Applying Structural Equation Modeling (SEM) to Analyze Factors Affecting the Entrepreneurial Intention of the Students of Vietnam National University, Hanoi45264

Applying Structural Equation Modeling SEM to Analyze Factors Affecting the Entrepreneurial Intention of the Students of Vietnam National University, Hanoi Nguyen Cam Ngoc 1 , Nguyen T

Applying Structural Equation Modeling (SEM) to Analyze Factors Affecting the Entrepreneurial

Intention of the Students of Vietnam National

University, Hanoi

Nguyen Cam Ngoc 1 , Nguyen The Kien 2 (*)

(1) VNU University of Social Sciences and Humanities, Hanoi, Vietnam

(2) VNU University of Economics and Business, Hanoi, Vietnam

* Correspondence: thekien.edu@gmail.com

Abstract: The study aims at identifying factors affecting Vietnam National University Hanoi students' entrepreneurial intention The samples are 300 students from the VNU University of Social Sciences and Humanities, VNU University of Economics and Business, and VNU University of Technology, Hanoi Based on the application of SEM model, the results show six factors affecting students' entrepreneurial intention including confidence, entrepreurship, education, capital, technological skills, and work skills The higher the factors, the higher the entrepreneurial intention

Keywords: Structural Equation Modeling (SEM); entrepreneurial intention; students; Vietnam National University Hanoi

1 Introduction

Small and Medium Enterprises (SMEs) have long been a contributing factor to the improvement of the country's economy and the nation’s well-being (Hoelscher and Elango 2012; Hatten 2006) Developing countries face poverty, unbalanced development, and unemployment, and among these problems, unemployment is the most blazing problem (Dev and Mahajan 2003) Entrepreneurship plays an important role in developing the economy and creating many job opportunities for workers (Moica et al 2012) According to the statistics of Vietnam Chamber of Commerce and Industry in 2016, new businesses, especially small and medium enterprises, contribute to nearly 50% of GDP and attract more than 90% of new labor Therefore, the development of start-up activities will be a driving force to help solve employment problems and reduce unemployment and turn Vietnamese economy to be more dynamic at the same time

VNU Hanoi also has conducted many activities in recent years to provide advice on starting a business for students According to a study of iPrice group and 500 Startups Vietnam, in 2017, VNU was one of the national leading training units in the number of student startups In March 2017, the President of VNU Hanoi signed a decision to establish the Center for Enabling Startups and Knowledge Transfer with the mission of promoting the commercialization of science and technology products of VNU and other research institution as a bridge to create value from scientific research products and promote entrepreneurship development in VNU and Hanoi The start-up movement among students have been quite dynamic; one of the successful startups is the case of a student of the VNU University of Economics and Business, Nguyen Ngoc Quan, when starting a business called Wine Co., Ltd Vietnam in 2016 Starting from being a CEO of a media company when he was very young, just after 2 years of graduation, Quan's company has now grown into a

medium-sized one with about 100 employees The company operates in marketing and consulting brand strategy & provides communication solutions for many major partners such as: Dairy cow by BOO, Lincup by Lintimate, Vincom Ice Rink, The Elephant Company, Hailecao Studio, etc Another example is a startup search project of the University of Technology - VNU in 2018 when the student group created WORKSVN in two forms: website and mobile application with the goal of connecting businesses with students According to statistics from the Google Play app, up to now, the WORKSVN app has had more than 10,000 downloads and an average of about 2,000 visits per day with user appreciation

However, a large part of students still have poor initiative in finding jobs as well as starting a business The question is what factors affect the VNU students’ entrepreneurial intention? In Vietnam, SEM model is currently applied in many fields, but the number of studies into students’ entrepreneurial intentions using SEM model is quite limited Therefore, this study applies SEM model to analyze factors affecting VNU students’ entrepreneurial intention, thereby making policy suggestions to further promote the

start-up movement among VNU students in particular, and students from the whole country in general

2 Methodology

2.1 Literature review

Intention is a thinking situation that engages personal experience and behavior for a specific purpose or behavior (Gerbing and Anderson, 1988) The theory of Planned Behavior suggests that entrepreneurial intention is the result of the intent and actions of brave individuals described as modern-day heroes (Ajzen, 1987) Based on the analytical framework of GEM, Micozzi and Lucarelli (2016) analyzed data from 37 economies and concluded that gender and confidence in one's skills affect personal mood and strengthen the individual's entrepreneurial intention Gelderen et al (2008) showed two important factors that influence an individual's entrepreneurial intentions based on market awareness and self-improvement Ooi and Nasiru (2015) studied the impact of business education on Malaysian community college students with 235 final-year students as samples drawn from four schools located in northern Malaysia show the important role of universities and community colleges in promoting and nurturing entrepreneurship among graduates Sahinidis et al (2014) indicated that, there is a strong link between the Personal Attraction and Entrepreneurial Intention (EI), similar to the literature findings with other types of samples and also a strong relationship between Perceived Behavioral Control or Self – Efficacy and (EI) Finally, again as previously found in several studies Social Norms and Valuations are expected to affect both directly and indirectly EI Patrick et al (2019) reviewed the entrepreneurship intention literature within the context of business start-up among university students This study also proposes a conceptual model to increase the understanding of the factors hindering undergraduate students from starting a business while studying This model, which enhances knowledge of entrepreneurship intention

among undergraduate students, enriches the theory of Psychological School of Thought and Social Networking Theory in entrepreneurship intention research

In Vietnam, Le (2007) studied the startup decision process of young Vietnamese entrepreneurs The study was conducted with a sample of 159 entrepreneurs who established businesses under the age of 30 during 2000-2006 The research results indicate that young people tend to make the decision on starting a business when meeting all three groups of factors, namely personal qualities, accessibility to resources and business opportunities In addition, the research also shows that entrepreneurial thinking is strongly formed and develops over time, and is influenced by many environmental factors Do (2016) had the study "Factors affecting entrepreneurial intention of Business Administration students at the University of Labor and Social Affairs (Ho Chi Minh City campus)" The study used the entrepreneurial potential model of Krueger and Brazeal (1994) and the theory of planned behavior of Ajzen (1991) The study gathered 315 students at the University The research results indicate five factors influencing students' entrepreneurial intention, namely higher education and training, experience and experiences, family and friends, personality, and capital Do et al (2017), in the study "Analysis of factors affecting the start-up intention of students of Binh Duong University", conducted a survey of 250 samples and applied the linear structural equation model (SEM) for analysis The results show that the student's intention to start a business is influenced by seven factors, of which thre are directly affected by self-awareness, education and training, family and social environment Nguyen and Nguyen (2016), in the study "Factors affecting students’ start-up intention in Tra Vinh University", conducted a survey of 405 students in different disciplines; the results show that factors affecting entrepreneurial intention through the factor of confidence in entrepreneurial feasibility, reflected in teaching activities, extracurricular activities, people’s opinions and business interests which directly affect confidence The higher the confidence about the feasibility of starting a business, the higher the student's intention to start the business

2.2 Analysis methods

In this study, the authors applied the method of structural equation model (SEM), which used SPSS 22.0 software and AMOS version 20 software, via six steps: Cronbach's Alpha reliability analysis, Exploratory Factor Analysis (EFA); Confirmatory Factor Analysis (CFA), Structural Equation Modelling (SEM) analysis, Model estimation test by Bootstrap, Multi-group structure analysis, specifically as follows:

Step 1: Scale reliability test

Cronbach's Alpha coefficient (CA) is used to evaluate the reliability of the scale for each observed variable belonging to the factor groups Peterson (1994) suggested that any factor with CA less than 0.6 should be excluded from the research model (Peterson 1994) According to Bernstein and Nunnally (1994), observed variables with a total correlation coefficient less than 0.3 are considered as irrelevant variables and should be excluded from the model (Bernstein et al., 1994)

Step 2: Exploratory Factor Analysis (EFA)

The EFA allows describing the correlation between the impact variables, called

"factors" EFA is used in cases where the relationship between observed and latent variables

is unclear or uncertain The EFA analysis is conducted in a discoverable manner to determine the extent and degree of relationship between observed variables and the latent factors, which underlie a set of measurements to reduce the number of observed variables loading base factors Meyers et al (2016) reported that in the EFA, the Principal Component Analysis extraction method with Varimax rotation is the most commonly used method A condition for EFA analysis is the following requirements: factor loading > 0.3; 0.5 ≤ KMO ≤ 1; Bartlett test having statistical significance (Sig <0.05); total variance percentage > 50%

Step 3: Confirmatory Factor Analysis (CFA)

CFA can be used when the researcher has some knowledge of the latent variable structure CFA may be the next step of EFA to test whether there is a theoretical model that underlies a set of observations Indicators for measuring the relevance of the model with data include Chi squared (CMIN); Chi square adjusted according to degrees of freedom (CMIN / df); Comparative Fit Index (CFI); Tucker & Lewis index (TLI); Root Mean Square Error of Approximation (RMSEA) According to Hair Jr et al., if 1 <CMIN / df <3, the model

is considered to be a good fit Nguyen (2011) states that if the model receives CFI values, TLI

≥ 0.9; RMSEA ≤ 0.08, and P> 0.5, the model is relevant to the data

Step 4: Structural Equation Modelling (SEM)

The structural equation model (SEM) is the following step of EFA and CFA, helping

to test a set of regression equations at the same time In this study, the SEM model was applied with the aim to identify the influencing factors and the degree of influence of each factor on "VNU students' intention to start up"

2.3 Data collection methods

Determining the sample size is important However, although there is no specific standard in sample selection, the model still reaches a high practical significance with a relatively small number of variables (100-150) (Hoyle 1999; Hoyle and Kenny 1999; Marsh and Hau 1999) With the scope of research equivalent to district and inter-district levels, the practical significance of the data is high and very high at the sample size of about 200 samples (Hoogland and Boomsma 1998; Boomsma and Hoogland 2001; Kline 2005) In some studies analyzing employees’ satisfaction with management, the number of samples was

150 (Muthén 2002) According to Hair et al (1998), in order to perform the EFA factor analysis, the size of the sample applied in the study must be at least 5 times the total number

of observed variables The study has 29 observed variables, so the minimum number of samples is 29 x 5 = 145 Thus, In order to identify the factors affecting entrepreneurial intention, the study conducted in-depth interviews and a structured questionnaire survey for students of the University of Social Sciences and Humanities, University of Economics and Business, and University of Technology at VNU, Hanoi The number of responses collected was 300, of which 296 provided relevant information that could be used for analysis (98.67%)

The questionnaire was designed with closed-ended questions to ask students to assess, compare the current conditions and status of students related to entrepreneurial intention, in aspects related to confidence, technological skills, work experience, entrepreneurship, capital, and education The assessments use the Likert scale with 5 levels:

1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree and 5 = strongly agree

3 Application of SEM model to analyze factors affecting entrepreneurial intention

3.2.1 Evaluating the scale reliability by Cronbach's Alpha

The scale reliability evaluation was done by calculating Cronbach's Alpha coefficients for seven groups of factors through a number of cycles The results obtained (Table 1) show that the CA of all factor groups is greater than 0.6 and the correlation coefficient of the total variable is greater than 0.3 Removing one of the base factors in the group will result in a smaller CA Specifically, Cronbach's Alpha results of "confidence" through 2 cycles showed that variables NT2, NT3, NT4, NT5, and NT6 met the requirements Cronbach's Alpha results of "education " showed that variables GD1, GD2, and GD3 met the requirements Cronbach's Alpha results of " capital" showed that variables NVKD1, NVKD2, and NVKD3 met the requirements Cronbach's Alpha results of

"Entrepreneurship" showed that through two cycles, variables TDKN1, TDKN2, TDKN3, and TDKN5 met the requirements Cronbach's Alpha results of "work experience" showed that variables KNLV1, KNLV2, and KNLV3 were satisfactory Cronbach's Alpha results of

"technological skills" showed that through two cycles, variables KNCN1, KNCN2 and KNCN4 met the requirements Cronbach's Alpha results of "entrepreneurial intention" showed that through three cycles, variables YDKN2, YDKN3, YDKN4, and YDKN5 met the requirements and were retained to analyze the discovery factor in the following round Thus, the results obtained had 25 observable variables and were retained for discovery factor analysis (EFA) in the following round

Table 1: Results of the scale reliability evaluation

coefficient

Cronbach’s Alpha for excluding variables Confidence Cronbach’s Alpha = 0.836

1 NT2 You feel that starting a business is quite easy 0.372 0.829

2 NT3 You think that maintaining the value of your

business is not too difficult

0.461 0.694

3 NT4 You think you have the ability to control the

establishment of a new business

0.461 0.802

4 NT5 You think starting a business brings you more

development opportunities

0.379 0.784

5 NT6 You think you know what needed to start a

business

0.443 0.751

2 Education Cronbach’s Alpha = 0.792

6 GD1 The university provides the knowledge needed for

starting a business

0.510 0.773

7 GD2 The university has business skills training courses 0.361 0.698

8 GD3 My university often organizes career guidance

activities for students (start-up workshops, entrepreneurship contests)

0.476 0.783

3 Capital Cronbach’s Alpha = 0.921

9 NVKD1 I can borrow money from friends and relatives to do

business

0.327 0.749

10 NVKD2 I have the ability to accumulate capital (thanks to

saving and doing a part time job)

0.483 0.861

11 NVKD3 I can raise capital from other sources (banks, credit

funds, )

0.629 0.895

4 Entrepreneurship Cronbach’s Alpha = 0.820

12 TDKN1 I am confident in my ability to start a business 0.429 0.801

13 TDKN2 I have many social relationships 0.532 0.785

14 TDKN3 My relationships can help my start-up 0.460 0.699

15 TDKN5 I am not afraid of risks in business 0.398 0.769

5 Work experience Cronbach’s Alpha= 0.739

6 Technological skills Cronbach’s Alpha = 0.793

19 KNCN1 The 4.0 technology revolution gives businesses an

opportunity to expand their markets

0.425 0.738

20 KNCN2 The 4.0 technology revolution will help businesses

increase their ability to participate in global and

regional value chains

0.392 0.764

21 KNCN4 Entrepreneurs will study the advanced technologies

of Industry 4.0 and apply them to improve their

position in the value chain

0.580 0.691

7 Entrepreneurial intention Cronbach’s Alpha = 0.839

22 YDKN2 I tend to open my own business after graduating 0.542 0.790

23 YDKN3 Starting a business is appealing to me 0.389 0.827

24 YDKN4 I am a person with many business ambitions 0.493 0.819

25 YDKN5 I am confident in my ability to start a business 0.462 0.803

Source: Analyses of collected data

3.2.2 Exploratory Factor Analysis (EFA)

In this step, all 25 variables obtained through Cronbach's Alpha coefficient test ran EFA analysis, and EFA results through a number of qualifying variables were: 0.5 <KMO = 0.832 <1 as shown in Table 2 ; in conclusion, factor analysis is

consistent with actual data

- Testing the correlation between measurement variables (Bartlett’s Test)

+ Test of hypothesis H0: The correlation of the variables is zero;

+ Bartlett’s Test results are valid Sig = 0.000 <0.05; in conclusion, the measurement variables correlate with each other in each factor group;

- Testing of variance extracted of factors (% Cumulative variance):

In the table shows, the total variance extracted (Total Variance Explained) of Component 7 and Cumulative% column has the value of cumulative variance of factors as 56.263%> 50%, which meets the standard In conclusion, 56.263% change of factors is explained by factor's component measurement variables

Table 2 Results of testing KMO and Bartlett’s Test of CFA Kaiser-Meyer-Olkin Measure of Sampling Adequacy 0.832 Bartlett's Test of Sphericity Approx Chi-Square 6840.424

(Source: Analyses of collected data)

- Testing Factor loading

The results of EFA analysis for the independent variables of the factor rotation matrix showed that the factor loading of the measurement variables satisfied the condition when Factor loading was > = 0.5, so all 23 measurement variables were analyzed in the next step

Table 3: Results of Exploratory Factor Analysis (Pattern Matrix)

Factor

NT2 0.751

NT3 0.610

NT4 0.842

NT5 0.603

NT6 0.832

KNLV3 0.628

(Factor loading was at its minimum when > 0.3; significant when > 0.4 and practically significant when > 0.5

(Hair et al 1998) (Source: Analyses of collected data)

3.2.3 Confirmatory Factor Analysis (CFA)

After the EFA analysis, the scales continued to be tested again, using CFA confirmatory factor analysis The aim was to ensure greater reliability and validity of the scale based on a defined number of known factors in order to limit errors in identifying factors The CFA results showed the indicators evaluating the relevance of the theoretical model were GFI = 0.930; TLI = 0.931; CFI = 0.942; RMSEA = 0.047 Therefore, this model was relevant to actual data It can be concluded that components measuring "Student's entrepreneurial intention" and independent factors achieved discriminant validity Therefore, this model was relevant to actual data

3.2.4 Structural Equation Modelling

In this study, SEM model was applied to determine the influencing factors and the degree of influence of each factor on "Student's entrepreneurial intention" The original SEM model was adjusted to be more relevant to the study The model of satisfaction factors is

described in the Figure 1

Figure 1: Model to identify factors affecting VNU students’ entrepreneurial intention

(Source: Analyses of collected data)

The research results showed that the degree of impact of factors affecting VNU student’s startup intentions were: confidence as 0.082; entrepreneurial thinking as 0.106; education as 0.192; capital as 0.89; technological skills as 0.147; and work skill as 0.156 These factors have a positive relation with students' intention to start a business, which means that the factors above will strengthen the student's intention to start a business, thereby promoting students to start a business

4 Conclusion

This study aims to study factors affecting VNU students’ entrepreneurial intention With the structural equation model, the study found a positive relation between the factors

of confidence, entrepreneurship, education, technology skills, capital, and working skills and the formation of entrepreneurial intention The most important affecting factor was education, which reflects the reality because when the university focused on training, organizing seminars, startup idea competitions (such as the Business Challenges 2019 Contest organised by the University of Economics and Business, VNU), the university provided students with necessary knowledge for starting a business, and created a positive environment for the development of startup ideas The factor was followed by factors of work skills, technological skills, entrepreneurship, capital, and confidence The findings from this study are useful references in developing strategies to promote entrepreneurship

However, the study also reveals some limitations; for example, the student survey sample was not enough and did not cover all the universities in VNU, and the study did not compare the impacts of different factors on university students’ entrepreneurial intention

This is the premise for further studies by expanding the scope of research on the factors affecting student entrepreneurial intentions in the future

References

Abu Bakar, S.G., Salwa, K., Amina, M.N.(2014) An assessment of students’ entrepreneurial intentions in Tertiary Institution: a case of Kano State Polytechnic, Nigeria International Journal of Asian Social Science, 4(3), 434-443

Ajzen (1991) The Theory of Planned Behavior Organizational Behavior and Human Decision Processes 50, 179-211

Alexandros G Sahinidis, Evangelos E Vassiliou, Alina Hyz (2014) “ Factors Affecting Entrepreneurs' Intention to Start a New Venture: An Empirical Study” International Journal on Strategic Innovative Marketing, 1, 148-162

Amou & Alex (2016) Theory of Planned Behavior, Contextual Elements, Demographic Factors and Entrepreneurial Intentions of Students Kenya European Journal of Business and Management, 6(15), 167-175

Arenius, P., and Minniti, M (2005) Perceptual variables and nascent entrepreneurship Small Business Economics, 24 (4), 233-247

Bernstein, I.H and J.C Nunnally Psychometric theory (1994) New York: McGraw-Hill Oliva, TA, Oliver, RL, & MacMillan, IC (1992) A catastrophe model for developing service satisfaction strategies Marketing, 56, 83-95

Binks, M., K Starkey, et al (2006) Entrepreneurship education and the business school Technology Analysis and Strategic Managemenet, 18, (1), 1-18

Bird, B (1988) Implementing entrepreneurial ideas: The case for intention Management Review, 13(3), 442-453

Cho, B (1998) Study of Effective Entrepreneurship, Education Method and Its Process, Business Education Research, 2 (1), 27 – 47

Clark, B W., Davis, C., & Harnish, V (1984) Do Courses in Entrepreneurship Aid in New Venture Creation, Journal of Small Business Management, 22(2), 26 – 31

Dev, M.S and Mahajan, V (2003) Employment and unemployment Economic and Political Weekly, 38(12), 1252-1261

Dilts, J C and S M Fowler (1999) "Internships: Preparing students for an entrepreneurial career." Journal of Business and Entrepreneurship, 11(1), 51-63

Dyer, W.G (1994) Toward a theory of entrepreneurial careers Entrepreneurship Theory and Practice, 19(2), 7-21

Hair, Joseph F (1998) Multivariate Data Analysis, New Jersey, Prentice Hall, 730 Hart, S L (1992) An Integrative Framework for Strategy Making Process Management Review, 17, 327 - 351