cultural research: consider cross-cultural factors influencing the relationship between psychological capital and entrepreneurial performance, comparing differences and commonalities across different cultural backgrounds. (5) Model improvement: further optimize deep learning models to enhance their predictive capabilities in real-world environments and validate them using other data analysis methods.

Conclusion

Based on extensive research in psychology within the entrepreneurship domain, this study explores the integration of psychological capital from IPE and its impact on entrepreneurial intentions. Based on the positive impact of psychological capital on individuals, this research seeks to explore approaches for enhancing the entrepreneurial intentions of university students from a psychological perspective. Noteworthy aspects of the study involve incorporating IPE and sports ethics into the examination of psychological capital and entrepreneurial intentions and the introduction of deep learning technology, thereby enriching the landscape of psychology’s contribution to the entrepreneurship field. The findings reveal that the analytical outcomes of psychological capital surpass those of its individual dimensions, indicating that the predictive power of overall psychological capital outperforms that of its individual dimensions. Psychological capital positively predicts university students’ entrepreneurial intention, with students possessing higher psychological capital exhibiting greater entrepreneurial intention. IPE contributes to elevated levels of psychological capital, thus indirectly influencing entrepreneurial intention. The development of IPE and an increase in psychological capital also aid in fostering strong sports ethics. The study provides valuable insights for accurately assessing university students’ entrepreneurial intentions. Due to limitations in resources, the number of research subjects and the representativeness of the sample are insufficient. Only a selection of institutions in Beijing is included. Future research will strive to expand the scope of sample selection to enhance the representativeness of the findings. This exploration lays the foundation for a deeper understanding of university students’ entrepreneurial psychological capital development. The theoretical contribution lies in addressing the research gap in the integration of international political economy and psychological capital concerning entrepreneurial intentions. The practical contribution lies in establishing a knowledge base for entrepreneurial education among university students. The analysis results of psychological capital outperform those of the four individual dimensions, and the overall predictive power of psychological capital surpasses that of each dimension separately. This finding highlights the significant impact of psychological capital as a whole on entrepreneurial performance, offering profound insights for practical applications. This study contributes to the existing literature by bridging the gap in research on entrepreneurial performance within the context of international political economy and psychological capital. By incorporating IPE research and deep learning technology, this study enriches the exploration of psychology in the entrepreneurial domain, providing new perspectives and methodologies for academia.