The Effect of Self-efficacy on Acquiring Innovation Ideas among Food Vendors

The study aims at assessing the effect of self-efficacy on acquiring innovation ideas among food vendors in Zhenjiang. The study employed a cross sectional design with a sample size of 400 food vendors as the study participants. Quantitative data were collected from the participants through surveys. The data were analysed through the use of AMOS 23.0 software. The findings of the study revealed that food vendors acquired their innovative ideas through learning from parental, mentor and television models. The findings indicated that food vendors acquired innovative ideas about the cuisine industry from parents, television and mentors. The findings revealed that self-efficacy boosted the relationship between parental model and innovation. The findings further revealed that self-efficacy mediated the relationship between television model and innovation. The findings of the study have contributed to deepening the understanding of entrepreneurs learning innovation. The study also enriches literature in the field of social learning and career development. The study therefore recommends that entrepreneurs build on enhancing their self confidence level so as to be innovative. The study also recommends that educational institutions should include programmes, courses, seminars and training which will improve students’ individual self-confidence.


INTRODUCTION
In recent time, China has continued to gain fantastic economic performance. Entrepreneurs and small and medium enterprises (SMEs) have contributed immensely to the economic development of China. As a way of ensuring comprehensive, sustainable and equitable economic gains, entrepreneurs and SMEs need to constantly look for and implementing new ways of doing things.
To continue contributing immensely to the economic development of China, entrepreneurs have to be catalysts of change, creative destructors and innovators in general (Schumpeter, 1934). Stressing on the role of innovation to national economies, Huarng and Ribeiro-Soriano (2014) opined that innovative entrepreneurs make a difference in every economy. According to Baumol (2002) entrepreneurial innovation is considered as one of the main foundation of national competitive advantage. Baumol was of the view that entrepreneurs need to introduce new businesses that broke with reputable development routes and undermined established competencies. For Drucker (1985), innovation is considered as being the exact means by which entrepreneurs exploit opportunities. As being innovative entrepreneur is a must condition for entrepreneurs to remain viable in their business, the dynamic and ever-changing environment and competition are more compelling for SMEs and firms to be innovative in order to remain competitive and sustain their businesses (Cefis and Marsili, 2005;Necadova, 2010).
SMEs and entrepreneurs have continued to face the issues of coming out with new and meaningful ideas that may potentially This Journal is licensed under a Creative Commons Attribution 4.0 International License be developed into new goods or services attractive to some identifiable markets. Entrepreneurs identifying opportunities in their environment after wards, must determine how to successfully make use of the opportunity (Vaghely and Julien, 2010). Their ability to turn these opportunities to successful businesses rely on the entrepreneurial innovativeness. In the view of Joyce and Paquin (2016) innovation is considered as a key driver of productivity and business sustainability and thus can aid in solving social challenges at the very least cost. However, literature shows that individuals turn to innovate differently where some innovate more than others. For example, Carland and Carland (1991) opined that entrepreneurs in the United States of America significantly exhibit advanced levels of preference for innovation than managers.
Though, innovative entrepreneurs have been widely viewed as contributing significantly to making change in national economies of countries, there is little attempt to uncover how these entrepreneurs acquire their innovation tendencies. Entrepreneurs are always engaged in daily activities to advance the course of their businesses and therefore have to learn so as to come out with novel ideas, services and products to continue to be in business. Pavitt (2002) stated that in the current business world of competition in which success depends increasingly upon the ability to produce innovative or improved goods and processes, and that knowledge remains the key important construct for innovation-based value creation.
Engaging in business activities means that one has to engage in learning and as such a theory of entrepreneurship needs to be based on a theory of learning (Minniti and Bygrave, 2001). This means that entrepreneurs whose behaviors are embedded in learning will stand the possibility of succeeding in creating business start-ups and sustaining their businesses than their counterparts who do not have such learning behaviour. Many researchers over the years have conducted myriad studies which have paid attention to the connection between learning and entrepreneurship (Cope, 2005;Corbett, 2005;Holcomb et al., 2009;Sardana and Scott-Kemmis, 2010). However, not much research has focused on the antecedents of entrepreneurs acquiring innovation ideas. Evidences from many research findings showed that learning from the experience of others has long been identified as critical for individual and organisational success (Argote and Ingram, 2000;Bresman, 2013). According to Hitt et al. (2001), learning process is well recognised of been a key driver of entrepreneurs' success. It is also evident that learning style has been proved to have an effect on occupational interests and areas of domain expertise development (Kolb and Kolb, 2005). However, empirical studies exploring learning models for innovation such as close family members or a mentor are still few (Ravasi and Turati, 2005). For instance, Abecassis-Moedas et al. (2016) conducted a study on entrepreneurs learning innovation but did not include the symbolic process of observing vicariously. More importantly, no study has included how self-efficacy mediates on entrepreneurs learning innovation though many scholars have focused on self-efficacy in the entrepreneurship field.
Though, vicarious learning is not a new form of learning for individuals and firms, there is little empirical evidence on understanding entrepreneurs learning innovation by observing mainly as a result of the perceived appropriateness of exploration nature of vicarious learning (Bingham and Davis, 2012). Equally worth noting is the fact that, no literature has attempted to recognise how self-efficacy influences entrepreneurs learning innovation. With these deficits in empirical evidences, the study seeks to look at these questions; how entrepreneurs learn innovation through observation? Does self-efficacy boost food vendors ability to learning innovation?
The study in a way to find answers to the above questions examine food vendors in the service industry in Zhenjiang in Jiangsu Province. Food vendors are located in most parts of the city as one can easily stretch out to any food vendor. The researchers considered the food industry because in this sector entrepreneurs require some level of knowledge and experiences which can easily be observed and due to the competitive nature of the food sector. These food vendors mostly and directly engage in the preparation of cuisines themselves. Also, the entrepreneurs' self-efficacy plays an important role as far as learning innovation is concern and for that matter serve as a very useful context for understanding the mechanisms underlying entrepreneurs learning innovation. The large number of these food vendors make the sector a very competitive context hence creativity and innovation are essential (Svejenova et al., 2007). Furthermore, the food (cuisine) sector is associated with a milieu that apprenticeship and mentoring are key to enhancing the capabilities and processes of these food vendors (James, 2006).

Theory of Vicarious Learning
Vicarious learning was first introduced by Albert Bandura in his social learning theory. According to Bandura (1977) vicarious learning is a process of learning through the observation and imitation of a model and noticing that many individual behaviours are acquired through models in the form of observation. Bandura identified direct observation and replication of behavior through a four-stage which include; process of attention thus identifying a model, retention which means encoding the model's actions, motor reproduction which reflects accurately duplicating the action and motivation which portrays the reinforcing action (Bandura, 1977). He, however, expanded the scope of vicarious learning in his later work to include symbolic processes referring to the reproduction of a model's experience through written or pictorial means such as a written case, summary or televised display of action (Bandura, 1989). Learning by observation goes beyond just looking at and doing same but affords individual observer reproduce successful behaviours whilst avoiding unsuccessful ones.
Vicarious learning in the view of Davis and Luthans (1980) is an important perspective for organisational studies since it recognises that individuals often learn more from informal observation of others than through formal means noting that job descriptions, rules and policies are more likely to be interpreted from watching what others do than following written directions.
Engaging in entrepreneurial activities means continuous learning and therefore individuals in one way can learn by observing (Bandura, 1977) and that organisations similarly learn through observation (Srinivasan et al., 2007;Terlaak and Gong, 2008).
Entrepreneurs innovative ideas can be learnt through observation hence the application of entrepreneurs learning vicariously in this study. The study in the next paragraphs will discuss the three ways of learning innovation including self-efficacy.

Observational Learning of Innovation through Models
The study looks at how entrepreneurs in the service industry especially those in the cuisine industry learn innovation. Food/ cuisine is one of the sectors in the service industry which requires a lot of innovation by entrepreneurs to continue to be relevant and sustain their businesses due to the high level of competition. The study therefore considers the following models of learning innovation by entrepreneurs.

Vicarious Learning from Parental or Close Family Model
The household is the first and the most important factor affecting individual personality. Apart from parents influence on children from heredity, family social statute, economy, thought and belief, custom, will, education level of parents, age and family form and members behavior pattern to each other, and total atmosphere of family have influence on the individual personality hence personal and social behavior are equally affective. McCartan and Hargie (1990), argued that parents are indispensable factor/models on social skill forming.
In a study aimed at developing framework for entrepreneurial learning, Rae (2005) revealed a triadic model based on personal and social emergence, learning based on context, and the negotiated initiative. Rae stressed that entrepreneurial identity is formed based on personal and social emergence at the early life and family exposure. Rae findings revealed a positive relationship between entrepreneurial stories which are linked with personal relationships with family members (parents, spouses, etc.) and developing entrepreneurial traits. Abecassis-Moedas et al. (2016) in their study on how chefs learn innovation through observational means revealed that parental models positively associated with innovation. In a similar study, White et al. (2007) affirmed parental that entrepreneurship is explained by how individuals were brought up. The researchers explained that parents act as role models upon which children imitate, acquire basic values or utilities from their parents. Owing to this fact, the researcher advances the argument that since innovation is a behaviour, food vendors learned such innovations through the observation of their parents as a model.
Hypothesis 1: Vicarious learning through parental model positively and statistically relates with innovation.

Observational Learning from Television (TV)
Learning about new things can be traced to different hierarchical sources. Effective modeling helps to build individual capabilities and also improves the sense of individual efficacy required to change the knowledge and skills into useful courses of action (Bandura, 1977). When people watch TV by observing what models do on the television will end up learning the behavior of the models and will intend to utilise such experiences when similar situations arise. In a longitudinal study among adolescents, Martino et al. (2005) found that there is a positive correlation between adolescents who engaged in viewing more TV shows and discussions pertaining to the adoption of safe sex including negative repercussions of engaging in sexual intercourse and the desire to start sex. In this regard, the study argues that when food vendors at their early stages of life engage in watching models on the television with cuisines preparations content such individuals will vicariously learn the behaviours from the personalities.
Some researchers argued that the vicarious cultivation of beliefs have arisen over findings from correlational studies using global indices based on amount of television viewing (Gerbner et al., 1981). The influence of television is embedded in the contents of programs or shows people watch rather than just the mere amount of time spent viewing television. The researcher here again advances the argument that if food vendors continue to watch contents from professionals on how to prepare food, they will learn observationally of what the models are doing especially the positive aspects. According to Bandura (2009), television (the media in broader terms) helps to guide individuals and provide natural incentives and social supports for desired changes.
From the above review of literature, the study proposed the following hypothesis.
Hypothesis 2: Entrepreneurs observing models on TV directly and positively influences their innovative ideas.

Vicarious Learning from Apprenticeship (Mentor) Models
Mentoring refers to the process whereby two or more individuals thus the mentor and mentee work together to develop the career and abilities of the mentee (Goodyear, 2006;Mijares et al., 2013). Okurame (2012) offered a clearer definition of mentoring as a method by which novice practitioners are taught to adapt and succeed in new professional roles. Mentoring individuals on their careers at the beginning stage are crucially important for the mentees since it help them to do away with unnecessary dangers whilst acquiring useful skills and knowledge (Clutterbuck and Ragins, 2002). Through vicarious experiences, learners measure their self-efficacy in future performance based on observed successes or failures of mentors (Schunk and Usher, 2012;Zulkosky, 2009). Some scholars have established that mentors can help their apprentices to progress steadily in their professions, increase their incomes, and achieve greater recognition (Whitely et al., 1991). Greenwood (2017) in an exploration study of lived experiences of six novice nursing instructors found that observing mentors teaching students helped gain a greater understanding of what clinical teaching might look like and how they could incorporate what they observed into their own teaching.
Most of these studies were all conducted in organisational and educational contexts, though the influence of mentoring on the protégés' careers is expected to be similar for entrepreneurs. Entrepreneurial study examining vicarious learning among chefs in the cuisine sector in the U.S found that mentors positively influences innovation (Abecassis-Moedas et al., 2016). Similarly, Ozgen and Baron (2007) in an investigation to determine how mentors help entrepreneurs to recognise entrepreneurial opportunities revealed a positive link between mentors and innovation. The results showed that mentors facilitate opportunity recognition by providing valuable information based on their extensive experience. Also, related to the food services industry, James (2006) revealed that apprenticeship is a recognised learning method. With reference to the above review, we argue that mentees learn from mentors not only through the information that they transmit but also through the models they represent.
Hypothesis 3: Observation of mentor models and innovation are positively related.

Self-efficacy and Learning Innovation
According to Bandura (1977) though an individual acquire knowledge by observing desired behaviors from others, selfefficacy is instrumentally key for those behaviors to be enacted by the individual. Self-efficacy is an individuals' judgments on whether he/she can perform a behavior at different times and levels. Importantly, the desire that drives people, emotional states, and their activities are basically driven by their beliefs than what is objectively true (Bandura, 2006). It has been argued that selfefficacy beliefs better predict people accomplishments than their previous achievements, knowledge, or skills and for that matter such beliefs are linked with the level of energy an individual put towards his/her goals, perseverance, and how flexible one is in the face of difficulty. Self-efficacy beliefs are influenced by contextual factors which include directing and regulating one's motivation, thinking level, emotional conditions, actions, or environmental conditions. There exists some link between these beliefs and the expected outcomes. More precisely, when individuals believe in their capabilities, they tend to achieve the expected outcomes thus people who are confident anticipate positive results whilst people with no confidence will anticipate negative outcomes.
Self-efficacy has been commonly identified as having influence on performance as many research pieces have focused on entrepreneurial self-efficacy. Some scholars have established that entrepreneurial self-efficacy has the ability to positively influence performance (Bandura and Locke, 2003). Therefore, food vendors ability to vicariously learn innovation depends on their individual self-confidence (self-efficacy). Similarly, many research findings have established a positive relationship between the level of entrepreneurial self-efficacy and firm performance (Baum and Locke, 2004;Hmieleski and Baron, 2008;Kickul et al., 2009).
According to Zhao et al. (2005) entrepreneurs believe in their competencies positively influence the desire to established a new venture. This means that food vendors individual competences will determine their level of learning innovation from the three models. Some scholars observed that entrepreneurial selfefficacy is associated with certain behaviors such as opportunity recognition and innovation by entrepreneurs (Chen et al., 1998). Based on the prowess of self-efficacy, Lucas and Cooper (2005) argued that self-efficacy is one of the key psychological constructs linked to commitment to accomplish goals. It is therefore, argued here that self-efficacy wields positive influence on the ability of entrepreneurs to learn innovation from the three models. Based on this, the researchers propose these hypotheses.
Hypothesis 4: Self-efficacy positively mediates the link between parental model and innovation.
Hypothesis 5: The relationship between television model and innovation is boosted by self-efficacy.
Hypothesis 6: Self-efficacy positively influences the relationship between mentor model and innovation.

Conceptual Framework of the Study
The researchers developed the conceptual frame based on the review of literature on the constructs under consideration. The framework seeks to explore the link among learning models, selfefficacy and innovation (Figure 1).

MATERIAL AND METHODS
The study seeks to assess how food vendors in Zhenjiang vicariously learn innovation and the mediation effect of individuals believe on their capabilities on learning innovation. The study specifically looks at how these entrepreneurs in the food industry observationally learn innovation from models during their early part of life and the effect of self-efficacy on learning innovation.
Quantitative method was used in the data collection and analysis process. As a way of measuring the constructs, testing and verifying our hypotheses, the researchers administered a survey to food vendors. The researchers administered 460 surveys and 400 were successfully answered and retrieved from the respondents. This represented 87 % of valid administered surveys. The participants are owners of food joints (food vends). These food vendors have operated for more than a year and have to adopt strategies, innovative ideas and paths, in order to be competitive enough to sustain and increase their revenues. Due to the competitive nature of the cuisine (food) sector, entrepreneurs (food vendors) are faced with dynamic and rapid changes in innovative ideas and customers Figure 1: Conceptual framework of the study preferences which implies the entrepreneurs have to take key strategic decisions of the firm to have competitive advantage over others (Kreiser et al., 2002).

Structural equation model (SEM) was employed in analysing
the data through the application of AMOS 23.0 software. The structural equation model is appropriate in this study since it takes a confirmatory approach to analysing data by stating specific relationships among variables (Teo, 2013). Using SEM also enables the researchers to assess the factorial validity of the questions which make up the scales by revealing the extent to which it is likely to measure identical concepts or variables (Hardy and Bryman, 2009).
The questionnaire was translated into Chinese by a Chinese professor and was crossed checked by another Chinese professor to ensure that the questions were properly translated. The study further conducted a pre-testing of the questionnaire among ten food vendors as a way of improving the accuracy of the questions. The results were translated into English and some modifications were made to the questions and the wording. The final questionnaire was then design for the study.

Measures
All the constructs in the study were measured using a five-Likert scale. Though, in previous studies parental and mentor models were measured in binary coding, this study takes different perspective of measuring these models. The researchers believe that measuring these models by simple binary coding will not bring out deep processes that might arise during learning. With this argument as the underlying point, the researchers measured these models using a five-point Likert scale where the respondents were to indicate whether they agreed or disagreed from a range of 1 (strongly disagree) to 5 (strongly agree). Parental model (PARE) was measured by four indicators such as "My parents' profession is closely associated with cuisines." I had always learned from my parents as I observed them intentionally or unintentionally; my passion was to continue my career in line with that of my parents; there are many people in my close family who are into cuisine preparation. Mentor model (MENT) was measured by four indicators. For instance, I have worked as an apprentice before starting my own business; I have acquired innovative ideas from my mentor. Television model (TEL) was measured by four indicators such as I have watched cooking/cuisine related programmes before starting my own business; I have learnt of mixing ingredients from watching television; watching cuisines programs helped shaped by creative ideas on cuisines preparations.
Self-efficacy (SE) was measured based on Zhao et al. (2005) measurement of self-efficacy. Four indicators were used. The respondents were asked to indicate how confident they were at these four items: Creating new products; identifying new business opportunities; thinking creatively; commercialising new idea or product. A five-Likert scale ranging from 1 (no confidence) to 5 (complete confidence) was used to measure the indicators.
To measure innovation (INNO), this study differs from the many usual studies which measure innovation either based on R&D or intellectual property right hence limiting innovation to larger firms and invention. The study measured innovation at the entrepreneur level. The need to measure innovation based on multiple -dimension is important than a just a single indicator (Hagedoorn and Cloodt, 2003). According to Petruzzelli and Savino (2014), measuring innovation on the basis of multi-dimension is even more key in the cuisine industry where innovation can take different forms. Innovation was measured by four indicators. The study based on novelty, new product, new process, new service, and rebranding as indicators to measure innovation among food vendors. The researchers used a five-point Likert scale to measure the indicators.

Reliability and Validity Assessment
It is always important to conduct validity and reliability checks to ensure that the data collected adequately reflect the intended constructs measurement in the study. The need for reliability and validity test are very important in cases where psychological scales are used in measuring human behavior (Thompson, 2003). Hair et al. (2013) pointed out that it is important to verify that measurement scale accurately represents the concepts to be measured.
Internal reliability was achieved as the Cronbach's Alphas of the constructs which were calculated from SPSS exceeded 0.70. Hair et al. (2013) underscored the importance of composite reliability and the values of CR were above 0.6 revealing that composity reliability was met. The AVE were all above 0.5 acceptable threshhold showing that convergent validity was achieved. Table 1 shows reliabilty and convergent validity values.
Researchers often consider vadility issues very important and as such the study undertook a validity measure of the data. The data was fit since all the fitness indxes were achieved. This means that all the construct were valid. Table 2 represents the fitness indexes. The study also revealed that all the modification indices (MI) were below 15 which is the accepted MI value. Also, discriminant validity was achieved since all the square roots of AVE values are higher than the values of correlation between the constructs. Table 3 represents the discriminant validity values.

DATA ANALYSIS AND RESULTS
The study used AMOS 23.0 in analysing the data. The use of SEM enables the researchers to assess the factorial validity and established the relationships among the various constructs. The researchers first conducted a confirmatory factor analysis. The results showed that all the indicators had factor loadings above 0.60. Figure 2 represents the factor loadings and initial path analysis whilst Table 4 represents the regression results.
The study results revealed that parental model positively and statistically influence innovation learning (β = 0.176, P = 0.001).
As was expected, this result confirmed hypothesis 1. Also, the results indicated that television model had a positive and significant relationship with innovation (β = 0.324, P = 0.001). The findings therefore supported hypothesis 2. The study further revealed that mentor model positively and significantly relates with innovation. As the researchers expected, hypothesis 3 was also confirmed. Table 5 represents the direct effects of the three models on innovation learning.
In order to test the mediation effect of self-efficacy on the models of learning innovation and to determine the significance level, Preacher and Hayes (2008) bootstrap procedure of 2000 sub sample was used. The results showed that self-efficacy enhances the relationship between television model and innovation. Television model revealed a coefficient value of 0.07 on innovation through self-efficacy at 95% confidence interval with lower limit confidence interval (LLCI) of 0.086 to an upper limit confidence interval (ULCI) of 0.395. The indirect effect of television model is statistically significant at 0.05 since the confidence level here does not include zero. As the researchers envisaged, hypothesis 5 was supported hence self-efficacy played a mediation role in the relationship between television model and learning innovation. In   addition, it emerged from the analysis that self-efficacy influences the relationship between parental model and innovation such that parental model indirectly influences innovation through self-efficacy by 0.009 at a 95 percent confidence interval. This indirect effect was statistically significant since it had a lower limit confidence interval (LLCI) of 0.130 to an upper limit confidence interval (ULCI) of 0.459. The indirect effect of parental model is therefore statistically significant at 0.05 since the confidence level here does not include zero. Hypothesis 4 was confirmed. However, contrary to the researchers' anticipation, the results indicated no mediation effect on mentor model. Therefore, hypothesis 6 was rejected. Tables 6 and 7 show the mediation effects and mediation analysis respectively.

Discussion
The study sorted to understand how food vendors learn innovation at their early lives which they utilize later as entrepreneurs in their businesses. The study also assessed how self-efficacy influences food vendors learning innovation from the three models. Vicarious learning was examined at the individual level first by Bandura (1977) where individuals observationally learn from people. The scope of vicarious learning was further expanded to include the organizational level where by employees informally and observationally learn from people (Terlaak and Gong, 2008). Though, vicarious learning among entrepreneurs occur, Bingham and David, (2012) posited that its analysis is usually limited to the firm level. Innovativeness among entrepreneurs in the food sector is very important since these innovations can take different forms such as in the mixing of ingredients, new services, new product and methods. It is therefore a step in the right direction to investigate how food vendors in Zhenjiang learn their innovations and the consequences of such innovativeness on acquiring innovative ideas. Self-efficacy which has extensively been researched in the field of entrepreneurship but has not be considered in relation to learning innovation was captured in this study to assess its mediation effects on the learning models.
The study results of positive and significant parental model of learning innovation supported Abecassis-Moedas et al. (2016) findings that parental models positively associated with learning innovation. The findings also agreed with White et al. (2007) position that entrepreneurship is explained by how individuals were brought up by their parents. The findings continue to add to the explanation that entrepreneurs learn innovation at their carriers from parents and close relatives. This lays emphasis on the point that food vendors learn how to prepare cuisines from their parents at the early part of their carriers.
Secondly, the findings of the study showed that food vendors learn innovation observationally from their mentors at the later part of their careers. The findings agreed with Abecassis-Moedas et al. (2016) findings that chefs learn innovation from their mentors and that the innovation learnt positively influenced their performance. The study evidence also supported Ozgen and Baron (2007) conclusions that mentors positively influence innovation abilities of their mentees.
In addition, the findings showed a significant relationship between television model and innovation. Bandura (2009) assertion that television (the media in broader terms) helps to guide individuals and provide natural incentives and social supports for desired changes was confirmed in this study. This means that when individuals watch TV programmes on cuisines preparation related content, they will derive innovative ideas from such programmes.
The study revealed that self-efficacy boosted the relationship between television model and learning innovation. This finding    showed that though, entrepreneurs can acquire innovative behaviours from watch cuisines related content on television, their self-confidence in relation to applying what they watched is very important. This study support the evidence that self-efficacy has the ability to positively influence performance (Bandura and Locke, 2003). Finally, the results revealed that self-efficacy mediated the relationship between parental model and innovation.

Implication
The study has contributed theoretically by assessing entrepreneurs acquiring innovation ideas vicariously. The study therefore contributed to enriching the understanding of social learning of carrier development observationally. This research has added to deliberations of how people learn innovation observationally not only in the organizational or firm level but at the individual entrepreneur level. The inclusion of self-efficacy in the study brought a new dimension to innovation study especially learning innovation vicariously. This will open up a new discourse in assessing how self-efficacy of entrepreneurs influence their ability to acquire innovative behaviours.
Furthermore, the study contributes to establishing that parental, television and mentor models play an important role in the innovative behaviour of entrepreneurs especially in the cuisine industry. The findings provide literature in the field of learning innovation.
Finally, the study contributes practically by establishing that selfefficacy influences entrepreneurs learning innovation. This means that educational institutions, policy makers and change agents should develop curriculum, training programmes, seminars, among others which will enhance the development of self-confidence among young people.

CONCLUSION
The study intended to assess how self-efficacy boost food vendors vicarious learning of innovation in Zhenjiang. The findings revealed that food vendors learn innovation vicariously from mentor, television and parental models. Self-efficacy was found to mediated the relationship between parental model and innovation. It was also found that self-efficacy served as a booster to the relationship between television model and innovation.
The study was conducted on the basis of cross sectional and the results did show over time how these entrepreneurs learn innovation. A longitudinal study would have discovered the trend of learning by these food vendors. Despite the use of crosssectional data, the study results represent an important literature in the field of social and innovation learning.
Another limitation of the study is that, the researchers did not include performance of these entrepreneurs. By including performance, the study would have assessed whether these innovations they learn actually have any influence on the performance of their businesses. A future research should include performance of the business in assessing learning models and innovation.