“The mediating effect of entrepreneurial self-efficacy on business experience and performance of women-owned enterprises”

This study aims to investigate the mediating effect of entrepreneurial self-efficacy on the relationship between the business experience and performance of women-owned enterprises in South Africa. A quantitative methodology with random sampling was employed. Qualtrics was used to administer the online questionnaire; a sample size of 258 was attained to test the study hypotheses. This cross-sectional study design followed recommendations from scholars on a minimum of 100 as an adequate sample size for regression analysis. Hierarchical regression and mediation analysis were employed to analyze the primary data collected from women entrepreneurs in South Africa. Using ordinal data, a seven-point Likert scale was adopted to operationalize constructs. Out of all the entrepreneurial self-efficacy antecedents and dimensions that were tested, industry experience emerged as the most influential predictor of business performance (β = 0.496*) when mediated by the innovation dimension of entrepreneurial self-efficacy. Although the management dimension of entrepreneurial self-efficacy showed strong predictive power, it did not exhibit significant mediating effects. Consequently, there was only partial mediation of the innovation dimension of entrepreneurial self-efficacy in the relationship between industry experience and business performance. Industry experience and business performance are antecedents and outcomes of entrepreneurial self-efficacy, respectively; they were partially mediated by the innovation dimension of entrepreneurial self-efficacy.


INTRODUCTION
South Africa is one of the most unequal societies in the world.This inequality is reflected in the performance of women-owned enterprises when compared to those owned by men.The country's population is estimated to be approximately 60 million people, with 80% being black Africans and 51% being women (STATSSA, 2022).Several reports, including the World Bank (2021) and STATSSA (2022), have shown that women and the black community (who are the majority in the country) are the ones who must deal with poverty and low representation in business.Entrepreneurship can address inequality; however, the Global Entrepreneurship Monitor (GEM) reports show year after year that women are lagging when it comes to starting and running successful businesses.The gap has been slowly closing, but it started widening again due to the COVID-19 pandemic (GEM, 2022).Msimango-Galawe and Mazonde (2021) attribute this to limited busi-ness experience, low entrepreneurial self-efficacy, and limited exposure to entrepreneurship and gender roles that take up a lot of time from black women.It is, therefore, relevant to give attention to such problems to change the future of women entrepreneurs.
Challenging entrepreneurial environments of many developing countries contribute to poor entrepreneurial self-efficacy and limited business experience.Business experience and entrepreneurial self-efficacy are critical for a business to perform well in different environments.Hence, there is a need to examine the multi-dimensions of entrepreneurial self-efficacy within a developing country context such as South Africa as its level of effect may differ depending on the level and type of experience the woman entrepreneur has.However, women's entrepreneurship activities are increasing despite all the challenges (Catalyst for Growth, 2018; Seed Academy, 2019; Ogundana et al., 2022).Thus, there is a need to understand how to best support women in the forever-changing environment and identify the critical success factors that need to be explored further.

LITERATURE REVIEW AND HYPOTHESES
Entrepreneurial self-efficacy represents an entrepreneur's belief in his/her capability to influence and control circumstances and incidents that shape his/her life (Bandura, 1994).It indicates an entrepreneur's belief in the ability to execute entrepreneurial tasks and the numerous roles of an entrepreneur (Chen et al., 1998).Both definitions reflect the issue of belief, though they differ in emphasis on what that belief is about.Bandura (1999) emphasizes belief in controlling, while Chen et al. (1998) emphasize belief in execution.For this study, both are deemed key for a women entrepreneur's business to perform well.Entrepreneurial self-efficacy can be categorized into four antecedents (industry, work, entrepreneurial experience, and management experience), five dimensions (management, risk-taking, innovation, financial control, and marketing), and one outcome (performance).To ensure that the study focuses on certain variables, it was important to categorize entrepreneurial selfefficacy according to extant literature.Because entrepreneurial self-efficacy is domain-specific to entrepreneurial ventures, it is markedly appropriate for studying venture performance (McGee et al., 2009).This study draws from two theories (social learning and human capital theories) to explain the relationships between business experience, entrepreneurial self-efficacy, and business performance (Bandura, 1982 2008).Therefore, the relationship between entrepreneurial self-efficacy and firm performance can weaken for women entrepreneurs with little or no prior experience.
Pragmatic strategies for the entrepreneurial role, such as performance and skill strategies, are most likely gained from prior entrepreneurial experiences, including past failures (Zhao et al., 2005).Therefore, those with prior entrepreneurial experience would have had more opportunities to observe and gain valuable knowledge from successful role models.The likelihood of women with prior experience pursuing an entrepreneurial career is high because they will be more confident that they possess the requisite know-how to fulfill the roles and achieve the activities necessary to be successful entrepreneurs.
The positive relationship between entrepreneurial self-efficacy and performance is supported by vast empirical evidence (Hallak et  Based on the literature review, business experience and entrepreneurial self-efficacy affect business performance.Still, there are different types of experiences that women entrepreneurs have at various levels, as there are multiple dimensions of entrepreneurial self-efficacy.It is still unclear as to which of these antecedents and dimensions of entrepreneurial self-efficacy need to be prioritized for women enterprises to perform best.This study, therefore, seeks to dive deep into those business experiences (antecedents of entrepreneurial selfefficacy) and dimensions of entrepreneurial selfefficacy rather than generalizing, as most of the previous studies have done.
The study's objective is to investigate the extent to which entrepreneurial self-efficacy dimensions (management, innovation, risk-taking, financial control, and marketing) mediate the relationship between the different types of business experience (management, industry, work, and entrepreneurial) and business performance.It determines the kind of business experience that influences business performance the most.Business experience and business performance are also referred to as entrepreneurial self-efficacy antecedents and entrepreneurial self-efficacy outcomes, respectively.
The study suggests the following hypotheses: H1: Entrepreneurial self-efficacy (management, risk-taking, innovation, financial control, and marketing) mediates the relationship between business experience (management, industry, work, and entrepreneurial) and performance.
H2: Industry-specific business experience influences performance the most when mediated by entrepreneurial self-efficacy dimensions.

METHOD
A quantitative approach was adopted to test hypotheses based on existing theories (human capital and social learning theories).The study took a deductive approach through the positivist worldview.The positivist worldview assumes that research can be conducted objectively without the researcher interfering with the truth, which is the assumption on which this study is based (Creswell & Creswell, 2018).
This cross-sectional paper collected the data online using a self-administered questionnaire via the Qualtrics platform using random sampling (Cooper & Schindler, 2003) to give women entrepreneurs in South Africa an equal chance to respond to the questionnaire.The data collection process included sending a link to the questionnaire to women entrepreneurs from nine South African provinces via the Catalyst for Growth (C4G) database and social media platforms.Catalyst for Growth is an organization that does research in the business development support space and has a database of entrepreneurs and business development support providers from nine South African provinces.C4G assisted with follow-up and en-suring that enough responses were received from potential respondents.
The study extracted variables for three constructs from the dataset: business experience, entrepreneurial self-efficacy, and performance of womenowned enterprises.Entrepreneurial self-efficacy is made up of five dimensions to capture female entrepreneurs' self-efficacy concerning management, risk-taking, innovation, financial control, and marketing.This measure is the most widely used and validated in the extant literature (Chen et al., 1998).
Moreover, demographics and business information were extracted, which included eight variables, as shown in Table 1.The business experience as a predictor variable included four different types of experiences.Business performance as the outcome variable included five indicators.The details of all the variables and how they were measured or operationalized are detailed in Table 1.
The validity and reliability of the measurement scales and constructs were tested using various methods, including exploratory factor analysis and reliability analysis (Cronbach's alpha).Since the scales used were multi-item scales, exploratory factor analysis (EFA) was conducted to test the construct validity, employing principal axis factoring (PAF) to extract factors and Promax for factor rotation.According to previous studies, the factors are likely to be correlated, thus using PAF and Promax.The Kaiser-Meyer-Olkin Measure (KMO) and Bartlett's test were conducted first to test the sampling adequacy of the data to perform EFA.The results showed that the sample was adequate to continue with EFA with a KMO of KMO = 0.947 > 0.5, p < 0.05, which was significant.After rotation, four factors were extracted, and each was labeled as per Table 2. 75.83% of the total variance was explained in the dataset after removing crossloadings and wrong loadings, which is an excellent result as it explains more than 70% with only a 30% error margin.
The proposed study framework included five dimensions of entrepreneurial self-efficacy, but after EFA, only three converged into valid factors.
Similarly, data were collected on five business performance indicators, but only four remained eligible for further analysis.Lastly, there was no change in the number (four) of business experi- Slightly not confident ( 5) Confident (6) Totally confident (7) Note: ESE* = entrepreneurial self-efficacy.Age = Control variable.
ence types as all were eligible.Furthermore, all the measurement scales were reliable, with excellent Cronbach's alpha coefficients higher than 0.9 each.Therefore, the results confirmed that there was construct validity and reliability of the measurement scales.Hierarchical regression was employed to conduct mediation analysis to investigate the mediating effect of entrepreneurial selfefficacy dimensions on different types of business experiences and performance.The motivation to use hierarchical regression was mainly due to its capacity to analyze each variable's unique contribution without confounding effects from other variables in the model.

RESULTS
It is necessary to first present the sample characteristics and profile of the study's respondents to give a clear view in case there are biases as a result of most respondents.
The dataset had 662 cases, and 267 responses were from women entrepreneurs, but after cleaning the data, only 258 were eligible for further analysis.The respondents' profiles included 258 South African women entrepreneurs, of whom 74% were black, 10% colored, and the other 16% were distributed equally amongst other races.The race distribution was primarily black, which reflects the South African demographics in terms of race.It is encouraging to observe that about 97% of the respondents fall under the youth category (18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) age group as the country is driving youth and women entrepreneurship earnestly.The level of education of the women entrepreneurs was very high, with 82% having tertiary education.Most of the businesses are in Gauteng (91%), which is to be expected as the country's economic hub; this leaves the remaining provinces with less than 3% representation each, suggestive that results need to be interpreted and generalized with caution as other provinces are not well represented.Most (68%) businesses had one business owner, and 60% had been operating for more than 3.5 years.It is evident that most had no co-owners and only had one employee, with a limited number having two or more employees.This alludes to the fact that the business performance would reflect the owner's entrepreneurial self-efficacy and help not bias the findings in favor of the entrepreneurial self-efficacy of employees rather than that of the women entrepreneur, which is the unit of analysis in this study.
The study started by summarizing the data using descriptive statistics and determining the relationships between variables using Pearson correlation.Since mediation analysis follows linear regression assumptions, the study first had to confirm the linearity and normality assumptions to avoid violation.Additionally, for mediation to hold, the predictor and mediator variables should first show a significant association with the outcome variables.
Only those that show significant correlations are appropriate for further analysis, as they are likely to be significant predictors and mediators.
Table 3 shows that the dimensions of entrepreneurial self-efficacy have a mean of 6, ranging from 6.14 to 6.22, which is good as these were measured using a seven-point Likert scale.Therefore, the women entrepreneurs in this sample generally have a high entrepreneurial self-efficacy.Consequently, a high business performance mean score of five from a seven-point Likert scale is a reasonably good business performance scale.Several scholars have argued that women entrepreneurs have lower entrepreneurial self-efficacy, especially younger ones.However, these results show evidence contrary to findings from Dempsey and Jennings (2014), Nowiński et al. (2019), and Wilson et al. (2007).This might be partially attributed to the fact that the sample did not constitute only youth but also older entrepreneurs.As much as business experience is low (less than four years), entrepreneurial self-efficacy and business performance still look relatively high, divergent to what the current research suggests that low business experience will lead to low entrepreneurial self-efficacy and, consequently, low business performance.
Looking at the business experience dimensions, management experience, as shown in Table 3, does not significantly correlate with business performance (r = 0.122; p > 0.05) and is excluded from further analysis.This answers whether management experience influences entrepreneurial selfefficacy the most; evidence suggests it does not in this sample due to the weak and insignificant correlation.All the entrepreneurial self-efficacy dimensions significantly correlated with business performance, so the three dimensions (management: r = 0.451; innovation: r = 0.418; marketing: r = 0.379) all significant at p < 0.01 were retained.
Industry experience and entrepreneurial self-efficacy management emerged with the strongest association with business performance among other antecedents and dimensions of entrepreneurial self-efficacy, as presented in Table 3.However, further testing is required to determine whether this remains the case regarding predictive capacity and when a mediator is involved.Hierarchical regression analysis was employed to test the mediating effect of entrepreneurial selfefficacy on the relationship between business experience and performance.The regression model showed an R-square value of 0.387 and significant ANOVA results, which indicated that business experience and entrepreneurial self-efficacy dimensions significantly predict performance.
It is evident from Table 4 that industry experience, followed by work experience, influenced business performance the most.Therefore, it can be tested for mediation.Innovation is the only dimension that is significant and the strongest predictor of performance out of the three tested.
For the mediation analysis, only two types (work and industry) of business experience and one entrepreneurial self-efficacy dimension (innovation) are eligible for mediation.The study concludes that business experience and entrepreneurial selfefficacy positively influence business performance and, therefore, can be further investigated for mediating effects.
For the mediation analysis, the study used Baron and Kenny's (1986) method.The study tested if the variables did not violate any of these assumptions from the four models.When testing the four mod-els, work experience failed the mediation test on model 2 and was therefore excluded from further analysis.In contrast, industry experience did not violate any assumption and was retained for further analysis.Additionally, when the mediator was added to the model, the strength of industry experience was reduced but remained significant; thus, evidence of partial mediation could be confirmed.
The entrepreneurial self-efficacy dimension of innovation mediates the relationship between business experience (industry) and performance.Only one entrepreneurial self-efficacy dimension (innovation) and one type of business experience (industry experience) were tested for mediation.They were the only two variables that did not violate regression and mediation assumptions.The results presented in Figure 1 show a partial mediation of innovation on the relationship between industry experience and business performance.Each of the three direct paths was significant with (c: industry experience β = 0.496; b: innovation β = 0.418 significantly predicting performance) and (a: industry experience β = 0.214 significantly predicting innovation).The indirect path c was calculated from the path a x the path b = 0.089 and the total effect was computed from the direct effect (DE) added to the indirect effect (IE), which is DE + IE = 0.089 + 0.496 = 0.585.
The findings evidence that industry experience is the most important type of experience women entrepreneurs need to acquire to improve their entrepreneurial self-efficacy and, subsequently, business performance.Interestingly, innovation emerged as the only dimension significantly mediating the relationship between industry expe- rience and performance.Although management was also a strong predictor, it did not have significant mediating effects.Thus, contrary to other findings, entrepreneurial and work experience were not the most influential.Instead, industry experience proved significant in improving the performance of women-owned enterprises, especially when mediated by innovation.
Table 5 shows that two types of business experience (management and entrepreneurial) were not tested for mediation as the predictors did not significantly predict business performance.That was a violation of mediation assumptions.Work experience was tested for mediation but was not supported, which leaves only industry experience as the only assumption that was supported and significant.
In summary, among the various combinations of entrepreneurial self-efficacy (ESE) dimensions and types of business experience, only the hypothesis stating that ESE-Innovation acts as a mediator between industry business experience and performance was supported and found to be significant (as shown in row 2 of Table 5).All other combinations were not supported or showed insignificance.Lastly, the hypothesis that industry-specific business experience has the greatest impact on performance when mediated by the entrepreneurial selfefficacy dimension of Innovation was supported and deemed significant.

DISCUSSION
The study hypothesized that entrepreneurial selfefficacy mediates the relationship between business experience and the performance of womenowned enterprises in South Africa.The reason is that entrepreneurial self-efficacy accords women entrepreneurs with the satisfaction of self-dependence necessary for innovation (Hu et al., 2021).More studies posit that entrepreneurial self-efficacy and women entrepreneurs' past work  In this study, it is interesting to note that as much as the sample was highly educated with high levels of entrepreneurial self-efficacy, the business experience was still low at less than four years on average.
Regarding the level of education, in a developing world context, what reins in the gender and entrepreneurship literature are inconclusive studies on how the growth of an enterprise is influenced by experience and the level of education of women entrepreneurs (Ogundana et al., 2022).Therefore, this study contributes to the gender and entrepreneurship literature by shedding light on that debate with the result that a highly educated woman entrepreneur with some level of industry experience enhances the performance of her enterprise.

CONCLUSION
The study investigated the extent to which entrepreneurial self-efficacy dimensions (management, financial control, marketing, risk-taking, and innovation) mediated the relationship between different types of business experience (management, industry, entrepreneurial, and work) and how this affects the business performance of women-owned enterprises in South Africa.
This study argued that as much as most studies make general findings about the importance of business experience and entrepreneurial self-efficacy in running a successful SME, not all types of business experiences and entrepreneurial self-efficacy dimensions are equally significant.Therefore, it was important to establish which business experience and entrepreneurial self-efficacy dimension were most important, specifically for women entrepreneurs in South Africa.
This study concludes that industry experience is more important than work, entrepreneurial, and management experience.Additionally, the innovation dimension of entrepreneurial self-efficacy is the critical dimension compared to risk-taking, management, financial control, and marketing.Therefore, it is a significant mediator between business experience and business performance.
The study's findings have practical implications.Government policies can facilitate a conducive environment for a partnership between women entrepreneurs who need to acquire industry experience and successful businesses within the same industry.Moreover, practical activities that include engaging women entrepreneurs in different business scenarios proven to enhance an entrepreneur's enterprising abilities (Maritz & Brown, 2013) will also be beneficial.These findings also have implications for training providers, as they will need to design programs with learning objectives to improve women's entrepreneurial selfefficacy, especially in innovation, to help boost the performance of their entrepreneurial endeavors.
The fact that most of the respondents were from Gauteng is a limitation.Future research should endeavor to capture geographic distribution that accurately reflects the South African demographics of womenowned SMEs.The study suggests future research to utilize a longitudinal design to explore whether business experience and entrepreneurial self-efficacy affect venture performance differently in the long term.
It was evident that the business performance and entrepreneurial self-efficacy were still high despite low experience.Future research can investigate the reasons for this contradictory finding to establish if there are possible moderating factors.Additionally, it should explore how many years would suffice as business experience to enable performance for each type of business experience.Finally, a qualitative approach could be beneficial in understanding what women entrepreneurs perceive as the most significant type of business experience and entrepreneurial self-efficacy dimension to propel their business success.This significant outcome requires further study of its veracity, factoring in context and other environmental influences to establish other confounding factors that might have contributed to such results.

Figure 1
Figure 1.Mediation results (Miao et al., 2017;Ucbasaran et al., 2008)., 2017)uence venture performance(Bosma et al., 2004;Dimov, 2010;Miao et al., 2017).Entrepreneurial self-efficacy helps an entrepreneur understand the effort to invest.Entrepreneurs with previous entrepreneurial experiences appreciate where and how to expend effort to attain performance(Miao et al., 2017).Therefore, prior experience underpins the role of entrepreneurial selfefficacy in performance.Yet, entrepreneurs with little or no previous experience but endowed with high entrepreneurial self-efficacy are incapable of leveraging experience and knowledge into notable firm performance(Miao et al., 2017;Ucbasaran et al., 2008).Thus, the positive relationship between entrepreneurial self-efficacy and firm performance will fluctuate depending on the women entrepreneurs' differing experience levels.
(Bosma et al., 2004;Miao et al., 2017)ormance and entrepreneurial self-efficacy differs as entrepreneurs have varied levels of previous entrepreneurial experiences(Bosma et al., 2004;Miao et al., 2017).Extant research posits that entrepreneurial self-efficacy and previous This brings the current study to focus on the relationship between entrepreneurial self-efficacy dimensions, business experience, and business performance of women-owned enterprises in the emerging economy of South Africa.Prior research shows a positive link between the founder's entrepreneurial self-efficacy and commonly utilized measures of entrepreneurial firm performance(Hallak etal., 2012; McGee & Peterson, 2019; Newman et al., 2019).On the other hand, Cumberland et al. (2015) revealed that entrepreneurial self-efficacy concerning management, innovation, and financial control positively affected firm growth in competitive environments.

Table 2 .
Validity and reliability scores Note: Extraction Method: Principal Axis Factoring.Rotation Method: Promax with Kaiser Normalization.a. Rotation converged in six iterations.ESE = entrepreneurial self-efficacy.

Table 5 .
Mediation analysis results