Predicting Students’ Response to Entrepreneurship in Hospitality and Tourism Education: An Application of the Theory of Planned Behavior

ABSTRACT Globally, the study of entrepreneurship is increasingly becoming a prominent component of hospitality and tourism education. However, uncertainties arising from an unpredictable business environment, inadequate governmental support as well as the ramifications of the recent Covid-19 pandemic have heightened the risks of starting business ventures. Yet the role of risk-taking propensity in driving entrepreneurial attitudes and career decisions in the context of developing economies has received limited attention. Using a survey of 547 hospitality and tourism students from five public universities, the structural equation modeling technique of partial least squares was applied in examining students’ response to entrepreneurship. While entrepreneurial attitudes are driven by risk-taking inclination, the findings suggest that entrepreneurial career decisions are more complicated. The study outcome further underscores the critical roles of personal attitude, normative influence, and behavioral control in shaping entrepreneurial mind-sets.


Introduction
Entrepreneurship has been highly recognized for driving the growth of industries (e.g., Ahmad, 2015;Cunha et al., 2020;Zhang et al., 2020) including those in the hospitality and tourism sector (Esfandiar et al., 2019). As a complex phenomenon that requires the participation of many suppliers from various industries, the hospitality and tourism sector is typically populated by sole proprietorships ranging from street food vendors to small-sized hotels (Andringa et al., 2016). These small and medium scale businesses have shaped the economies of numerous towns and cities. In response, many tertiary institutions have in recent years incorporated entrepreneurial studies into the curricula of academic programmes, including those of hospitality and tourism-related degrees (Ahmad, 2015;Boubker et al., 2021;Mat et al., 2015). These programmes are designed to inspire and equip students for start-up ventures after graduation. Though this approach has earned encouraging reviews in the literature (Bogatyreva et al., 2019) and has received significant research attention across disciplines (e.g., Boubker et al., 2021;Camelo-Ordaz et al., 2016;Nowiński & Haddoud, 2019), there is limited understanding of how individual factors shape entrepreneurial career decisions.
Anecdotal information from developing countries suggests that education is seen as an expensive proposition with the principal aim of enabling the fortunate beneficiary to escape from poverty and redeem their family members. There is the felt exigent need for fresh graduates to find sources of income right after completing higher education. The high cost of accessing capital and the unpredictability embedded in creating and growing a start-up venture in an unstable economy may scare away potential entrepreneurs (Bird & Wennberg, 2016;Edelman et al., 2016). Even in a developed country like the US, Andringa et al. (2016) observed that approximately 60% of small businesses close down before their fourth year with 25.2% collapsing in the first year alone. Still, the role of risk-taking inclination in driving entrepreneurial attitudes and decision making in developing economies has been given limited attention. This study, therefore, adds to the literature in this field by exploring the influence of the predisposition of hospitality and tourism students to take risks on their attitudinal response and formation of entrepreneurial intention. Through the lenses of the Theory of Planned Behavior (TPB), the study further explores the roles of personal factors such as behavioral control and normative influence.

Literature Review
Several scholars recognize entrepreneurship as a job creation tool for economic growth (e.g., Ahmad, 2015;ElSaid & Fuentes Fuentes, 2019;Zhang et al., 2020). Baron and Shane (2008, p. 8) termed entrepreneurs as "engines of economic growth." Entrepreneurs in hospitality and tourism industries are considered instrumental in the sustainable development of their destinations (Cunha et al., 2020). The composite nature of hospitality and tourism industries require several players in meeting the needs of customers (Chou et al., 2020). Koh and Hatten (2002) observed that jurisdictions with the needed resources for tourism will not advance without entrepreneurs. In developing countries where there are often high rates of unemployment, entrepreneurs are needed to generate tourism businesses, create jobs and enhance visitor experiences (Esfandiar et al., 2019).
In an era of fierce competition, entrepreneurship is crucial to the development of tourism and hospitality, particularly at the embryonic stage (Chang, 2011) of destinations, as captured in Butler's (1980) tourism area life cycle (TALC). During the exploration and involvement stages where there are no or few basic facilities for visitors, the local people try to provide some products to meet the needs of guests. Owing to the limited numbers of people visiting the destination at these stages, multinational companies are less likely to set up their businesses in such locations. Hence the importance of small tourism trades and entrepreneurs in delivering the needed services, providing memorable experiences and in forging the economy in such destinations cannot be overemphasized (Cunha et al., 2020).
Entrepreneurship has also become a very prominent subject in secondary and tertiary education systems. The goal is to train students to gain fundamental knowledge in entrepreneurship and to guide them in career decision making (Boubker et al., 2021;Mat et al., 2015). This has become more crucial due to the rising number of graduates leading to increased competition in the job market (Othman et al., 2012). Binti Othman and Othman (2017) aver that university students' views of entrepreneurship can improve when entrepreneurship education is carried out effectively.

Theory of Planned Behavior (TPB)
The TPB is a modification of the Theory of Reasoned Action (TRA). This was triggered by the TRA's limitation of not capturing perceived behavioral control. The TPB suggests that behavior is determined by intention to act (Ajzen, 1991). In other words, intentions are the activators of behavior; thus, they reflect how much effort people are willing to exert to perform a given behavior. Ajzen further proposed that intention is influenced by attitude, subjective norm as well as behavioral control. This study employs the TPB because of its relevance to the issues in the study as well as its proven applicability to various contexts in predicting personal behavior.

Students Attitude Toward Entrepreneurship
According to Ajzen and Fishbein (1980), attitude denotes an individual's favorable or unfavorable assessment of a behavior; thus, the person's evaluation of desirability or otherwise of the outcome of a behavior. In the context of this study, attitude alludes to students' favorable or unfavorable disposition toward entrepreneurship. Agarwal et al. (2020) posit that younger people have a higher tendency to be entrepreneurs. Mat et al.'s (2015) study of engineering technology students and, Nowiński and Haddoud's (2019) study of Polish university students further underscore these favorable attitudes among young people. These tendencies need to be developed through education.
The association between attitude and intention has been considered in several settings (e.g., Boubker et al., 2021;Esfandiar et al., 2019;Karimi et al., 2013;Roy et al., 2017). A study in Iran by Esfandiar et al. (2019), for example, found support for the influence of attitude on entrepreneurial intention. Similarly, Roy et al. (2017) found Indian students' attitude to have the most notable relationship with intention to start one's own business.
Based on the TPB, we argue that the more favorable students' attitudes are toward entrepreneurship, the higher the tendency to want to start new ventures, hence the hypothesis: H1: Attitude directly influences the intention to become an entrepreneur

The Influence of Referent Group on Career Decision Making
Referent groups play a prominent role in daily decision-making. Subjective norms symbolize an individual's view of social reference groups (people regarded as important), whether they support the idea to perform the behavior in question or not (Benchrifa et al., 2017). ElSaid and Fuentes Fuentes (2019) highlight the critical role of family, society, and university in entrepreneurial behavior. Theories on attitudes toward entrepreneurship underline the important role of the environment in inspiring people to venture into new and creative prospects (Shane (2000). Boubker et al. (2021) argued that students may be reluctant to pursue their entrepreneurial goals unless they are pushed by their environment. Some studies have recognized subjective norm as an important factor in shaping entrepreneurial decisions (Mat et al. (2015). However, the intensity of the support and/or pressure from the environment often varies from one jurisdiction to the other. For example, ElSaid and Fuentes Fuentes (2019) found differences in the level of support for Spanish and Egyptian students. Similarly, Elfving et al. (2009) observed that individuals are encouraged more in the United States to venture into entrepreneurship as compared to Europe. In contrast to the positive relationships noted in most studies on entrepreneurship, Shook and Bratianu (2010) detected a negative relationship between subjective norms and entrepreneurial intention. This notwithstanding, we propose: H2: Subjective norms positively influence the intention to become an entrepreneur

The Place of Behavioral Control in Career Decision Making
Perceived behavioral control also known as self-efficacy is defined by Ajzen (1991, p. 188) as the "perceived ease or difficulty of performing the behaviour." It relates to the individual's confidence or ability to handle and manage a situation (Benchrifa et al., 2017). The higher the perceived control over the behavior, the more willing the individual is to act. Ajzen (1991) posits that perceived behavioral control can directly influence behavior. Laguía et al. (2019) observed that students prefer to be entrepreneurs when their educational institutions emphasize the role of creativity in setting up a business. This could instill confidence in students to make informed decisions regarding their career choices. In the corporate world, Shepherd and Krueger (2002) argue that those with high efficacy are more willing to undertake entrepreneurial actions in their businesses. The effect of perceived behavioral control on entrepreneurial decisions among those already working has also been recognized in the literature (Biraglia & Kadile, 2017).
There is some evidence to suggest a positive relationship between behavioral control and entrepreneurial intention. For instance, Esfandiar et al. (2019) identified self-efficacy as the second predictor of intention to start one's own business. In a recent study, Nowiński and Haddoud (2019) reported that Polish students with high entrepreneurial self-efficacy tend to have greater motivation to engage in entrepreneurial activities. Another study by Karimi et al. (2013) found entrepreneurial intention to be influenced by perceived behavioral control. We therefore hypothesize that H3: Perceived behavioral control positively affects the intention to become an entrepreneur.

Willingness to Take Risks
Based on the assumption that the propensity to take risk is not a trait for all, the cultural theory developed by Douglas (1978) explains how people's thoughts and acceptability of risks are influenced by social and cultural traits. Risk-taking is a crucial attribute of entrepreneurship and most entrepreneurs have a strong desire to face unforeseen occurrences (Choi et al., 2019). Zhang et al. (2020) argue that it takes individuals with a highrisk propensity to identify and seize entrepreneurial opportunities. Entrepreneurship requires taking uncertain decisions which may or may not be pleasant. As Hisrich et al. (2017) observed, entrepreneurs are people who could use raw resources to design new things and are willing to take the risks associated with the action. Entrepreneurs take initiative to change the current state of affairs and develop new ways to tasks (Kerr et al., 2019). Empirically, few studies have indicated that risktaking is an important element of entrepreneurial orientation (Chou et al., 2020;Lurtz & Kreutzer, 2017;Zhang et al., 2020). A study in Brazil offered evidence to suggest that readiness to undertake risk could influence entrepreneurial behavior (Ferreira et al., 2017). Hence, we can assume that hospitality and tourism students who tend to take risks are more likely to have favorable disposition toward entrepreneurship: H4: Willingness to take risks directly influences attitude toward becoming an entrepreneur H5: Willingness to take risks directly influences the intention to become an entrepreneur

Measures
A multi-item questionnaire was employed in the collection of data. The instrument was in three main parts. The first part questioned respondents on their sociodemographics, the second was on the traditional TPB variables, and the third covered their willingness to take risk. All variables were measured on a seven-point Likert scale except for attitude. A semantic differential scale was used to collect data on the attitude of respondents toward entrepreneurship. The items for the measurement of attitudes, aubjective-norms, perceived behavioral control, and intentions were adapted from the original TPB scales established by Ajzen (2006) whereas that for willingness to take risk was adapted from Yurtkoru et al. (2014). Subjective norm was treated as a second-order construct with normative beliefs and motivation to comply as its dimensions.

Sample and Procedure
The study adopted a quantitative approach to research which conforms to a post-positivist philosophy. A census was taken of all final year hospitality and tourism students from five public universities in Ghana, comprising two technical universities and three traditional universities. As aforementioned, hospitality and tourism businesses in the country are predominantly owned by small and medium-scale entrepreneurs. Thus, the choice of hospitality and tourism students for the study. The selection of final year students was based on the rationale that, at that stage, all the students in the selected universities would have benefited from an entrepreneurship course and may be better informed about their future career choices.
Out of the 569 students that made up the target population, questionnaires were distributed to 547 of them. The undistributed questionnaires were due to the unavailability of the students during the period of data collection that spanned over two weeks.

Data Analysis
The IBM SPSS Statistics (version 21) was first used to process the data. The data collected were screened for incomplete responses, outliers, etc. In total, 519 valid responses remained for the final analysis. In determining the psychometric properties of the measures and evaluating the structural model, a variance-based structural equation modeling (SEM) technique was applied; specifically, Partial Least Squares SEM using SmartPLS version 3. Given its appropriacy for exploratory and prediction-oriented research (Hair et al., 2011), this approach was deemed most suitable for this study. The PLS algorithm is popular for addressing problematic model identification issues as well as for placing less demand on residual distributions, sample sizes and measurement scales (Hair et al., 2011;Henseler et al., 2009).
The evaluation of normality showed a fairly normal distribution for all measurement items. The values of kurtosis were satisfactorily within the recommended thresholds (Hair et al., 2014). However, four out of the 26 items were negatively skewed at levels slightly below −1. Given the fairly large sample size which minimizes the effect of non-normality on results (Hair et al., 2014), the robustness of PLS-SEM when it comes to nonnormal data (Hair et al., 2011) and the argument around transmutation of data (Pallant, 2013), we considered it most appropriate to keep these variables at this stage without transformation for further analysis.

Measurement Validation
We assessed the psychometric properties of the scale as regards collinearity, items loadings, discriminant validity and internal consistency. The variance inflation factor (VIF) and tolerance values of each of the variables in the model were assessed to check the multicollinearity among the constructs. A VIF of more than 10 and a corresponding tolerance value of less than 0.1 would suggest a very high degree of multicollinearity (Hair et al., 2014). None of the variables were close to these values except one indicator each for intention (intention_d) and perceived behavioral control (control_d) which failed to meet the more stringent criteria of VIF < 5 (Hair et al., 2011). Since the remaining items were adequate for measuring intention and perceived behavioral control respectively, these two items fraught with multicollinearity concerns were consequently dropped from the final analysis.
Next, the reliability and validity of the reflective constructs were assessed. To evaluate the consistency of the constructs based on their measurement indicators, we applied the Composite Reliability Test. Both the composite reliability and Cronbach's alpha values met the 0.7 criterion, confirming the reliability of the latent constructs. Indicator reliability was established by observing the factor loadings (Table 1). All item loadings (outer model) were significant and above the 0.7 threshold. The Average Variance Extracted (AVE) for each construct was greater than 0.5; verifying convergent validity (Fornell & Larcker, 1981;Hair et al., 2014).
Discriminant validity among the constructs was assessed by checking whether: (1) The squared correlations between each pair of constructs were less than the square root of the AVE for each construct; (2) None of the indicators loaded higher on an opposing construct (cross-loadings); and (3) The values of the Heterotrait-Monotrait ratio of correlations (HTMT) were below the threshold of HTMT. 90 (Gold et al., 2001;Henseler et al., 2015). Tables 2 and 3 as well as in Appendix A1, these criteria were all met. The square roots of the AVEs are greater than the corresponding correlations. The loading of each measurement item on its allotted latent construct is larger than its loadings on other constructs. The HTMT criteria were also computed for each pair of reflective constructs based on the item correlations (Henseler et al., 2015). The maximum value is 0.624 in respect of HTMT (MTC, Normative belief) . Comparing the results with the threshold values as defined in HTMT .90 (Gold et al., 2001;Henseler et al., 2015) further confirmed discriminant validity. Table 4 shows the socio-demographic characteristics of respondents. The greater proportion of respondents were females (72.1%), with males constituting 27.9%. This mirrors the general gender ratio of students enrolled in hospitality and tourism programmes both locally and globally. In the United States, for example, Lee et al. (2010) observed that there were more females than males pursuing   hospitality and tourism degrees. Kim et al. (2007) noted a similar trend in Asia. In Ghana, Alhassan and Sakara (2014) attest that there is a higher enrollment of females than males in hospitality programmes. As expected, most of the respondents were in their mid-twenties (46.4%) followed by those in their late teens and early twenties (42.1%). Majority of the respondents considered their family income to be in the middle-income bracket (71.1%). Ghana is ranked by the World Bank as a lower middle-income country. Access to education is free only up to the secondary school level. Students make significant contributions to their tertiary education even in public universities were government covers tuition fees except for a number of fee-paying programmes. It is thus largely middle-income earners who are able to access university education. A significant number of respondents had their mothers (44.7%) and other close relatives in entrepreneurship -predominantly as small-scale entrepreneurs.

Structural Model and Hypotheses Testing
First, we examined the residual root mean square (SRMR; Hu & Bentler, 1998). The SRMR represents the extent of the variance between the matrix of observed correlations and the correlation matrix implied by the model. The SRMR value of 0.08 was within the recommended threshold (Hu & Bentler, 1998), implying the model is a good fit. We applied the Stone-Geisser's Test (Geisser, 1974;Stone, 1974) to assess the predictive validity of exogenous latent variables. To calculate the cross-validated redundancy measure Q 2 , the blindfolding procedure was adopted. Table 5 shows that the Q 2 values of the key endogenous constructs are significantly above zero, indicative of the variables' high predictive power (Q 2 > 0 suggests predictive validity, Q 2 < 0 implies lack of predictive relevance; Chin, 1998). The percentage of explained variance for entrepreneurial intention and attitude were 25.2 and 6.9 respectively.
To test the significance of the hypothesized relationships, t-values were computed using the bootstrapping technique. We applied the non-parametric bootstrapping procedure, using 519 cases, 5000 subsamples and individual sign changes, as recommended by Hair et al. (2011). The results demonstrate that four out of the five hypothesized paths in the structural model show statistically significant figures (see, Figure 1).
In line with the theoretical assumption of TPB, the direct positive relationship between attitudes and entrepreneurial intention was supported (H 1 : β = 0.125, t = 2.611, p < .01). The second order construct -Subjective normholds a positive and significant influence on entrepreneurial intention (H 2 : β = 0.335, t = 6.121, p < .01). The study also finds perceived behavioral control to be significantly related to entrepreneurial intention (H 3 : γ = 0.148, t = 2.887, p < .01).  Figure 1. Structural Path Analysis. Note: These significance levels are determined via bootstrapping analysis (Hair et al., 2011). *Significant at p < .01.
In contrast, the effect of willingness to take risk on entrepreneurial intention was not supported in this context (H 5 : γ = 0.037, t = 0.807, p > .05) even though willingness to take risk favorably impacts attitudes toward entrepreneurship (H 4 : γ = 0.263, t = 5.681, p < .01). A post hoc analysis revealed that the indirect effect of willingness to take risk on intention via attitude was significant (γ = 0.033, t = 2.309, p < .05), implying that attitude mediates the relationship between risk propensity and entrepreneurial intention.

Discussion
Using TPB as a theoretical lens, this study applies partial least squares structural equation modeling technique to examine hospitality and tourism students' risk-taking propensity and attitudinal response in the context of a developing economy as set in an African culture and societal norms perspective.
The results of our study underscore the role of attitude in driving conative response to becoming entrepreneurs. The more favorable attitudes students have toward entrepreneurship, the higher their tendency to want to start their own ventures and vice versa. The findings offer support for the TPB as well as previous empirical research (e.g., Esfandiar et al., 2019;Karimi et al., 2013;Roy et al., 2017) which have established that when individuals feel the outcome from an activity is desirable, they are more likely to be engaged. Developing countries are often confronted with a high incidence of poverty as well as soaring levels of unemployment. This situation usually kindles in students the urgent need to find sources of income while in school or immediately after graduation.
The TPB posits that subjective norms drive intention to act (Ajzen, 1991). In the context of this study, subjective norms reflect the professed support (or lack thereof) from referent groups to start a new business (Laguía et al., 2019). The results imply that when individuals believe that they have support from their social environment, they tend to want to start their own ventures. Some previous studies offer support for this link (e.g., Boubker et al., 2021;ElSaid & Fuentes Fuentes, 2019;Mat et al., 2015;Roy et al., 2017). Individuals are generally motivated to act when they have support from people they regard as important. Referent groups serve as inspiration and dependable sources. The tight-knitted social relationships and support of African societies may play a critical role here. The prevalence of the extended family system enables people to gain financial, emotional, and psychological help from a vast number of people when the need arises. So, if the extended family members urge a person on, there is the likelihood for them to be better cushioned in hard times. Lumpkin et al. (2008) observed that extended families aggregately impact businesses based on the trust, interdependency, loyalty, stability, and traditions shared among the individual members. Research has also highlighted the role of family background in shaping entrepreneurial attitudes (Aldrich & Cliff, 2003). As noted by Farrukh et al. (2017), children who have successful entrepreneurial parents find entrepreneurship valuable and desire to follow their parents' path. Our findings however contrast that of a study by Shook and Bratianu (2010) who reported a negative relationship between subjective norms and entrepreneurial intention in a Romanian context. The Romanian experience was explained by the view that some of those referent groups experienced the communist and socialist rule where people relied solely on the government for employment and might have not embraced the idea of creating new ventures.
Our study further found entrepreneurial intent of students to be influenced by their perceived behavioral control. When individuals believe that they possess the ability to perform an activity that will lead to an expected outcome, they embrace it. This finding corroborates observations in other contexts (Biraglia & Kadile, 2017;Esfandiar et al., 2019;Farashah, 2015;Nowiński & Haddoud, 2019;Shepherd & Krueger, 2002). Students may be moved to make concrete decisions to start their own hospitality/tourism business ventures when they believe that they have the capability and skills to do so.
Willingness to take risks was observed to drive students' attitudes toward entrepreneurship. Individuals with greater propensity to take risks tend to have more a favorable affective response toward entrepreneurial ventures. Contrary to expectations, the direct effect of risk-taking propensity on entrepreneurial intention was not supported. This implies that risktaking propensity may influence how favorably an individual views entrepreneurship but may not necessarily determine the decision to start a new venture. This finding contradicts some studies in other contexts that suggest a direct relationship between risk-taking propensity and intention (Lurtz & Kreutzer, 2017). First, this could be explained by contextual differences. The findings of these previous studies may not necessarily apply to our context. As intimated in the cultural theory (Douglas, 1978), an individual's thoughts of risk are influenced by their cultural and social traits. Boateng (2018) underscores the important role sociocultural factors play in entrepreneurship. Studies in other regions have their unique socio-cultural circumstances that may influence the acceptability of risk and how it influences entrepreneurial decisions. From a corporate perspective, Díez-Esteban et al. (2019) observed that risk-taking is influenced by dimensions of national culture such as power distance, long-term orientation, governance, and the quality of formal institutions. Secondly, while risk-taking propensity drives respondents' attitudes toward entrepreneurship, it appears their entrepreneurial career decisions are driven by additional considerations (possibly, financial and/or environmental constraints) as evident in the significant influence of normative beliefs.
Our study further revealed that the effect of risk propensity on intention is mediated by students' attitudes. Thus, affective response is essential in determining the influence of willingness to take risk on respondents' conative responses to becoming entrepreneurs. Students who have traits of handling risky tasks tend to have favorable disposition toward start-ups, and this subsequently impacts their entrepreneurial decisions.

Conclusion & Implications
Entrepreneurship courses are captured in educational programmes with the aim of developing students to consider setting up their own enterprises as a career option (Benchrifa et al., 2017). Exploring students' entrepreneurial decisions and the determinant factors is pertinent to entrepreneurship research particularly for the hospitality and tourism sector which is largely characterized by small and medium scale entrepreneurial businesses (ElSaid & Fuentes Fuentes, 2019). This is even more important given the issue of unemployment in developing countries. This study therefore examined the attitudinal response of hospitality and tourism students in the context of a developing economy through the lenses of TPB. The model also explored the role of risk-taking propensity to help with our understanding of the formation of entrepreneurial mind-sets. While entrepreneurial attitudes are driven by risk-taking inclination, the findings suggest that entrepreneurial career decisions are more complicated. Nonetheless, the study outcome draws attention to the crucial roles of personal attitude, normative influence, and behavioral control in shaping entrepreneurial mind-sets.

Theoretical Implications
This study adds a fresh perspective to the ongoing discourse in the literature by clarifying hospitality and tourism students' risk-taking propensity and attitudinal response in the context of a developing economy as set in an African culture and societal norms perspective. While entrepreneurial attitudes are driven by risktaking inclination, the findings suggest that entrepreneurial career decision-making is more complicated. A key contribution to theory is unpacking the intricate relationship in this context between risktaking inclination and the formation of entrepreneurial intention via attitude. Findings demonstrate that although risk-taking predisposition drives attitude toward becoming an entrepreneur, it does not necessarily inform the intention to do so. The findings further accentuate the critical roles of personal attitudes, normative influence, and behavioral control in shaping entrepreneurial mind-sets among hospitality and tourism students in the Global South.

Practical Implications
The study's findings have implications for hospitality and tourism education in entrepreneurship. Schools need to focus entrepreneurship training in at least two critical areas: attitudinal change and skills training. Attitudinal training is crucial given the prominent role of attitudes in driving the formation of entrepreneurial decisions. Skills training is essential to improve students' confidence and behavioral control. Giving the significance of normative influence, it would be valuable to have community-wide education to help referent groups to appreciate entrepreneurship as a career path since they constitute a major part of students' career decisionmaking process. As Binti Othman and Othman (2017) observed, successful local entrepreneur stories and guest speakers inspire students' entrepreneurial mind-set. Academic institutions are encouraged to offer mentorship programmes aimed at matching students to entrepreneurs in the hospitality and tourism industry to ease their apprehensions and guide them in their entrepreneurial decision making.
The world has been hit hard by the Covid-19 global pandemic which has led to economic devastation in many countries (Ozili & Arun, 2020). Various startups -both small and medium-scale businesses -have been affected disproportionately. The risk of starting a business venture has become even more profound, entrenching the traditional belief in the Global South that government employment is a safer option. This further highlights the need for including risk management as a core part of any curriculum for entrepreneurship education. Among others, students should be trained in risk management and failure recovery. This could help improve entrepreneurial attitudes and empower students to adopt a rational approach to address concerns.
The turn of events has a spill over effect on jobs. Jain (2020) argues that the aftermath of the Covid-19 pandemic is more likely to result in employment challenges.
The role of entrepreneurship for post Covid-19 economic recovery cannot be overemphasized. Therefore, students should be trained to be more innovative and take advantage of emerging opportunities. Educational institutions could adopt a business generation model in their training programmes to enhance exposure (Bogatyreva et al., 2019). Practical and internship components of educational programmes could be strengthened to expose students to operations in the hospitality and tourism industries to inspire them to take up entrepreneurial opportunities. Policymakers in developing economies should also institute favorable laws and conditions aimed at encouraging graduates into start-up ventures.

Limitations and Future Research
The scope of this study was limited to propensity to take risk and the TPB components (since these have been generally accepted as heuristically appropriate for determining attitudes and intention). For future research, environmental support, self-awareness and availability of alternative careers as well as cultural dimensions can be considered to further enlighten our understanding of students' entrepreneurship decisions. As the Covid-19 pandemic continues to reshape businesses around the world, subsequent studies can assess the impact of the pandemic on students' interests, attitudes and concerns about entrepreneurship in the hospitality and tourism sector.

Disclosure Statement
No potential conflict of interest was reported by the author(s).