Academic entrepreneurship in Sudanese universities: explaining entrepreneurial intention using the Theory of Planned Behavior (TPB)

This study aimed to fill the literature gap of entrepreneurial intention antecedents from the academic staff’s perspective using the Theory of Planned Behavior (TPB) in an underdeveloped country context. Empirical results were derived from a quantitative approach based on a survey method with a selected academic staff sample of 97 Sudanese public universities. Hypotheses testing was done using the structural equation modeling (SEM) through the partial least squares (PLS) method to test the impact of attitude, subjective norms, and perceived control as the antecedents of entrepreneurial intention. The study results revealed the applicability and consistency of the Theory of Planned Behavior (TPB) in explaining Sudan’s academic entrepreneurial intention. The results indicated that attitude, subjective norms, and perceived control explained 61.70% of the entrepreneurial intention (EI) variance. Perceived control (PC) proved to be the primary antecedent of EI, which explained 42.20% of EI variance, while the business environment and experience were the leading influencers of perceived control. This result can guide the authorities to formulate policies that encourage the transformation of the Sudanese universities to be entrepreneurial.


INTRODUCTION
In the last few decades, entrepreneurship has gained attention universally as an essential instrument to improve economic growth. Numerous studies have stated that entrepreneurship in developing countries, including Africa, is a creative solution to resolve and counter the significant challenges and enhance economic growth, create employment, and reduce poverty (Omoruyi, Olamide, Gomolemo, & Donath, 2017). The policymakers should consider the technological knowledge that the publicly funded research institutions such as universities, laboratories, and research centers produce and transfer as drivers of this development and guarantee economic growth sustainability.
As an underdeveloped country, Sudan, suffering from widespread unfavorable conditions at all levels, entails serious initiatives to overcome its drawbacks. From an economic viewpoint, the country faced high inflation rates and broad segments of unemployment. Despite these crises, there are initiatives oriented towards entrepreneurship to create and overcome unemployment problems. Many studies emerged in this field ( (Mwange, 2018). TPB is becoming central in social psychology studies concerned with explaining behavior (Lucas & Cooper, 2012). According to Ajzen (1985), the researchers extended the Theory of Reasoned Action to the Theory of Planned Behavior, which deals with the type of behavior that people have limited control over attempting to perform or not a given behavior. TPB relies on the intention to determine the intensity of an attempt to perform a given behavior.
TPB proved that three antecedents influence the intention: the attitude toward trying, the subjective norm concerning trying, and the perceived behavioral control of internal and external factors (Ajzen, 2012). Esfandiar, Sharifi-Tehrani, Pratt, and Altinay (2019) indicated that attitude refers to an individual's appraisal of the outcomes of behavior in question and the degree of favorability. Social/ subjective norms refer to the individual's perception toward the other's role in performing a given behavior and to the extent his behavior is consistent with a leading person or group's thoughts. Perceived behavioral control refers to the individual ability to perform a given behavior and the perception of the controllability of the behavior (Liñán & Chen, 2009). In the psychological literature, intentions have proven to be the best predictor of planned behavior. Accordingly, the intention models ideally suited entrepreneurship as it is a type of planned behavior (Krueger, Reilly, & Carsrud, 2000). Consequently, entrepreneurial intention (EI) would be the key to understand the complicated entrepreneurial process and explain entrepreneurial behavior (Miranda, Chamorro-Meraa, & Rubiob, 2017).
F. Guo, Zou, J. Guo, Y. Shi, Bo, and L. Shi (2019) defined academic entrepreneurship as "found-ing of companies by academicians who had previously worked in laboratories or university departments that are technology originated". From this perspective, academic entrepreneurship encompasses activities beyond the traditional roles of the academic or research institutions that generate financial and social benefits for entrepreneurial academicians or the university (Abreu & Grinevich, 2013). Such activities imply innovation and risk (Urban & Chantson, 2019 Sudan's entrepreneurial environment is still at the infantry stage; this weakness can partially be attributed to the government's failure to implement a suitable policy for entrepreneurship development besides its failure to implement macroeconomic policies for maintaining stable economic conditions. This failure manifested in high inflation rates, massive budget deficit, fluctuating exchange rates, and inadequate education and infrastructure (Gangi & Timan, 2013). Musa (2012) indicated that businesswomen have huge problems in accessing financing from the traditional banking system. These problems are mainly due to the collateral requirements, high finance cost, and short lending period. These businesswomen have also experienced similar problems with the government's funding programs, such as the microfinance initiative.
The future of entrepreneurial activities in Sudan depends mainly on the improvement of the political and economic conditions. The government needs to revise and fine-tune its regulatory and institutional measures to facilitate and expedite starting new businesses. As far as the cultural barriers are concerned, training and develop-ment of potential entrepreneurs can change their attitudes and empower them with the necessary skills to successfully manage an investment project (Gangi & Mohammed, 2017). In the same direction, Khattab, Sid Ahmed, and Mohamed (2017) stated that entrepreneurship requires a favorable environment to grow and nourish. Startups require low barriers at the outset. Moreover, to achieve scale, they require a legal and regulatory framework that rewards entrepreneurial initiative, ensures fair competition, and protects private property rights. Thus, a robust policy environment must encourage sustainable entrepreneurship, financial, educational, legal, and other support.
Regarding the triggering factors that motivate entrepreneurship, Mansour and Omer (2020) stated that the respondents' entrepreneurial intention is predictable from the attitude towards becoming entrepreneurs, perceived behavioral control, and subjective norms. Their findings revealed that TPB is a valuable tool in understanding university students' entrepreneurial intention in Sudan. Similarly, Ibrahim et al. (2018) suggested that subjective norms and perceived control significantly predicted entrepreneurial intention. Moreover, Mansour and Omer (2020) stated that TPB was a valuable tool in understanding university students' entrepreneurial intention in Sudan. They emphasized that the students' attitudes towards entrepreneurship and perceived behavioral control positively influenced entrepreneurial intention.
The previous studies concerned about the entrepreneurial intention and its antecedents' in the Sudanese universities were entirely focusing on students and the postgraduates, and neither of them discussed the issue from the academic staff perspective.

AIMS AND HYPOTHESES
The study aims to evaluate the applicability of the intention model in underdeveloped countries, mainly in Sudan as well as to study the entrepreneurial intention and its antecedents from the academic staff viewpoints in the Sudanese universities.

Hypotheses
The theoretical framework stood on TBP; according to Aizan (2012), the TBP model suggests that intentions guide behavior. The paper by Krueger and Carsrud (1993) is probably responsible for making the TPB the "reference" theory in EI research (Liñán & Fayolle, 2015). According to the Theory of Planned Behavior, the intention to perform entrepreneurial behavior is influenced by attitude toward the behavior (EA), subjective norms (SN), and perceived behavioral control (PC) (Aizan, 1991, 2012). The higher attitude, subjective norm, and perceived behavioral control, the stronger the person's intention to perform the behavior in question (Vamvaka, Stoforos, Palaskas, & Botsaris, 2020).
This study replicated the intention-based model, which was elaborated by Miranda et al. (2017) who integrated various entrepreneurial intention predictors into one cohesive framework. The model included the antecedents of entrepreneurial attitude (creativity, self-confidence, perceived utility, and business experience).
Having stated the above, the study hypotheses were formulated as follows: H1: Attitude positively influences academicians' entrepreneurial intention.

DATA AND METHOD
This study's population was the Sudanese teaching staff of agriculture, animal production, and veterinary medicine faculties, which is more promising to lead the country's entrepreneurship ventures. The Sudanese economy's main characteristic is that it depends mainly on the agricultural sector (Gangi & Mohammed, 2017). About 40 percent of Sudan's population is involved in the agricultural sector (The GlobalEconomy.com, 2019), which contributes by 30 to 35 percent of the annual country Gross Domestic Product (GDP) (The World Bank, 2017), as the country is rich in agricultural capabilities and natural resources. Due to time and budget constraints, the sample respondents were the academic staff of the mentioned faculties of Omdurman Islamic, Sudan, El Nilain, and Khartoum universities, located within Khartoum State, the capital of Sudan. The survey faced some constraints in accessing the academic staff as live at Khartoum was still not fully back to normal after the Sudanese Revolution of December 2018. However, the data collection took place between November 2019 and January 2020. Consequently, the sample size was limited to 97.
The instrument used for collecting data for this study was a questionnaire administered in Arabic that went through several rounds of pre-tests and refinement taken by a panel of academics and The data were analyzed using the Partial Least Squares (PLS) approach to Structural Equation Modeling (SEM), appropriate to small samples.

RESULTS
The study implemented the Confirmatory Tetrad Analysis (CTA) via computing all vanishing tetrads and examining the upper and lower Bonferroni-adjusted confident intervals "CI Low adj." and "CI Up adj." The study hypothesized that if any of the intervals do not include zero value, then the construct is formative (i.e., it has a value that ranges between + + or --compared to reflective constructs that range between + and +) (Gudergan, Ringle, Wende, & Will, 2008). The CTA results indicated that subjective norms (SN) was the only formative construct for this study, as presented in Table A1.
The reflective measurement model's evaluation procedures applied to EA, CR, PU, SELF, BE, ET, ENV, PC, and EI and were presented in Table A2.
All the reflective construct loadings exceeded the 0.700 reference values, except 12 items, which were excluded from further analysis (see Table A4 for the final list). These were 5 PC items, 4 CR, two items of ENV, and one of ET.
The Cronbach's alpha, which reflects the reliability of the indicators for each construct, ranged between 0.707 (the minimum acceptable level) and 0.872 as indicative of excellent reliability among the indicators (except CR, which converged to a single construct as its items did not satisfy the reliability and validity conditions with Cronbach's alpha equal to 1). The composite reliability, which tells the internal consistency, exceeded the 0. The formative constructs' measurement model assessment included three testes: convergent validity, collinearity among indicators and significance, and relevance of the outer weights.
The evaluation of formative constructs differs from procedures applied to the reflective models. According to Edwards and Bagozzi (2000), the internal consistency reliability concept is inappropriate to formative constructs as they are assumed to have lower internal correlation compared to the reflective ones.
The study assessed the convergent validity of the formative constructs (EA and SN) through the path coefficient magnitude linking each of the two constructs with a global item obtained by averaging each construct's constituting indicators. This test is termed redundancy analysis as per Chin (1998). High path coefficients between the construct and the global item indicate the existence of convergent validity. The path coefficient of the SN construct and its global items was equal to 0.999. i.e., comfortably above the minimum 0.700 acceptable level for convergent validity indicated by Hair et al. (2016).
The study assessed the collinearity among the formative constructs' indicators through the variance inflation factor (VIF). According to Hair et al. (2016), VIF is the reciprocal of tolerance. It represents the amount of variance of one formative indicator not explained by the other indicators in the same block, i.e., The maximum acceptable VIF level is 5, according to Table  A2, i.e., the SN construct was free from the collinearity problem as the maximum VIF was 1.75.
The last step of the formative model assessment tested the significance and relevance of formative items' outer weights to verify how they contribute to the SN construct via bootstrapping procedure (Hair et al., 2016). Two SN items were dropped as their outer weights proved insignificant.
The structural model assessment included the predictive relevance (Q 2 ) test, which tells how the predicted value of a reflective endogenous construct deviates from their original values (Hair et al., 2016). Hair, Ringle, and Sarstedt (2013) also stated that the path model's predictive relevance for a reflective endogenous construct would be assured if its Q 2 values were larger than zero. Table A3 reveals that all the endogenous constructs considered predictive, as their magni-tude of Q 2 was greater than zero; thus, the model's predictive relevance was satisfactory. The model's predictive power (R 2 ) was acceptable (Table A4) as it contributes to explaining 61.70% of the variance of EI, 64.10% of EA, and 12.70% of the PC. Table A5 and Figure 2 present the results of the evaluation made with our structural model. The arrows indicate causal relationships, the number beside each is the respective standardized coefficient, and in parentheses is the product of the standardized coefficient and the correlation coefficient between the two constructs expressed as a percentage (Dijkstra & Henseler, 2015). The bootstrap resampling procedure determined the t-values and significance of the causal relationships with 500 sub-samples.
Perceived control (PC) was the construct that contributed most (46.20%) towards explaining entrepreneurial intent. The EA and SN were significant and contributed by 27.40% and 26.00% in explaining this intention. In turn, the EA is explained by PU (56.90%), SELF (23.50%), and CR (18.50%), while PC is explained mainly by BE (30.00%) and ENV (18.10%).

DISCUSSION
The intention models offer us a significant opportunity to increase the ability to understand and predict entrepreneurial activity (Krueger et al., 2000). The study endeavored to analyze the determinants of the Sudanese academicians EI by using the intention-based model built based on the Source: Survey results.
Note: * Significant at 95%, ** non-significant. Two reasons might lie behind having business experience not significantly contributed to the prediction of the entrepreneurial attitude. The first reason is that entrepreneurship is a newly emerging phenomenon in Sudan, which is lagging while most business activities are within the consumer-oriented service sector (A/Rahman et al., 2018). The second justification is that entrepreneurship education is inadequate and insufficient across all levels and types as there is low awareness at the individual level and deficiency in governmental policies that encourage entrepreneurship.

Figure 2. Structural model estimation
The business environment (BE) was the highest predictor in explaining the perceived behavioral control variance, with 30%, followed by ENV with 18%. The training effect on PC proved to be insignificant, with just 0.50%. The weak impact of training on PC likely reflects Sudan's inappropriate entrepreneurial education and training conditions. In this regard, Timan and Gangi (2015) stated that just a few Sudanese universities' entrepreneurial courses or programs do not reflect the importance of the issue to the Sudanese economy or the pressing need for it. Offering such courses and launching related programs would raise the Sudanese academicians' skills in the field and raise their intention towards it.
Despite the strong support for this research's entrepreneurial intention model, the findings should be deliberately treated because of the lack of previous work using the Sudanese academicians' sample. The small sample was another limiting factor confined to the uncalm security situations that prevailed recently in Sudan. Nevertheless, the model and findings seem quite robust because they explained a very high percentage of entrepreneurial intention and EA and PC variance. According to the study findings, the study suggests the following implications.
The findings indicated the relative effects of PC on entrepreneurial academic intention (H2), which is necessary for policymakers in higher educa-tion and universities to formulate concrete policies to encourage entrepreneurship in universities. Academic scientists' entrepreneurial intentions come about when individuals with suitable psychological characteristics live in environments favoring such activities (Foo, Knockaert, Chan, & Erikson, 2016). Also, the universities must transform into entrepreneurial institutions supported by adequate policies and atmosphere. Entrepreneurial university community (academics, students) can explore, evaluate, and exploit the knowledge transferable into new ventures that eventually pours into the country's economic development via knowledge production and commercialization (Guerrero & Urbano, 2014), which is considered a vital ingredient in Sudan. Unfortunately, academic research at universities in Sudan is mostly basic and does not generate immediate economic value and commercialization opportunities. Another implication regarding training and education are their essentiality in building the academic staff's abilities and skill in establishing and managing their businesses. Thus, the Sudanese government and universities must adopt suitable policies and regulations that facilitate and support entrepreneurship. The applied model is to be elaborated by integrating demographic factors in future researches.

CONCLUSION
This study was based on the Theory of Planned Behavior (TPB) to explain the influence of perceived control, subjective norms, and attitude on the academic staff's entrepreneurial intention in Sudanese universities. The findings proved the theoretical model's ability to explain the entrepreneurial intention in Sudan as one of the underdeveloped countries. The study results indicated a significant impact of perceived control on entrepreneurial intention. The findings also manifested that both the business environment and experience positively influence the perceived control. This result may be attributed to encouraging business environment resulting from the support provided by microfinance institutions to the small enterprises and the in-depth technical knowledge possessed by the academicians. However, the result indicated that training had no significant influence on perceived control. This can also be attributed to the general lack of capacity building programs in the business management area rendered by the universities to its academic staff. Note: * Collinearity statistics (VIF) is for the formative items, ** t-values are for the formative items, *** significance of outer weights is for the formative items.