Application of Entrepreneurial Intention Model in Comparative International Entrepreneurship Research: A Cross-Cultural Study of India and Saudi Arabia

Abstract

The development of entrepreneurial activities helps in realization of multiple goals of sustainable development. This study aims to inspect the applicability of the entrepreneurial intention model (EIM) to explain entrepreneurial intention in different cultural contexts. A survey was conducted using the Entrepreneurial Intention Questionnaire. Based on 535 valid responses received from business students in India (N = 285) and Saudi Arabia (N = 250), the model was analyzed using the partial least square regression method. The results show that the EIM is validated for the combined sample (R²_combined = 0.522) as well as for the individual samples (R²_India = 0.644; R²_Saudi = 0.412). Demographic variables and human capital variables were found to impact the entrepreneurial intention differently in the combined sample and the individual samples of the two countries. Multigroup analysis was performed to test if the country variable influences the hypothesized paths in the model. The results of the multigroup analysis show that many of the hypothesized relationships (gender-personal attitude; gender-subjective norm; gender-perceived behavioral control, role model-subjective norm, role model-perceived behavioral control, subjective norm-personal attitude, and subjective norm-entrepreneurial intention) are significantly different in the two countries. This study addresses the gap in comparative international entrepreneurship (CIE) research by testing the applicability of EIM in two different cultures. The findings are also significant for policymakers at both national and international levels for facilitating entrepreneurship.

1. Introduction

Economic development and entrepreneurship have been considered to be closely related to each other [1,2]. Entrepreneurs contribute to a nation’s growth in many ways such as innovation, employment generation, better resource utilization, wealth distribution, and societal well-being [3,4]. The entrepreneurial activities may help in realization of multiple sustainable development goals such as alleviate poverty, gender equality, bridge inequality among countries. Hence, it has been specifically targeted under the goals of equitable education and equitable work environment [5].

The role of entrepreneurship in the socio-economic development of developing countries is well supported in research [6,7]. A popular strand of entrepreneurship research has been the understanding of reasons why an individual decides to become an entrepreneur [8,9,10,11]. Research has supported the theory that being an entrepreneur is a planned action that is preceded by a “cognitive state” [12], famously known as “intention” [2,13]. Ajzen [13] is credited with theorizing the relationship between “intention” and “action”, popularly called the Theory of Planned Behavior. This theory has been applied to entrepreneurship as well and has been found to successfully explain the intention to become an entrepreneur or the “Entrepreneurial Intention” [11,13,14]. Entrepreneurship research has found several other theoretical models that help explain EI, including the Entrepreneurial Intention Model developed by Linan and Chen [14]. EIM has been adapted from the theory of planned behavior by adding the demographic and human capital variables in addition to including the indirect effect of subjective norms on entrepreneurial intention.

Research on entrepreneurial intention of students is overwhelming [1,2,15,16,17,18]. One of the major reasons for this is the emphasis of policymakers on increasing new businesses to overcome youth unemployment and economic growth [2]. Policymakers have been targeting university students to promote entrepreneurship because they can be easily influenced during their career selection [19]. More specifically, business graduates have been found to be more ambitious in starting their own venture [19]. Therefore, student population has been the prime target of entrepreneurial research [20].

Maalaoui et al. [21] classified EI research into three types namely, studies exploring antecedents of EI; explaining the conversion of intention to action; and improving the model by adding new variables. Studies have found the role of culture in entrepreneurial behavior as well as new venture creation [22,23]. A cross-cultural understanding of entrepreneurial behavior also has a “cognitive and utilitarian value” [24]. It may provide deeper insights into the “causative factors of entrepreneurship” in different national and cultural contexts [25].

A particular strand of entrepreneurship research dealing with a cross-cultural or international perspective has been named as comparative international entrepreneurship (CIE) [24,26]. It is characterized by its focus on entrepreneurial behaviors in different countries and cultures [1,14,27,28,29]. The CIE research is a part of the broad area of international entrepreneurship (IE). Several studies have pointed to the need for more research in all the areas of IE, including CIE. More specifically, researchers point to the need to focus on neglected/under-represented countries/regions in the CIE research [25]. Therefore, there is a need to consider the student population from developing and under-developed countries. The cross-cultural studies can test and establish the generalizability of the models [14,30,31]. The role of culture in career intentions of students is also required to understand the predictors of intentions [32]. The article [33] mentioned compelling reasons for conducting a cross-cultural validation of the entrepreneurial intention model. A cross-cultural examination of the model will reveal if the antecedents are or are not affected by culture or the country of origin.

Considering university student populations in two developing countries namely Saudi Arabia and India, the present study addresses the need for more cross-cultural research on entrepreneurial intention. The current research has used Linan and Chen’s Entrepreneurship Intention Model [14]. The model along with the instrument have been tested in a variety of samples. The main research questions (RQ) dealt in the current study may be listed as follows:

RQ1: Can the entrepreneurial intention model (EIM) explain entrepreneurial intention of students in different cultural contexts?

RQ2: Does culture affect the relationship between the main antecedents of entrepreneurial intention namely “personal attitude (PA), perceived behavioral control (PBC), and subjective norms (SN) with entrepreneurial intention (EI)”?

RQ3: How do demographic and human capital variables included in the EIM affect the antecedents of entrepreneurial intention (EI) in general and in cultural perspectives?

The rest of the paper is organized as follows. Section 2 presents the literature review. Section 3 describes the materials and methods. Section 4 illustrates the analysis and results. Section 5 completes the article with discussions and conclusion.

2. Literature Review

2.1. Entrepreneurial Intention Model

Entrepreneurial intention (EI) as a behavioral construct has several definitions. It has been defined as “the attempt to create new businesses, including self-employment or the expansion of an existing business by an individual, a team of individuals or an already established business” [32,34]. Entrepreneurial intention has been more precisely defined in terms of commitment to start a business [35]. Entrepreneurship research has found several theoretical models that help in explaining EI. Some of the models identified in the literature survey are “the Theory of Planned Behavior (TPB) [13]; Luthje and Franke’s model [36]; Entrepreneur Event Model [37]; social networking theory [38]”. Among these models, TPB has been the most popular [39,40,41]. Theory of planned behavior has generally been validated in EI research [12]. It has also been applied in cross-cultural research [42,43], as well as in longitudinal studies [12]. This theory explains intention to be formed by three determinants, namely attitude toward the behavior, subjective norm and perceived control over the behavior. Attitude toward behavior refers to the experience or appraisal of the individual towards the proposed action. Subjective norm depends upon the perception of others towards that action. The third determinant perceived behavioral control is the individual’s perception of the viability to perform that action. Lortie and Castogiovanni [44] mention that TPB has generally been used in adapted form in the EI research. Researchers added or adapted the TPB model to explain EI in different contexts [45,46,47].

The entrepreneurial intention model (EIM) is an adaptation of TPB developed by Linan and Chen [14]. The antecedents of EI are called “personal attitude (PA), perceived behavioral control (PBC) and subjective norms (SN)”. Personal attitude means the personal liking of an individual for entrepreneurship, perceived behavioral control implies the firm belief of an individual about his/her success as an entrepreneur, and subjective norms are the notions an individual has about others’ perceptions about entrepreneurship [32].

2.1.1. Personal Attitude and Entrepreneurial Intention

Personal attitude or attitude towards entrepreneurship is the “personal valuation”, “tendency”, or attractiveness of an individual in opting for entrepreneurship [13,42]. This valuation may be positive or negative, but it may vary in degree based on the contextual factors [14]. Vamvaka [48] equated entrepreneurship with self-employment that is different from being “organizationally employed”. Research shows a direct and positive relationship between attitude and EI [4,12,18,49,50]. It has been mentioned that this antecedent of EI comes into play even before a student or an individual starts evaluating business ideas [50]. Further it is elaborated that this factor represents an individual’s expectation of positive results, and people with such evaluation of entrepreneurship may create their own venture [51]. An individual’s expectations that lead to the personal attitude can also vary. These expectations or rewards are termed “intrinsic” and “extrinsic” [51].

As one of the three antecedents of EI in the EIM, we hypothesize the following:

Hypothesis 1 (H1). 

Personal Attitude positively affects entrepreneurial intention.

2.1.2. Subjective Norms and Entrepreneurial Intention

Subjective norm is the perception of an individual regarding the approval or disapproval of others for their decision to start their own business [2]. It is also a perception of the “social appropriateness” for engaging in entrepreneurship [52]. This factor includes seeking “suggestions and support” from an individual’s close connections [20]. Extant research has reported mixed results regarding the direct impact of this antecedent on EI [2]. Some confirm it [27,51] while others reported no significant impact [6,14,15,53]. In a cross-cultural study [43] that included twelve countries, subjective norms showed significant effect.

We therefore hypothesize the following:

Hypothesis 2a (H2a). 

Subjective norm positively affects entrepreneurial intention.

Because the impact of subjective norms on EI is not universally established, researchers also considered the indirect effect of subjective norms. In the entrepreneurial intention model, Linan and Chen tested the effect of subjective norm on the factors personal attitude and perceived behavioral control [14]. These two relationships were included in the adapted model of EI called EIM.

We therefore hypothesize the following:

Hypothesis 2b (H2b). 

Subjective norm (SN) positively affects personal attitude (PA).

Hypothesis 2c (H2c). 

Subjective norm (SN) positively affects perceived behavioral control (PBC).

2.1.3. Perceived Behavioral Control and Entrepreneurial Intention

The third antecedent of EI included in the EIM is called the “perceived behavioral control”. It is the perception of ease or difficulty to start a business [2,13]. Perceived behavioral control is said to be “situational and behavior-specific” [54], and it represents the controllability of the behavior [51]. More simply, this factor represents the ability to do business and control over it [52]. Further, the study [2] highlights the importance of this antecedent in awareness and control of the factors related to success of a venture. Positive impact of perceived behavioral control on EI is reported in numerous studies [12,27,48,55,56].

We therefore hypothesize the following:

Hypothesis 3 (H3). 

Perceived behavioral control positively affects entrepreneurial intention.

2.2. Demographic Factors, Human Capital and Entrepreneurial Intention

Apart from the adaptation in the relationship of subjective norms with EI, the EIM also included demographic factors and human capital [14]. Demographic and human capital variables included in the model are “gender, role model, work experience and years of business education”. Each of these factors are discussed below.

2.2.1. Gender

The impact of gender on EI has been found to be significant [27,57,58]. A study dealing with the EI of students in UAE [59] also confirmed that gender affects EI. Research suggests that among the male and female population, the interest for entrepreneurship is higher in males [48,60]. They also add that the males show higher perception towards all the three antecedents of EI compared to females.

Consequently, we hypothesize the following:

Hypothesis 4a (H4a). 

Gender positively affects personal attitude.

Hypothesis 4b (H4b). 

Gender positively affects subjective norm.

Hypothesis 4c (H4c). 

Gender positively affects perceived behavioral control.

2.2.2. Role Model

Role models, in terms of career decisions, are defined as “those who can encourage others to pursue certain career paths or pursue certain goals” [61]. A historical detail of role models shared by Yang [62] shows that the term was first introduced by Robert K Merton in 1957. Bosma et al. [63] defined role models as “individuals who provide examples for other people to emulate, and who stimulate or inspire to make decisions of others’ career as well as attain certain goals”. The role models inspire others, and they may be categorized as “partial, stage, option, charismatic and negative” [64].

The influence of role models in entrepreneurship has been proven in research [63]. These role models may also help in the development of the enterprise. Several entrepreneurial studies have highlighted the influence of role models in entrepreneurial intention [65,66].

Earlier studies that included this variable in the entrepreneurial intention models are by [35,67]. In Linan and Chen’s EIM, this variable was included as one of the human capital variables [14]. It is empirically established in entrepreneurial research that role models do influence an individual’s EI significantly [65,68,69,70]. Role models may impact the “cognitive or emotional process” of an individual that is the antecedent of EI included in EIM [68,71,72]. Further, role models also provide information as well as means to tackle frustration. Fellnhofer [72] also suggested that even the online role models can positively impact an individual’s EI.

Consequently, we hypothesize the following:

Hypothesis 5a (H5a). 

Role model positively affects personal attitude.

Hypothesis 5b (H5b). 

Role model positively affects subjective norm.

Hypothesis 5c (H5c). 

Role model positively affects perceived behavioral control.

2.2.3. Self-Employment Experience

Self-employment or prior entrepreneurial experience implies the exposure of an individual to entrepreneurship [73]. This factor has been included as a predictor of EI in entrepreneurship research and is suggested to have influence on EI [74,75]. Hockerts [76] studied the impact of prior experience in intending to start a social enterprise. A type of prior experience through family business was also considered in research [77]. Gird and Bagraim [78] confirmed the positive influence of self-employment experience on EI of students. Linan and Chen [14] included this variable as a predictor of the three antecedents of EI.

Accordingly, we hypothesize the following:

Hypothesis 6a (H6a). 

Self-employment experience positively affects personal attitude

Hypothesis 6b (H6b). 

Self-employment experience positively affects subjective norm.

Hypothesis 6c (H6c). 

Self-employment experience positively affects perceived behavioral control.

2.2.4. Business Education

Business education includes learning various aspects of business including entrepreneurship that may also be a specialization. Direct effect of business education on EI has not been established [79], but its impact on the antecedents of EI has been confirmed in several studies [80,81,82]. Research suggests that the influence of business education on EI may vary by country [83]. The number of years an individual is exposed to business and entrepreneurial education is an important measure of this variable.

Accordingly, we hypothesize the following:

Hypothesis 7a (H7a). 

Years of business education positively affects personal attitude.

Hypothesis 7b (H7b). 

Years of business education positively affects subjective norm.

Hypothesis 7c (H7c). 

Years of business education positively affects perceived behavioral control.

2.3. Comparative International Entrepreneurship

International Entrepreneurship (IE) is the strand of research dealing with entrepreneurial behavior and activities across nations and borders [84]. IE research is a new area that has gained prominence in the last decade [24]. It deals with entrepreneurship across borders and includes “internationalisation and international comparisons of domestic entrepreneurship in many countries” [85]. This research area has also been projected to be called an “independent scientific discipline” [23,84,86]. Zahra [87] made a notable contribution to the IE research by extending the focus beyond just internationalization of enterprises.

Comparative international entrepreneurship (CIE) is a field of entrepreneurship research that deals with entrepreneurship in different countries [24,26]. McDougall and Oviatt [88] were the early proponents of CIE research. Glodowska mentions that CIE research was first included in the area of IE research by Jones [23]. They also add that in terms of chronology, the order in which IE research progressed is “internationalization, comparisons of entrepreneurship and comparative entrepreneurial internationalization” [23].

Two typologies of CIE research were found in the literature survey. The first typology, proposed by Oviatt and McDougall [85], was included in a review article by Terjesen [26]. As per this typology, CIE research may be divided into either a comparison of entrepreneurs or their behaviors. Caviello [89] classified CIE research into three types, namely “Type A: Entrepreneurial internationalization; Type B: International comparisons of entrepreneurship (countries and cultures); and Type C: Comparative entrepreneurial internationalization”. [84] used the IE research typology given by Jones (2011) and IE research field classification by [90] to review articles on IE. They found that the share of CIE research is just 14.8%.

A lack of consistency in CIE research has been noted leading to a “scientific immaturity” of the IE research [24]. However, they emphasize that it is an extremely prospective research area. A systematic review by Terjessen, et al. [26] is an important contribution in the literature on CIE.

Comparison of India and Saudi Arabia

India and Saudi Arabia are two prominent developing countries of the Asian continent. They have similarities as well as dissimilarities (Table 1). The population of the two countries are significantly different. India has a population of 1.33 billion people whereas Saudi Arabia has a population of 34.3 million [91]. The real gross domestic product growth for 2020–21 for India and Saudi Arabia is 8.8% and 3.1%, respectively [92]. This means that India is growing at a faster pace than Saudi Arabia. The Indian economy is much more diversified compared to that of Saudi Arabia. The International Conference of Labor Statisticians (ICLS) consider the unemployed as “those persons of working age who are without work, seeking work (carried out activities to seek employment during a recent past period), and currently available for work” [93]. The unemployment rate for India and Saudi Arabia are almost the same at 5.4% and 5.9%, respectively [91]. India has been categorized as a lower-middle-income country, whereas Saudi Arabia is a high-income country [94]. Both countries have been trying to reduce unemployment by supporting entrepreneurship. The “total early-stage entrepreneurship” activity is defined as the percentage of the adult population engaged in entrepreneurship [94]. The percentages for India and Saudi Arabia are 14.97% and 13.96%, respectively.

Global Entrepreneurship Monitor (GEM) has developed a national entrepreneurship context index (NECI) that is a composite rating of various countries based on twelve entrepreneurship conditions [94]. The National Entrepreneurship Context Index (NECI) was created by Global Entrepreneurship Monitor in 2019. It represents “the average state of the entrepreneurial environment conditions of a country or territory in which it has been measured” [95]. It is also measured by the NES questionnaire. It reflects the “ease of starting and developing a business” in a particular country [96]. The NECI index includes twelve pillars, namely “entrepreneurial finance, government policies (support and relevance), government policies (taxes and bureaucracy), government entrepreneurship programs, entrepreneurial education at school stage, entrepreneurial education at post-school stage, research and development (R&D) transfer, commercial and legal infrastructure, internal market dynamics, internal market burdens, physical infrastructure and cultural and social norms” [95]. On this index, the current ranking of India is 5th with a score of 5.8, and Saudi Arabia is ranked 17th with a score of 5.04.

Based on ranking by World Economic Forum, India ranks 68th whereas Saudi Arabia ranks 36th in “global competitiveness” [94]. These data show that both countries are comparable on many parameters of entrepreneurship conditions.

A comparison of the two countries on Hofstede’s cultural dimensions, Hofstede, Hofstede, and Michael, 2010, shows that they differ from each other on these dimensions but to varying degrees. The six dimensions in Hofstede’s framework are “individualism/collectivism, power distance, uncertainty avoidance, masculinity/femininity, long-term/short-term orientation, and indulgence/restraint” [97]. It may be inferred by looking at the scores that the two countries are quite different from each other on most of the six dimensions.

Table 1. Comparison of India and Saudi Arabia.

	India	Saudi Arabia
Population (million) [98]	1.33 billion	34.3 million
Real GDP forecast [98]	8.8%	3.1%
GDP per capita [98]	US$2099	US$23,139
UN classification [99]	Lower middle income	High Income
Youth unemployment [100]	5.4%	5.9%
Total Entrepreneurial Activity Rate [101]	14.97%	13.96%
National Entrepreneurship Context (NECI) Index [101]	5.8	5.04
Global competitiveness ranking [102]	68	36

Table 1. Comparison of India and Saudi Arabia.

	India	Saudi Arabia
Population (million) [98]	1.33 billion	34.3 million
Real GDP forecast [98]	8.8%	3.1%
GDP per capita [98]	US$2099	US$23,139
UN classification [99]	Lower middle income	High Income
Youth unemployment [100]	5.4%	5.9%
Total Entrepreneurial Activity Rate [101]	14.97%	13.96%
National Entrepreneurship Context (NECI) Index [101]	5.8	5.04
Global competitiveness ranking [102]	68	36

Studies comparing different countries or regions identified through the literature review are listed in Table 2.

Differences in the model elements or variances have been reported in cross-cultural studies on EI [43]. More studies are required to test the applicability of EIM in CIE research [28,103]. This study aims to investigate the applicability of EIM in CIE research by analyzing the influence of a country on various hypothesized paths.

Table 2. Cross-cultural studies on EI.

Authors	Countries	Theoretical Model
DePillis 2007 [104]	USA and Ireland	Personality traits
Paul 2016 [31]	India and Japan	Self-developed multifactor
Linan 2006 [105]	Spain and Taiwan	EIM
Uslay 2002 [106]	USA, Turkey and Spain	TPB
Cruz et al. [103]	Brazil and Portugal	TPB
Autio 1997 [42]	USA, Finland and Sweden	TPB
Bogatyreva 2019 [54]	Various countries of Europe and Asia.	Self-developed based on Hofstede
Moriano 2012 [107]	Various countries of Europe and Asia	TPB
Hueso 2020 [30]	UK and Spain	TPB
Trivedi 2017 [9]	India Malaysia Singapore	Entrepreneurial Intention Constraint Model (EICM)
Martínez-González 2019 [32]	Spain and Poland	TPB
Nieuwenhuizen 2015 [108]	South Africa and Poland	TPB
Bouncken 2009 [109]	Germany and Poland	Self-developed based on Hofstede
Ozaralli 2016 [33]	USA and Turkey	TPB
Brancu 2015 [110]	Romania and Iceland	Self-developed based on Hofstede
Kibler 2019 [111]	Austria and Finland	TPB
Engle 2010 [43]	Various European, Asian countries and USA.	TPB

Table 2. Cross-cultural studies on EI.

Authors	Countries	Theoretical Model
DePillis 2007 [104]	USA and Ireland	Personality traits
Paul 2016 [31]	India and Japan	Self-developed multifactor
Linan 2006 [105]	Spain and Taiwan	EIM
Uslay 2002 [106]	USA, Turkey and Spain	TPB
Cruz et al. [103]	Brazil and Portugal	TPB
Autio 1997 [42]	USA, Finland and Sweden	TPB
Bogatyreva 2019 [54]	Various countries of Europe and Asia.	Self-developed based on Hofstede
Moriano 2012 [107]	Various countries of Europe and Asia	TPB
Hueso 2020 [30]	UK and Spain	TPB
Trivedi 2017 [9]	India Malaysia Singapore	Entrepreneurial Intention Constraint Model (EICM)
Martínez-González 2019 [32]	Spain and Poland	TPB
Nieuwenhuizen 2015 [108]	South Africa and Poland	TPB
Bouncken 2009 [109]	Germany and Poland	Self-developed based on Hofstede
Ozaralli 2016 [33]	USA and Turkey	TPB
Brancu 2015 [110]	Romania and Iceland	Self-developed based on Hofstede
Kibler 2019 [111]	Austria and Finland	TPB
Engle 2010 [43]	Various European, Asian countries and USA.	TPB

Therefore, we hypothesize the following:

Hypothesis 8 (H8).

Country significantly affects the paths in the entrepreneurial intention model.

3. Materials and Methods

3.1. Survey Instrument: Design and Procedure

To measure EI, the entrepreneurial intention questionnaire [14] was used. This scale measures the four constructs of TPB, namely “personal attitude, subjective norm, perceived behavioral control, and EI”. A number of studies have validated EIQ [28,49]. In India, the available English version of EIQ was employed. For Saudi Arabia, the Arabic version of the instrument was developed with the help of an expert Arabic translator. Thereafter, the instrument was pilot tested with 10 students to establish the validity of the constructs and its items. Based on their input, slight modification was done to generate an exact response.

3.2. Target Population

The target population consists of business management students enrolled in bachelor and master programs in India and Saudi Arabia. Their approximate numbers are estimated to be 713 thousand in India [112] and 430 thousand in Saudi Arabia [113,114]. Therefore, the total targeted population in both countries is 1143 thousand.

3.3. Sample Size

The minimum required sample size was calculated with the following formula:

$$n=\frac{z^2\left(p\right)\left(q\right)}{e^2}$$

(1)

where, n = sample size; z = 1.96 (z table value with 0.05 significance level); p = prevalence level 0.5 (prevalence level is unknown); q = 1 − p; e = error term 0.05 (significant level 0.05)

The value of n was calculated with above mentioned values as follows:

$$n=\frac{{1.96}^2\left(0.5\right)\left(0.5\right)}{{0.05}^2}$$

(2)

$$n=384.16$$

(3)

The value of minimum required sample size was further validated with the online Raosoft calculator [115]. The required minimum sample size for the target population of 1143 comes out to be 385 that is approximately the same as in Equation (3).

3.4. Sampling Techniques

The nonprobability convenient sampling was used to extract the required number of respondents from the target population.

3.5. Inclusion Criteria

i.

Valid response from the perspective of data accuracy

ii.

Response from the targeted population

3.6. Exclusion Criteria

• Incomplete response
• Duplicate response
• Students from other programs

3.7. Time Frame for Online Survey Process

Initially, the link of the survey using Google Form was distributed among target students in both countries through the academic staff using emails. The staff intern shared the link with their students by various means such as emails, learning management systems and social networking. For India in the first attempt in the month of June 2020, 50 percent of the data was collected, and in the second attempt in the month of July 2020, the remaining data was collected. Similarly, in Saudi Arabia in the first attempt within the first and second weeks of the month of June 2020, most of the responses were received.

3.8. Statistical Tools for Data Analysis

The statistical tools used for the data analysis are as follows:

• Descriptive statistics for frequency and percentage with the help of SPSS to know the sociodemographic comparison of respondents.
• Measurement model was assessed with the help of outer loading (OL ≥ 0.7), composite reliability (CR ≥ 0.85), average variance extracted (AVE ≥ 0.5) and Fornell–Larcker discriminant validity using SmartPLS [116].
• Structural model was evaluated using coefficient of determination (R- square) for endogenous variables, path coefficients, and statistical significance through bootstrapping [117].
• Multigroup analysis
  • A three step (configural invariance, compositional invariance, and invariance of mean and variance) for measurement of invariance of composite model (MICOM) analysis as a precondition for multigroup analysis is conducted [118,119,120].
  • Multigroup analysis for comparison of both countries using path coefficient of both subsamples, their differences and significance of difference.

4. Analysis and Results

Firstly, the descriptive statistics for frequency and percentage were analyzed with the help of SPSS software to know the sociodemographic comparison of respondents. Thereafter, the EIM as represented in Figure 1 was analyzed with SmartPLS (v3.3.2) software (SmartPLS GmbH, Bonningstedt, Germany) [121]. Five routines of this software were utilized, namely PLS algorithm, bootstrapping, blindfolding, permutation, and multigroup analysis. The settings of the routines were kept default except for a maximum number of iterations for the PLS algorithm set at 1000 and samples for bootstrapping set at 5000 and samples for permutation set at 5000. The overall analysis was conducted in two steps. First, the measurement model was assessed for “reliability and validity” based on the total sample. In the second step, the structural model was assessed. Finally, the permutation and multigroup analysis was done to test the differences in the samples of the two countries.

4.1. Descriptive Statistics

A total of 555 responses were received from various universities in Saudi Arabia and India. As mentioned earlier, the application of inclusion and exclusion criteria finally yielded 535 valid responses. The summary of responses is presented in

Table 3. Descriptive statistics.

		India		Saudi Arabia		Total
		Count	Percent	Count	Percent	Count	Percent
Gender	Male	170	59.60%	158	63.2%	328	61.3%
	Female	115	40.40%	92	36.8%	207	38.7%
Self-employment	Yes	73	25.60%	70	28.0%	143	26.7%
	No	212	74.40%	180	72.0%	392	73.3%
Role model	Yes	169	59.30%	139	55.6%	308	57.6%
	No	116	40.70%	111	44.4%	227	42.4%
Years of Business Education	1–2	189	66.30%	168	67.2%	357	66.7%
	3–4	61	21.40%	79	31.6%	140	26.2%
	5 and above	35	12.30%	3	1.2%	38	7.1%

. A total of 61.3% of the respondents were male, whereas 38.7% were female. The majority of respondents (73.3%) did not possess self-employment experience. A total of 57.6% of respondents claimed to have a role-model. A total of 66.7% of respondents had received one to two years of business education, while the remaining had three to four years or more. Furthermore, the subsamples of India and Saudi Arabia are similar. The Pearson’s chi-square tests for gender (χ2 (1, N = 535) = 0.556, p = 0.451), self-employment (χ2 (1, N = 535) = 0.275, p = 0.600), and role model (χ2 (1, N = 535) = 0.602, p = 0.438) reveal that the difference between two subsamples is not significant.

4.2. Measurement Model

Analysis of the measurement model showed that the 19 indicators out of 20 have loadings above 0.7. Only PA1 has a value of 0.604 and is dropped. The model is rerun after removing the PA1, and results are reported in

Table 4. Measurement Model: Reliability and Validity (Total Sample).

Construct/Indicators	Outer Loadings *	Composite Reliability *	AVE *
Entrepreneurial Intention		0.930	0.741
EI1	0.819
EI2	0.861
EI3	0.873
EI4	0.909
EI5	0.847
EI6	0.852
Personal Attitude		0.865	0.713
PA2	0.843
PA3	0.788
PA4	0.878
PA5	0.865
Perceived Behavioral Control		0.886	0.639
PBC1	0.730
PBC2	0.855
PBC3	0.865
PBC4	0.827
PBC5	0.779
PBC6	0.727
Subjective Norms		0.763	0.678
SN1	0.792
SN2	0.853
SN3	0.824

Note: * Recommended values of outer loading (OL ≥ 0.7), composite reliability (CR ≥ 0.85), average variance extracted (AVE ≥ 0.5).

. All items were found to be significant based on the bootstrapping procedure. The values of AVE and composite reliability (shown in Table 4) show acceptable loading for all of the items. The composite reliability scores are above 0.85 [116]. The values of AVE were above 0.5, indicating acceptable convergent validity [116].

The discriminant validity is presented in

Table 5. Discriminant validity (Total Sample).

	EI	G	PBC	PA	RM	SEE	SN	YBE
Entrepreneurial Intention (EI)	0.861
Gender (G)	0.151	1
Perceived Behavioral Control (PBC)	0.498	0.11	0.799
Personal Attitude (PA)	0.655	0.059	0.353	0.844
Role Model (RM)	0.178	0.071	0.157	0.066	1
Self-employment Experience (SEE)	0.149	0.107	0.171	0.141	0.125	1
Subjective Norms (SN)	0.274	0.05	0.194	0.339	0.015	0.023	0.823
Years of Business Education (YBE)	0.009	−0.126	0.139	−0.022	0.153	0.006	0.03	1

. The matrix comprises AVE values in diagonal positions, and below, squared correlations between each pair of constructs are given. The discriminant validity is adequate because the diagonal values are more than the values below (Fornell–Larcker criterion) [116].

4.3. Structural Model

As described above, the reliability and validity of the measurement model is established. Therefore, analysis of the structural model is done at this stage. In this procedure, “coefficient of determination (R-square) for endogenous variables, path coefficients, and statistical significance through bootstrapping” are analyzed. This analysis was done for the total sample as well as the subsamples of each country. The results based on the total sample are shown in Figure 2. R-square values are present inside the circles that range from moderate to weak in strength [116]. Furthermore, the results of the total sample, and Indian and Saudi subsamples are shown in

Table 6. Direct effects. Bias-corrected 95% confidence intervals and indirect effect multigroup comparison results.

(Hypotheses) Paths	Total Sample			India			Saudi Arabia
	Path	t	p-Value	Path	t	p-Value	Path	t	p-Value
(H4a) Gender→PA	0.020	0.498	0.618	0.122	2.150	0.032	−0.124	2.204	0.028
(H5a) RM→PA	0.049	1.181	0.238	0.037	0.643	0.520	0.031	0.465	0.642
(H6a) SEE→PA	0.125	3.504	0.000	0.177	3.804	0.000	0.067	1.193	0.233
(H7a) YBE→PA	−0.037	0.869	0.385	−0.086	1.597	0.110	0.042	0.664	0.506
(H4b) Gender→SN	0.053	1.174	0.240	0.129	2.089	0.037	−0.064	0.758	0.448
(H5b) RM→SN	0.004	0.079	0.937	0.094	1.538	0.124	−0.090	1.158	0.247
(H6b) SEE→SN	0.016	0.363	0.717	−0.022	0.361	0.718	0.027	0.389	0.697
(H7b) YBE→SN	0.036	0.854	0.393	0.030	0.544	0.586	0.025	0.352	0.725
(H4c) Gender→PBC	0.094	2.251	0.024	0.229	4.237	0.000	−0.029	0.449	0.653
(H5c) RM→PBC	0.110	2.519	0.012	0.001	0.026	0.979	0.179	2.964	0.003
(H6c) SEE→PBC	0.142	3.432	0.001	0.135	2.657	0.008	0.210	3.605	0.000
(H7c) YBE→PBC	0.128	3.212	0.001	0.087	1.740	0.082	0.197	3.023	0.003
(H2b) SN→PA	0.336	6.886	0.000	0.393	6.454	0.000	0.194	2.059	0.040
(H2c) SN→PBC	0.181	3.913	0.000	0.396	6.810	0.000	0.116	1.605	0.108
(H1) PA→EI	0.550	13.189	0.000	0.566	9.638	0.000	0.500	8.580	0.000
(H2a) SN→EI	0.030	0.686	0.493	0.136	2.342	0.019	−0.131	2.229	0.026
(H2) PBC→EI	0.298	7.923	0.000	0.242	4.395	0.000	0.318	5.664	0.000

and

Table 7. R² and Q² indicators (in Total N = 535, in India N = 285, in Saudi Arabia N = 250).

Total Sample			India		Saudi Arabia
	R2	Q2	R2	Q2	R2	Q2
Personal Attitude (PA)	0.137	0.094	0.234	0.169	0.065	0.030
Subjective Norms (SN)	0.004	−0.002	0.028	0.012	0.014	−0.002
Perceived Behavioral Control (PBC)	0.105	0.064	0.270	0.168	0.142	0.074
Entrepreneurial Intention (EI)	0.522	0.381	0.644	0.516	0.412	0.263

. The paths with a value of p > 0.05 are significant. A negative value seen in the case of gender-personal attitude indicates a negative influence. Further, the value of cross-validated redundancy index (Q²) is more than 0 in all the samples except for subjective norms (shown in Table 7. The values indicate that “the predictive relevance of the model” is established [122]. In the Indian subsample, the predictive relevance of the subjective norm is established, but in the total sample and Saudi subsample, it is not established.

4.4. Multigroup Analysis

The “measurement of invariance of composite model (MICOM)” analysis is a precondition for multigroup analysis [118]. MICOM establishes the measurement (outer) model invariance essential to carry out the multigroup analysis. This analysis is done in three steps, namely “configural invariance, compositional invariance, and invariance of mean and variance”. The usage of SmartPLS itself ascertains the first step, the configural invariance [119]. The results of step two, the compositional invariance, are shown in

Table 8. Measurement invariance test: step two compositional invariance.

Constructs	Original Correlation	Correlation Permutation Mean	5.0% Quantile	Permutation p-Values	Compositional Invariance
Entrepreneurial Intention	1.000	1.000	1.000	0.122	Yes
Gender	1.000	1.000	1.000	0.385	Yes
Perceived Behavioral Control	0.999	0.999	0.997	0.564	Yes
Personal Attitude	1.000	1.000	0.999	0.388	Yes
Role Model	1.000	1.000	1.000	0.148	Yes
Self-employment Experience	1.000	1.000	1.000	0.122	Yes
Subjective Norms	0.999	0.996	0.987	0.749	Yes
Years of Business Education	1.000	1.000	1.000	0.113	Yes

. The permutation p-values are more than 0.05, and the original correlation is greater than 5.0% quantile for all constructs; hence, the compositional invariance is established [120]. Further, in step three, invariance of mean and variance both are not established for some constructs (

Table 9. Measurement invariance test: step three invariance of mean and variances.

Constructs	Mean-Original Difference	95% Confidence Interval		Invariance of Mean	Variance-Original Difference	95% Confidence Interval		Invariance of Variance
		2.5%	97.5%			2.5%	97.5%
Entrepreneurial Intention	−0.017	−0.172	0.164	Yes	0.210	−0.255	0.259	Yes
Gender	−0.073	−0.165	0.174	Yes	0.034	−0.079	0.079	Yes
Perceived Behavioral Control	0.688	−0.174	0.162	No	−0.009	−0.199	0.211	Yes
Personal Attitude	−0.240	−0.176	0.166	No	0.213	−0.336	0.339	Yes
Role Model	0.075	−0.168	0.166	Yes	−0.023	−0.049	0.054	Yes
Self-employment Experience	−0.054	−0.173	0.167	Yes	−0.056	−0.180	0.180	Yes
Subjective Norms	−0.422	−0.173	0.168	No	0.166	−0.303	0.312	Yes
Years of Business Education	0.058	−0.170	0.167	Yes	0.792	−0.275	0.273	No

). Hence it can be concluded that a partial invariance is established, sufficient to carry out the multigroup analysis.

Table 10. Multigroup analysis (in India N = 285, in Saudi Arabia N = 250).

H8 on Different Paths	Path (India)	Path (Saudi)	Path Difference	Path Permutation Mean Difference	2.5%	97.5%	Permutation p-Values
Gender→PA	0.122	−0.124	0.246	0.000	−0.164	0.163	0.004
RM→PA	0.037	0.031	0.006	−0.001	−0.168	0.163	0.947
SEE→PA	0.177	0.067	0.110	0.001	−0.143	0.138	0.122
YBE→PA	−0.086	0.042	−0.128	−0.002	−0.172	0.171	0.147
SN→PA	0.393	0.194	0.199	0.002	−0.191	0.194	0.044
Gender→SN	0.129	−0.064	0.193	0.001	−0.181	0.183	0.037
RM→SN	0.094	−0.090	0.184	0.002	−0.179	0.185	0.048
SEE→SN	−0.022	0.027	−0.050	−0.001	−0.177	0.176	0.576
YBE→SN	0.030	0.025	0.006	0.002	−0.163	0.170	0.945
Gender→PBC	0.229	−0.029	0.258	0.001	−0.161	0.167	0.002
RM→PBC	0.001	0.179	−0.178	−0.001	−0.175	0.168	0.041
SEE→PBC	0.135	0.210	−0.075	−0.002	−0.163	0.158	0.372
YBE→PBC	0.087	0.197	−0.110	−0.001	−0.159	0.157	0.183
SN→PBC	0.396	0.116	0.281	−0.001	−0.183	0.178	0.002
PA→EI	0.566	0.500	0.067	0.001	−0.168	0.170	0.440
SN→EI	0.136	−0.131	0.267	0.000	−0.165	0.163	0.002
PBC→EI	0.242	0.318	−0.076	−0.001	−0.157	0.150	0.335

shows the path coefficient of both subsamples, their differences and significance of difference. The paths having p-value greater than 0.05 have significant differences between the subsamples such as gender and personal attitude.

5. Discussion and Conclusions

There were two main objectives of this study—first, to test if the EIM can predict the EI in a sample from two different countries; second, to test if the country variable makes a significant difference to any of the hypothesized paths in the model. Student samples have been studied in numerous researches on entrepreneurial intention [20,48,56].

Overall, the EI of students in both samples show similar intention to start their own venture. In a study on Germany and Poland [109], the EI was found to be different. Socio-economic environment of the countries was hypothesized to be the reason for the variations. In another study [33], comparing students from both the USA and Turkey had a lower EI but higher PA.

Based on the analysis, we found that two antecedents of EI, namely PA and PBC, significantly impact the EI in individual samples of India and Saudi Arabia as well as the combined sample. A positive and significant influence of PA and PBC have been confirmed in several studies [2,14,15,45]. This study also reveals a similar influence of PA on EI (H1). It has been reported in several studies that PA is the strongest of the three predictors of EI [123,124,125]. This predictor has remained significant in varying cultural contexts, a proposition confirmed in other studies [9,126]. A nonsignificant role of PBC has also been found in few earlier studies [127]. Karimi et al. [128] point out that the role of PBC may be affected by the economic status of the country where the research was conducted. In a developed country, the students might feel more confident about their success, whereas in developing countries or underdeveloped countries, they might be less or not confident.

The antecedent SN failed to show any significant effect on EI in the combined sample in this study. Interestingly, the relationship SN–EI was found significant in both the samples individually. An earlier study on several countries including India [107] also found a similar relationship. In general, the effect of SN in predicting EI has generally been ambiguous [105,129,130]. Some studies confirmed a direct relationship of SN with EI [12,29], while others did not [2,21]. The varying results of SN have been thought to be influenced by culture. Article [131] mentions that Western cultures may have less dominant influence on future entrepreneurs compared to non-Western nations. Moriano [107] also supports the view that individualistic and collectivistic cultures will show different results on the entrepreneurial intention, especially the subjective norms. In individualistic countries like the USA, Germany, etc., subjective norms would not be significant, whereas in collectivistic countries like India and Russia, they would be significant. Considering this ambiguity, [14] postulated that SN might influence EI indirectly through PA and PBC. This study confirmed that SN does influence PA and PBC significantly. This is similar to earlier studies [9,132]. Linan and Santos [133] first proposed that this relationship should be included in the entrepreneurship intention model. [131] opine that the SN scale may also be required to be revisited, despite adequate factor loadings.

The entrepreneurial intention model is an adapted version of the TPB that retains the same predictors of EI. The results show that overall, the model explains the measured construct EI reasonably well (R²_combined = 0.522; R²_India = 0.644; R²_Saudi = 0.412). Similar findings have been reported earlier regarding the applicability of EIM in CIE research [20]. One of the distinguishing traits of culture is the individualism and collectivism dimension. Countries differ from each other on this dimension to different degrees. On the scale developed by Hofstede, India is a mix of individualism and collectivism, whereas Saudi Arabia is a highly collectivistic society. The current findings show that EIM explains the EI quite well with the least variance reported in Saudi Arabia. A study on Zambia [134], which is a collectivist society, also confirmed the same results. A comparative study of Poland and South Africa [108] confirmed the validity of TPB in samples from both countries. Furthermore, the country moderator significantly influenced the EI in the two studied countries, PA and PBC. In a study [107], per capita income of a country was found to influence EI. Similarly, for the current study, EI for Saudi Arabia is higher than for India, although marginally because the per capita income of Saudi Arabia is more than India. The Q² values are all above zero, indicating that the model has good predictive value for individual as well as combined samples. PA and PBC were found to significantly influence EI across the samples.

The demographic factors and human capital variables considered in the model show mixed results. The Gender-PA (H4a) path was significant in India and Saudi samples individually, but there is no significant effect in the combined sample. Gender-SN (H4b) and Gender-PBC (H4c) were not found to be significant in any sample. In a study on Saudi students [75], gender was found to be significant. The effect of gender on the antecedents of EI have been found to be contradictory. Diaz-Garcia and Jimenez-Moreno [123] reported that gender was found significant in some studies and was not found significant in others. The article [134] opines that a lower EI for females may be because of experience, support or role models. Integrating these thoughts with the findings of the present study, it would be interesting to investigate a combined effect of gender with experience, social support and role models. It is possible there may be some new findings and implications.

Role model did not have a significant influence on either PA (H5a) or SN (H5b) in combined as well as individual samples. Bosma et al. [63] have suggested earlier that students or budding entrepreneurs do not take “icons” as their role models. Therefore, this variable needs to be defined more accurately in order to obtain a clearer picture of its role on EI. Bosma et al. found that an entrepreneurial role model is someone who the student/potential entrepreneur “knows personally and with whom he/she has frequent interactions”. The role model just shows significant influence on PBC (H5c) in the Saudi sample. In another study in Saudi Arabia [126], the antecedent SN was influenced by role models.

Self-employment experience (SEE) has a significant impact on PA (H6a) and PBC (H6c) in the combined sample and individual samples, but the SEE-PA is not significant in the Saudi sample. Significant effects of SEE have been reported in earlier studies [64]. Years of business education have a significant influence on PBC in combined and Saudi samples only. As discovered in [20,135], university support through education may significantly influence entrepreneurial attitude and perceived behavioral control. Education was found to influence PBC (H6c) in this study.

Cultural influence might be responsible for varying results in the two samples. The results of the multigroup analysis show the role of the country in the various paths between the predicting variables and the EI. Gender–PA, Gender–SN, Gender–PBC, SN–PA, RM–SN, RM–PBC, SN–PBC, and SN–EI relationships are significantly different for the two countries. The role of gender is perceived differently in both countries, and therefore, all the relationships between gender and predictors of EI are found to be significantly different. The same is the case with SN. In earlier studies, too, SN was influenced by country [14,136].

Implications

The results contribute to the scarce literature on CIE research. Moreover, the comparison of India and Saudi Arabia makes the study much more interesting and important because both the countries differ substantially in terms of culture and still face the common economic challenge of unemployment. The validation of EIM in these two culturally different nations adds to the strength of the model. The results are also particularly important to the ministries, educational institutions, government bodies as well as financial institutions. For example, the ministry of education and universities can apply the findings to introduce more targeted entrepreneurship education that bolsters the EI. The findings can also aid educational institutions in improving the “intermediate processes” [137] leading to entrepreneurship. Actions that are relevant at a cross-cultural level can also be derived from the findings of this research—for example, inclusion of entrepreneurship courses in curricula that directly influences the perceived behavioral control [138]. The level of entrepreneurial intention measured in this study would help financial institutions in determining the current propensity for entrepreneurship that might translate into future need for financing.

The findings have multifold implications on the university support for entrepreneurship. We should update curricula to include entrepreneurial lessons in varying degrees, from full courses to short modules in various programs. A method of objective evaluation of the policies may also be developed by understanding the results in cross-cultural context. Each antecedent of EI and the model would reflect the propensity of the students to take entrepreneurship as their career. The effect of various types of interventions in curricula should also be examined to see what would result in the most positive outcomes. These interventions should also include training and interactions with successful entrepreneurs. The faculty exposure to a practical entrepreneurial ecosystem is also important. The faculty exert a strong influence on the students’ career intentions and knowledge. In fact, their role in direct or indirect influence in the EIM should be explored. The results would help in devising a pedagogy that would result in increased interest of students in entrepreneurship. A “dual tutor system” [20] will be a good initiative by universities to positively influence entrepreneurial behavior. More specifically, there should be a strong focus on students who report an “indecisive attitude” [132] toward entrepreneurship.

The results of this study are pertinent to government agencies and ministries responsible for formulating policies in higher education. The policy makers should formulate policies that support entrepreneurship as a career option, as [20] point to the development of a “scientifically rational entrepreneurship course system”. Profiling of the students is also possible by considering the overall model and the various variables. Accordingly, appropriate programs may be developed for students in different cultures and different behavioral traits. Further, there should be a synergy between the government and universities to educate the students about the available support to entrepreneurs in the form of soft skills as well as financial resources.

On a special note, the findings of this study related to gender are also worthy of attention. Adequate measures should be taken to develop a system that encourages positive entrepreneurial intention among females.

In terms of research implications, this study addresses the call for more research in the area of CIE that involves comparison of entrepreneurial behaviors across cultures. Both the countries considered for this research share common features with many other countries, especially in the broad category of “Asian culture”. Therefore, there is a significant potential for generalizability of the results to such countries.

One of the limitations of the study is that the researchers could not apply random sampling, limiting the generalizability of results. Secondly, entrepreneurship can be pursued by other population groups. Because only university students were considered in this study, we cannot generalize the findings to other groups. Third, self-reporting research instruments do suffer from bias, and the results should be corroborated by other methods too. Policymakers need to be cautious in interpreting the results for application because the intention to start entrepreneurship might not always convert into a real act. Further longitudinal studies are required to understand the process of transformation of intention into action. However, even with the limitations, the current study applied all the scientific protocols so that the results are meaningful, the measurement of constructs meets standards, and the overall model is tested considering all parameters.

The entrepreneurial activities are one of the fundamental engines to drive economic growth and well-being of the people, thus supporting the goals of sustainable development. Entrepreneurs are the shining gems that carve out new opportunities for society. Therefore, institutional support is very essential to develop entrepreneurship. This study brings good comparative insights in two countries, namely India and Saudi Arabia, to develop and support the cause of entrepreneurship.