Gender related drives and aspirations
of nascent entrepreneurs
Diána Kürtössy
11374977
23-06-2017
Final
MSc Business Administration
Entrepreneurship & Innovation
University of Amsterdam
Tsvi Vinig
Author
Student number
Date of submission
Version
Course
Track
Institution
Supervisor
Statement of originality
This document is written by Diána Kürtössy who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
TABLE OF CONTENTS
1 Introduction ... 5
1.1 Research goals and research question ... 7
2 Literature review ... 9
2.1 Female entrepreneurship and gender studies ... 9
2.1.1 Gender discrimination ... 9
2.1.2 Masculinity of entrepreneurship ... 11
2.1.3 Career choices of females ... 13
2.1.4 Causes of female underperformance ... 14
2.2 Motivation to start a business ... 18
2.3 Self-efficacy ... 19
2.4 Attitude toward growth ... 20
3 Methodology ... 23
3.1 Research Design ... 23
3.2 Research instrument and data collection ... 23
3.3 Sample Characteristics ... 25
3.4 Variables and Measures ... 26
3.4.1 Independent Variable ... 26
3.4.2 Mediator 1: Self-efficacy ... 27
3.4.3 Mediator 2: Motivation to start a business ... 28
3.4.4 Dependent variable ... 29
3.5 Statistical Procedure ... 30
4 Results ... 31
4.1 Preliminary steps ... 31
4.2 Descriptive statistics ... 32
4.2.1 Normality and outliers check ... 32
4.2.2 Reliability testing ... 33
4.3 Correlations ... 34
4.4 Hypothesis testing ... 35
4.4.1 Mediator: Self-efficacy ... 36
4.4.2 Mediator: Motivation to start a business ... 37
4.4.3 Summary of effects ... 38
4.5 Additional analysis: Crosstabs ... 39
5 Discussion ... 40
5.1 Discussion and interpretation of results ... 40
5.2 Entrepreneurial implications ... 44
6.1 Conclusion of the research ... 44 6.2 Limitations ... 45 6.3 Future research ... 46 7 References ... 47 8 Appendix ... 52 8.1 Survey ... 52 8.2 Variables measurement ... 55 8.3 Crosstab results ... 58 8.3.1 Independence items ... 58 8.3.2 Recognition items ... 58
8.3.3 Financial success items ... 59
8.3.4 Marshalling items ... 60
8.3.5 Implementing people items ... 61
8.3.6 Implementing financials items ... 63
8.3.7 Gender and Growth_ ... 64
LISTOFFIGURESANDTABLES Figure 1. - Conceptual model ... 22
Figure 2. - Mediational model ... 35
Table 1. – Scales to measure the factors of self-efficacy ... 27
Table 2. – Items to measure the factors of self-efficacy ... 27
Table 3. - Scales to measure the factors of motivation to start a business ... 29
Table 4. - Items to measure the factors of motivation to start a business ... 29
Table 5. - Skewness and Kurtosis ... 32
Table 6. - Reliability of variables ... 34
Table 7. - Correlation matrix ... 35
Table 8. - Regression analysis on self-efficacy variables ... 37
Table 9. - Regression analysis on motivation to start a business variables ... 38
Abstract
The purpose of the thesis is to explore whether gender differences appear before entrepreneurs establish their businesses, in the phase when they only have the intention to do so. The study aims to examine the relationships between gender and growth aspiration, and two hypothesised mediators, namely self-efficacy and motivation to start a business answering the following research question: “Does gender have an effect on growth aspiration of nascent
entrepreneurs?”. This research tests the occurring gender-related differences in the literature
which casts its academic relevance by highlighting possible connections explained in academic discussions contributing to further understanding of motivations and perceptions of female entrepreneurs. 99 entrepreneurs’ answers are tested in the analysis, recorded by a web-questionnaire. The results indicate that neither the direct relationship nor the mediations are supported by the current research. However gender is found to be significantly related to one of the motivational and one of the self-efficacy factors, respectively achieving higher independence and implementing financials. These findings indicate that social role stereotyping and females lower self-perceptions proposed by the academics are supported by the current study.
Key words: entrepreneurship; gender; growth aspiration; motivations; self-efficacy; nascent entrepreneur; female entrepreneurship; gender discrimination;
1 INTRODUCTION
Female entrepreneurship. This is how academic texts refer to a phenomena, when women are involved in entrepreneurial activity. One question emerges immediately: why do we put the female tag in front of it? Let’s take sports as an example, for instance, swimming. Even though gender segregation appears, we do not label the races for women as women swimming and dedicate the swimming expression to one done by men. Why is it, that when it comes to economic activity the labels are used one-sidedly? In sports, sexes are separately competing as they are evolutionarily different, the biological systems of sexes do not work exactly the same. Why is it that in economic activities, the performance of genders are contrasted, labelling females as underperformers? This study aims to tap into this gender discrimination in economic activities, focusing on the field of entrepreneurship.
Being an entrepreneur comprises a positive and a prestigious essence, as it takes courage, endurance and tolerance for uncertainty. Its rising reputation certainly owing to the most successful entrepreneurs who shaped the 21th century’s thinking and way of living; Steve Jobs, the founders of Google, Sergey Brin and Larry Page, the visionary leader, Elon Musk or Jack Ma the e-commerce mogul, founder of Alibaba are among the most successful businessmen. In line with the structural advancement in technologies, they uplifted entrepreneurship to a whole new level by showing the perspective of self-actualization in this choice of occupation, hence being a startupper is something the new generation aspires. Another question emerges: where are the women founders from the toplists? Do they even desire to be among top entrepreneurs? Even though the Golden Age of entrepreneurship (TechCrunch, 2016), evidences show that this hazardous and unpredictable form of employment is not equally engaging to males and females. Regardless of slow progression, self-employment concerning gender still shows remarkable discrepancy in the OECD countries in favour of men. OECD Data from 2015 shows that self-employed men with paid employees are two and a half times more represented in
entrepreneurship than women regarding the same situation. This means, 2.2% of employed women, while 5.6% of employed men opt for establishing their own businesses with at least one employee on board (European Commission, 2016). In spite of the increasing number of women entrepreneurs lately, they still have less passion for entrepreneurship than males do (Verheul, Thurik, Grilo, & van der Zwan, 2012).
Researches affirm that female and male entrepreneurs are not even close with regard to their performances either. Recent studies result in a general agreement that female-led ventures are underperform the men-led ones in most cases. Although there are few established evidence on the reasons, women’s general characteristics and personal considerations such as risk aversion, family orientation, less focus on aggressive growth, keeping the work life balance could be the causes of this diversion between genders (Cliff, 1998; Fairlie & Robb, 2009). They are said to be less successful owing to less startup capital, smaller networks, in addition, their preferences and different motivations may also influence their performance (Fairlie & Robb, 2009). On the other hand, no difference was found in the business performance of ventures regarding the gender of founder, when the research is carefully measured and executed (Robb & Watson, 2012). Gender does not seem to be a significant direct explanation of the differences between small businesses’ financial performance and business growth (Collins-Dodd, Gordon, & Smart, 2004; Johnsen & McMahon, 2005). However, negligible amount of proven studies can be found on the contention, when female-established businesses outperform their counterparts. I note, that because the above mentioned papers have a slightly different scope, it is hard to conclude any generalizable result from them. Most of the reviewed articles are based on already existing large datasets gathered not especially for the purpose of the study, which makes further examining this issue academically relevant.
1.1 RESEARCH GOALS AND RESEARCH QUESTIONS
Before starting to approach the question of differences in business performance related to gender, one may need to take a step back and examine the circumstances of establishing and operating a venture. Each stage of this progress should be carefully studied in order to be able to articulate a precise overview on it. Due to their fast changing processes, different motivations and intentions may appear and disappear during the evolution of a business. Considering gender discrimination in entrepreneurship as a complex problem, I believe that examining it in smaller, carefully conducted steps can reflect on substantial information of what researches with broader scopes might miss out on.
The following study aims to tap into this research gap and examine nascent entrepreneurs’ growth aspirations in the phase, when they have not established their firms yet, they only have the intention in the nearest future to do so. By doing so, I strongly believe that this study can contribute to the bigger picture of the puzzle, namely to the literature of female entrepreneurship. Light can be shed on the question, whether gender differences are already existing in the zero phase of establishing a new venture and therefore they might exerting an effect on plans, or it is rather a later occurring and intensifying phenomena as the researches on business performance tell us so.
According to Davidsson (1989), when we study growth willingness, it is advisable to start with initial motivation aspects, which this study intends to do. Gundry and Welsch (2001) also reports, that there are few empirical work on entrepreneurial motivation in the stage beyond establishing a business. Cassar (2007) supports the idea to relate motivation to growth preferences. Self-efficacy is also shown to be strongly affected by gender and related to entrepreneurial intentions (Minniti, 2010, Sweida, 2013). Therefore I insist to introduce the following research question:
1. Does gender have an effect on growth aspiration of nascent entrepreneurs?
Additional factors influencing the relationship of gender and growth aspiration would like to be explored by this master thesis, namely:
2. Are there any possible motivational factor of starting a business which affects growth aspiration of a female entrepreneur?
3. Does perceived abilities to start and run a venture can influence the growth aspiration of a female entrepreneur?
The research is carried out by an online survey targeting nascent entrepreneurs, mainly masters and alumni students in universities in Hungary and the Netherlands. The research model builds up on the explained connections in the academic literature, the scales which were used to capture data were adopted from A* and A level peer-reviewed journals. The hypotheses are tested by a simple mediation regression analysis. To have a deeper understanding on the effects contributing to the results, a cross tab analysis were carried out checking frequencies of the items of the variables in light of gender.
The thesis is structured the following way: in section 2 a literature review is introduced on female entrepreneurship. Furthermore, the two possible gender related connections are discussed, which the master thesis is going to test, namely motivation to start a business and entrepreneurial self-efficacy, all of them are related to growth aspirations of nascent entrepreneurs. In section 3, the methodology is explained which describes how the data collection is initiated and method of data collection was carried out. Additionally, the sample characteristics and statistical procedure are briefly described. Next comes the results in section 4, where all the findings are presented which is followed by a critical discussion in section 5, reflecting back to the previously introduced knowledge according to the literature. Moreover, entrepreneurial implications are concluded from these learnings. A conclusion is closing the
thesis in section 6, summarizing all the contributions which the study unfolds. It discusses limitations of the current research and advancements, and lastly future research directions are suggested.
2 LITERATURE REVIEW
2.1 FEMALE ENTREPRENEURSHIP AND GENDER STUDIES
In the following review of academic publications on female entrepreneurship the reason why female tag is usually put in front of the expression, entrepreneur is discussed. These gender studies are not only appearing in economic journals, but also the field of sociology and psychology are dealing with the topic. In most cases, females and males are contrasted while studies focusing on females only appear as well. In this section a piece of the whole academic gender study literature is taken out and the focus is directed on becoming and being an entrepreneur as a female, from which the reader is about to learn just as much on males. The logic of order starts with gender discrimination in entrepreneurship and masculinity of entrepreneurship, afterwards occupational choice and eventually the myth of female underperformance are discussed. Each topic is reviewed from many perspectives contrasting several beliefs.
2.1.1 Gender discrimination
Gender discrimination is a common occurring discourse within socio-economic context. To further elaborate on how this real - or often just a perceived - inequality can affect the behaviour of women in entrepreneurship, first the term itself should be comprehended. Gender is
understood to be socially constructed, a product of historic, social and cultural meanings. It is understood to provide “socially produced distinctions between male and female, masculine and
feminine” (Patterson, Marvin, & Turner, 2012, p.: 397). In other words, gender is all the
characteristics and behaviours attributed to sexes in a certain culture.
In line with the review of Bruni, Gherardi and Poggio (2004), three kinds of obstacles have been found in the literature in connection with female entrepreneurs. The first is the previously prospected social role within the meaning of this provision women should come up to the society’s expectations, usually primarily being responsible for the family, including domestic life and taking care of the children (Marlow & McAdam, 2013). These beliefs are deeply embedded in several, even developed countries’ cultures, putting them into a disadvantaged position, since they are anticipated to behave likewise the socially constructed roles, traditionally being the homemakers while men are the breadwinners (Patterson, Marvin, & Turner, 2012).
When responding to these overwhelming social imperatives and ascribed roles, women face the second barrier that is to say they have less remaining time to socialize and to build relationships and shape their own network, which are strategic resources of starting and running, if so growing a new business successfully. They usually develop smaller networks, resulting in fewer investors (Klyver & Grant, 2010) and they ought to seek other women’s circles rather than the company of males when in the early phase of business development (Hampton, Cooper, & McGowan, 2009). This eventually leads to having a smaller network as there are fewer women entrepreneurs compared to their counterparts, however these relationships can be deeper. Additionally, male networks have their own characteristics, based on formal but just as much on informal social activities (Hampton, Cooper, & McGowan, 2009) of which it is hard to connect to for women in a similar way than men can do so. Males utilize these connections more often and effectively, but these do not associate with firm performance eventually (Watson, 2011).
Even when women do have opportunities to network, it is still more difficult for them to get access to angel investors or venture capital (Bruni, Gherardi, & Poggio, 2004) which counts as the third obstacle for women entrepreneurs. The masculinized cultures and the behaviours accepted accordingly makes nearly impossible for women to win capital in the SET (science-engineering-technology) industries (Marlow & McAdam, 2013). In addition to this distinctness regarding access to capital, women feel the pressure of financial constraints, furthermore they perceive more often that funding is a barrier to starting a new venture (Kwong, Jones-Evans, & Thompson, 2012), which can be one reason that they tend to launch enterprises in industries which are less capital intensive (Coleman & Robb, 2009). Fortunately more opportunities emerging with the spread of crowdfunding platforms, which democratizes the possibility to launch a project by winning capital even for women entrepreneurs, however it is noticed that they largely differ by their fundraising targets (6,300 $ versus 9,400$) on average, resulting more considerate amount targeted by women fundraisers (Marom, Robb, & Sade, 2014). This can be explained by their risk-aversion, that they need to reach the targeted goal in order to keep that capital or even by their lack of self-efficacy, that they underestimate their capabilities to run a successful campaign to reach the target goal. But what is the root of these insecurities and hardship they are facing within entrepreneurship? In order to be able to highlight possible explanations, digging deeper in the discourse of entrepreneurship is inevitable.
2.1.2 Masculinity of entrepreneurship
When trying to identify the reasons behind these discriminations against female entrepreneurs, it is advisable to start with unfolding the past. It is proposed that the masculinisation of
entrepreneurship as well as male gender-role stereotype is deeply rooted in the entrepreneurship
discourse (Gupta, Turban, Wasti, & Sikdar, 2009; Orser, Elliott, & Leck, 2011) which could be a possible deterrent for a woman weighting her prospects before entering the career market. This is supported by the finding that not only males, but females are perceiving
entrepreneurship as a masculine occupation, and this masculinity is found to be positively correlating with entrepreneurial intentions (Gupta, Turban, Wasti, & Sikdar, 2009). Consequently females who have less masculinity might be reluctant to opt for starting and running a business or at least it is raising their doubts in themselves. A positive reinforcement is that women can relate the entrepreneurial and feminine attributes in entrepreneurship, however men cannot connect them (Gupta, Turban, Wasti, & Sikdar, 2009), which is not satisfactory if we would like to have the embeddedness of this stereotype eliminated in the future, in order to encourage women taking up entrepreneurial occupation.
Orser, Elliott, & Leck (2011) refer on different authors and uses a really strong expression, ghettoized on the treatment of female entrepreneurs. With this expression in my opinion the authors exaggerated a little bit too much, even though several studies (Bruni, Gherardi, & Poggio, 2004; Marlow & McAdam, 2013 among others) concluded that it is harder for women to get access to capital and finance and get recognition. It is also emphasized that institutional actors often behave in line with social perceptions even unintentionally (Bruni, Gherardi, & Poggio, 2004). Despite all these arguments, their situation is not that forcefully disappointing, for instance in the OECD countries many steps are being taken to emphasize the importance of female led ventures and to encourage them, as they can provide different perspectives and approaches to business management, organizational culture among other issues (OECD, 2016). What is seen from the already introduced views, findings and initiations regarding female entrepreneurs, these culturally produced and socially learned stereotypes affect occupational segregation in a complex way (Gupta, Turban, Wasti, & Sikdar, 2009) and each author come to a conclusion about what effect could be the strongest explanation of the issue of having less female than male entrepreneurs, such as risk aversion, hardship in accessing venture capital or the masculinity of the occupation. Although many studies examine females and males and their characteristics and roles attributed to them in the early phase of entrepreneurship, Gupta,
Turban, Wasti, and Sikdar (2009) argue that gender identification could be even more related to the decision for a person to start a business than simply gender. It would be an interesting direction to extend the reviewed researches.
2.1.3 Career choices of females
A conclusion from the previous section could be derived, namely that stereotyping and gender discrimination somehow and to some extent affects career choices of women. Despite entrepreneurship is seen as a manly work among members of the society, women have broader views on the gender-role stereotype and they can relate themselves to being an entrepreneur. However man still not concede its feminine attributes, which causes a hurdle for women to be taken seriously by their counterparts (Gupta, Turban, Wasti, & Sikdar, 2009). These hardships occur more often in the stage of career choice, not once they are running their businesses. This can be the phase where the problem is rooted. Women are dissuaded by the attitude they face by men and certainly not reaching the phase to start a business in several cases (Verheul, Thurik, Grilo, & van der Zwan, 2012). To reflect back on the previous argument, if a woman has more masculinity, she might be able to overcome those given obstacles. But what if she does not own those masculine characteristics? The issue can be treated either by encouraging women (as governments and organizations should do so (OECD, 2016)) to neglect those judgements and carry on with their aspirations or by promoting female entrepreneurs’ success as much as we do it with men.
According to Sweida and Reichard (2013) women entering self-employment face dual gender stereotypes: one is that certain industries have higher masculinity included and the other is that entrepreneurship is a masculine type of employment. These pressures may hold back females to enter high-growth industries, however it is found that women generally have lower preferences for even entering self-employment (Verheul, Thurik, Grilo, & van der Zwan, 2012).
It is interesting to contrast the probability to become an entrepreneur and age in light of gender. In the study carried out by Noguera, Alvarez, and Urbano (2013) involving 4000 Spanish entrepreneurs it is found, that while women show an inverted U-shaped curve, men have a negative relationship. This means, men tend to start their entrepreneurship career in a really young age while women are considerate at first, but with time it is more likely for them to start their businesses, until it reaches a certain point where it starts to decline (Noguera, Alvarez, & Urbano, 2013). Despite the slight generalizability of results, there is a reason to believe from the above discussed points that it can be a valid case in other developed countries of the European Union. The results indicate that women tend to collect experience and self-confidence beforehand and it is most probable that they choose self-employment once they are re-entering the labour market.
2.1.4 Causes of female underperformance
As discussed earlier, women prefer less to become entrepreneurs and there are some structural differences such as they are eager to run smaller ventures, underrepresented in manufacturing and construction industries, are less export-oriented, and disproportionately dependent on households as customers (Du Rietz & Henrekson, 2000). They are more likely to have insufficient education, less optimal or “feminine” management strategies or experience and less desire to start a business. In addition the claims that they are more risk averse, facing unique startup difficulties or training needs, behaving irrationally and turning to family for help and not networking optimally (Ahl, 2006), may have an effect on their lower level of performance (Verheul, Thurik, Grilo, & van der Zwan, 2012), often referred as the phenomena of female “underperformance”. Women’s businesses are usually characterised by smaller size, constrained growth and lower profitability compared to their counterparts, they are pictured as caring, nurturing, less innovative because of their duties not only at work but in the family as well, therefore they expected to be less successful (Orser, Elliott, & Leck, 2011). This
underperformance is seen problematic according to the economic growth theory (Ahl, 2006). In contrast, entrepreneurs are portrayed as having characteristics of a hero for instance daring, decisive, ambitious, having will to conquer (Orser, Elliott, & Leck, 2011) and determined, forceful, adventurous, brave, visionary, active, risk taking, motivated and achievement oriented (Patterson, Marvin, & Turner, 2012). These masculine standard norms are fostering the discrimination of female entrepreneurs. They are judged according to the majority, whose performance and behaviour set the standards. Therefore women are marked out as ‘others’ or to say so, as female entrepreneurs (Lewis, 2006). There is an invisible impression of masculinity (Ahl, 2006; Bruni, Gherardi, & Poggio, 2004), which directs behaviours without even being recognized, making men and entrepreneur interchangeable words, while female characteristics marked by the “female entrepreneur” expression often formulate something different or “other” (Patterson, Marvin, & Turner, 2012).
The first cause of the so called female underperformance found in several studies is women’s preference to enter much low-returning service industries than their counterparts, who tend to dominate high-growth industries. This can explain the observed difference between genders, however it should not be called underperformance, strongly emphasized by Marlow and McAdam (2013). They reflect to this discrimination or the misused “under-performance” phenomena on women as it is created by the lack of control for business size and sectoral distribution statistically. These misused analyses creates axioms on women, and this femininity which impede growth in this mythical theory is used as evidence for the performance differences in entrepreneurship. Furthermore Sweida and Reichard (2013) are also concerned about statistical measures. They think that the reason that men are major in growth, high-tech and manufacturing industries, while women tend to enter low-growth, low-skilled industries which indicate lower growth on the most common performance measures (employees, sales, profitability, market share) can account for the underperformance noticed in
many researches. However other considerations are emerging, for instance, that the measures used in researches are masculine in nature.
The attention is caught on the so called ‘malestream’ phenomena, under which masculine norm has become the reference point from what women’s traits and behaviour regarding entrepreneurship is measured. When economics started to influence the discourse of entrepreneurship, researchers started to implement economic measures as growth, profit, firm size together with these objective (masculine) ontology. Many researchers in the past have neglected the gendered nature of their research besides the importance of economic growth rationale. By doing so the field suffered from this and masculinity set the standard measure lacking the discovery of power relations of gender social role (Patterson, Marvin, & Turner, 2012).
In addition to the biased performance measures, authors mention the ignorance of the roots of these differences between genders. Historical and cultural concepts are often left unchallenged and sex specific personal attributes are highlighted instead, which are predecessors of the essential differences between them, leading to a female underperformance conclusion, building on their inadequacies and shortfalls (Patterson, Marvin, & Turner, 2012, p.: 404). In line with this thought, Marlow and McAdam (2013) claim that when researchers explain the underperformance assumptions with the concentration of women in low performing sectors and the higher likeliness that they work from home or part time, they certainly do not consider that these characteristics have different epistemological roots and require careful explanatory analysis. They rather call it constrained performance which still carries the gendered subordination but derives from different roots.
Patterson, Marvin, and Turner (2012) collected some evidences on the masculine root of the concept of entrepreneurship, such as Ahl (2006) found disparity between feminine and entrepreneur characteristics described. Bruni, Gherardi and Poggio (2004) found that either men
or women business owners adjusted their behaviours to be masculine in order to meet with the masculine discourse of entrepreneurship. Women who act masculine mismatch their socially perceived female bodies causing discrepancy, and eventually disrupting the gender social order which is devalued in the eye of males.
Eventually some authors argue against this underperformance with several evidences found in their studies. For instance, Marlow and McAdam (2013) consider underperformance a myth which places emphasis on performance indicators which are in fact deeply gender biased. This causes the spread of the female deficiency when it comes to competency and ambitions of an entrepreneur. Du Rietz and Henrekson (2000) confirmed the female underperformance phenomena but only on an aggregate level. When they included control variables in their analysis, this phenomena only appeared in one performance measure out of four, which is sales. This, however is weak in large firms and not existing in businesses with one employee. All in all, they found no evidence for the female underperformance hypothesis, as profitability measure was not statistically significantly different in any of the cases.
Another interesting perspective reflected by Patterson, Marvin and Turner (2012), is that most of the males do not perform according to the masculine economic growth theory, but their few growth oriented fellow male entrepreneurs actually save their image (Ahl, 2002), thus leaving only their female counterparts to count as the underperformers. In contrast, it is found in the literature that males do start small and stable businesses, just as well as women (Lewis, 2006) and no difference was found between genders when they are operating with the same input (Watson, 2002).
Recent literature reports that women these days can experience better conditions when it comes to entrepreneurship. The entry barriers of the service sector are getting lower which can provide them with more opportunities as one can not only prevail with a possession of niche knowledge or enormous capital anymore. It might be the case that self-employment is going to be as good
alternative for women in the future as employment. Traditionally, they enter the low performing industries which are seen less masculine, not like SET (science, engineering and technology) sectors (Marlow & McAdam, 2013), but new technologies and younger generations may find their creative way to open up those barriers for women, just like crowdfunding did with the hardship of accessing capital (Marom, Robb, & Sade, 2014). There are initiations across the world to involve women more into coding and IT development, hence giving them better understanding and thereby easier access to entrepreneurship in dynamic SET industries.
2.2 MOTIVATION TO START A BUSINESS
As it is discussed in the previous review, one of the possible causes of female underperformance phenomena is that women often found to have different motivations to start their businesses than their counterparts, which are not in line with the commonly used performance measures. Based on their socially expected roles, they tend to pay attention to balance work and family life, they enter self-employment for greater independence and autonomy, and they search for self-fulfilment, for income and to pursuit a social mission (Bruni, Gherardi, & Poggio, 2004) and not especially to build wealth or grow their ventures big. Males often put greater emphasis on economic values, quantitative measures of achievements, while women tend to appreciate social values and qualitative measures of success (Cliff, 1998).
The motivation to start a business is found to be related to growth aspiration in Cassar’s (2007) study, therefore I intend to adapt the measure used in the study in order to better explore how gender may affect initial motivations and how these factors of motivation can be related to growth aspiration. He found that the emphasis put on gaining recognition and self-realization is associated with greater growth intention, while striving for independence is negatively related to growth aspiration. Therefore, women are expected to be motivated by reaching independence and by this, their lower growth aspiration is forecasted. On the other hand, males are rather opt
for greater growth, hence gaining recognition and self-realization are expected to be higher in their case.
2.3 SELF-EFFICACY
Owing to the gender discrimination and masculine discourse of entrepreneurship, females perceive themselves less prepared and less competent of being an entrepreneurs in many cases, also men do perceive women less entrepreneur-like (Gupta, Turban, Wasti, & Sikdar, 2009). As women according to the literature tend to own low-growth and low-performing businesses, often deliberately (Cliff, 1998), it is expected that they are lower on self-efficacy. This is the reason the study proposes that self-efficacy should be discussed in a gender relation.
Entrepreneurial self-efficacy (ESE) is an important explanatory variable in determining both the strength of entrepreneurial intentions and the likelihood that those intentions will result in entrepreneurial actions (Boyd & Vozikis, 1994, p.: 66). Self-efficacy affects one’s commitment
and goal settings, moreover it is seen as self-confidence regarding tasks. Someone with high self-efficacy is expected to plan more challenging goals and develop more efficient strategies to achieve them, to exert more effort for longer time and to own stronger commitment to one’s goals. (Boyd & Vozikis, 1994, Shane, Locke, & Collins, 2003). Women are found to be lower on entrepreneurial self-perception which may make them insecure about their skills and knowledge for taking up self-employment (Verheul, Thurik, Grilo, & van der Zwan, 2012).
Recent studies found relationship between gender and entrepreneurial self-efficacy, and it is shown that self-efficacy mediates the relationship of gender and entrepreneurial intention (Sweida & Reichard, 2013). It is recommended to analyse entrepreneurial self-efficacy’s effect on high-growth intentions of individuals. Among business owners, women perceive themselves better in time management, controlling organizing, planning and informing, while males are reported on higher commitments on quantitative, decision making, problem solving and
financial skills compared to the other gender (Orser, Elliott, & Leck, 2011). Noguera, Alvarez and Urbano (2013) summarize previous studies in which women perceive their entrepreneurial skills lower regardless of their real skills especially when it comes to masculine industries. They also point out that men perceive themselves more positively in terms of skills than their counterparts.
2.4 ATTITUDE TOWARD GROWTH
Growth aspiration has been chosen to be the dependent variable, therefore to be the focus of this study for several reasons. First of all, I believe that it has a connection with future business performance of a firm, as Cassar’s (2007) longitudinal study pointed out. Secondly, growth in the number of employees and/or in the number of sales are really common to be the dependent variable in the previously reviewed studies on business performance and gender. Last but not least, traditional categorization of firms also uses the number of employees to define an entreprise as micro, small, medium or large, which all have their own characteristics based on their size. It is expected that different motivations and different levels and kinds of self-efficacy can result in different planned businesses regarding their size as well, making growth aspiration an appropriate indicator.
Moreover, growth (let it be in the number of employees or sales) is also a common occurring natural phenomenon, when it comes to profit maximizing in economic theory according to Kolvereid (1992). The assumption that people act on behalf of growth until a firm reaches its optimal size is not considered valid anymore among small businesses. People dive into establishing and operating a business beyond maximizing economic returns, they are capable to and they do control the intensity and the level of growth (Kolvereid, 1992). Therefore I assume that growth aspiration can be a good indicator of different future intentions of business owners, and motivations can relate to this intention, which is assumed to vary among genders
taken the reviewed literature into account. Since the study builds on nascent entrepreneurs coming from different countries, the study does not measure sales as an indicator of growth intention, but the number of employees, because latter is a measure which is generally accepted and easily comparable through different economies.
Growth willingness and aspirations often occur as dependent variables in connection with motivation and self-efficacy. Davidsson (1989) suggests that if one would like to study growth willingness, start with psychological theories of motivation. Sexton and Bowman-Upton (1990) points out that “those in control of the firm may initiate, foster, nurture, or prune growth in
accordance with their own propensity for growth and their abilities to manage it” (1990, p.:
137), which thought underpins the idea to examine growth based on the characteristics, perceived abilities and motivations of entrepreneurs. Theoretically, women and men often happen to start businesses for different reasons, therefore it is assumed that their motivation mixes are different. Genders differ in relation with how they value business expansion. While men feel the pressure from society to be competitive and dominant, therefore they more likely to measure success by the size of the venture, women put emphasis on interpersonal relationships which are more subjective values. These different focuses can result in difference of the growth aspiration of the genders. (Cliff, 1998). Kolvereid (1992) studied Norvegian entrepreneurs whose motivations are related to growth aspiration, however the relationship does not seem to be strong. The difference reported between male and female owners is in the expected direction – female are more considerate than men –, but not statistically significantly. He argues that it is possibly owing to Norway’ egalitarian culture, where there is no sharp distinction among genders. It is important to note that this conclusion derives from established firm owners, and gender and their aspiration to grow might have a less strong relationship once the firm is established as it enters into a different stage, where initial motivations are not quite relevant anymore therefore biases often occur when it is examined retrospectively. In addition,
those entrepreneurs who did not succeed are not involved in the sample (Cassar, 2007). Cliff’s (1998) qualitative findings suggest that women may either choose to run low-growth ventures deliberately, or they rather choose to grow slowly, while growth is in line with the family roles of men. He reflects on many findings which supports this idea, but he points out that studies are often based on women-only samples.
Based on the previously introduced theories and researches, the current study intends to test a newly proposed conceptual model, where growth aspiration of nascent entrepreneurs is hypothesised to be affected by different motivations to start a business and different aspects of self-efficacy. All these three variables are expected to be affected by gender. Accordingly, the master thesis hypothesizes the following relations:
H1: Women are negatively associated with growth intentions.
H2: Self-efficacy mediates the relationship of gender and growth aspiration.
H3: Motivation to start a business mediates the relationship of gender and growth aspiration.
H2 H3 H1 Growth Aspiration Gender (Female) Motivation to Start a Business Self-efficacy
3 METHODOLOGY
In the upcoming chapter the research approach and research design of this master thesis are discussed. First of all, the research design is introduced, then the research instrument, the target sample and the data collection procedure follow. Next, the sample characteristics, then the measurements and variables are presented and lastly the statistical method part is closing this section.
3.1 RESEARCH DESIGN
The thesis is conducted by a positivist philosophy as the qualitative research aims to disclose causes and effects of variables which are observable and measurable. In line with that, theory can be proposed, tested and refined up to predicting reality accurately (Saunders & Lewis, 2012). Deductive approach is applied throughout this study as the literature on female entrepreneurship is used as the baseline of the research from which the conceptual model and hypotheses are derived and tested. Taken into account that the timeframe of the master thesis is limited, a cross-sectional study was conducted, which intends to collect and analyse data at only one period in time (Saunders & Lewis, 2012). In order to have the chance to generalize the findings, a questionnaire was used for data collection, which is the most common method to test the hypotheses in my model conducting a quantitative research.
3.2 RESEARCH INSTRUMENT AND DATA COLLECTION
The distribution of the questionnaire followed a multi-channel approach as it was posted on formal and informal online channels. Social media channels were the primary source for spreading the survey, groups of entrepreneurs and graduate students from Hungary and the Netherlands were asked in Facebook and LinkedIn groups. In addition, the survey was sent out
to all current Economics and Business masters students at the University of Amsterdam in the year of 2016/2017 by email. Furthermore, the Alumni of the Economics and Business faculty of University of Amsterdam were asked by a newsletter to join the survey. Moreover, it was sent out on a mailing list of members of a Hungarian student organization of Corvinus University of Budapest, namely the EVK College for Advance Studies.
A low response rate was expected beforehand due to high-time season among master students to complete different surveys and the difficulty to detect nascent entrepreneurs. Therefore the complete sampling frame was not expected to be reached and a non-probability sampling technique were chosen combining purposive and self-selective sampling methods (Saunders & Lewis, 2012). Initially the groups of relevant respondents were selected (purposive sampling) – the students of business administration and entrepreneurship at Corvinus University of Budapest and University of Amsterdam, and then they could decide whether to participate in the research (self-selective sampling). Filling out the web-questionnaire was anonym, confidentiality were promised and kept.
Graduate students are expected to understand business English, hence the questionnaire was only spread in English language in two countries, the Netherlands and Hungary. Before it was posted, post-graduate students from the University of Amsterdam tested each blocks, who found everything in a good order. The questionnaire was open to fill in from 9th of April until the 1st of June. In order to look more appealing and professional, an electronic survey by Qualtrics software was used, which contained the headline and style of the University of Amsterdam, Faculty of Economics and Business (Appendix 1).
The survey is specialized on nascent entrepreneurs, hence a filter was originally built in, and only those respondents were allowed to carry on the survey who definitely or probably wants to start a business within a coming two years. To reach a higher response rate, entrepreneurs who already started their business no more than one year, were allowed to answer as they are
expected to remember correctly to their initial motivations. The scales used in the survey are adopted from peer reviewed journals and were all validated in those studies (Appendix 10.2).
The survey is divided into five blocks: The first part covers asking on the independent variable, gender and a filter which only let nascent entrepreneurs to respond the survey. The second block contains scales on motivations to start a business, the third is about perceived self-efficacy and the fourth one is the dependent variable, growth aspiration. A short demographic section closes the survey, asking information on age, level of education and current residency. The survey took 5-6 minutes to fill in.
3.3 SAMPLE CHARACTERISTICS
By running the survey for one and a half months, I managed to collect 276 responses. Even though the response rate can only be estimated, I assume that approximately 2000 people were reached by the online questionnaire from which 276 people were willing to answer, therefore the response rate is almost 14%. It is important to note that even though the strong emphasis put on people who willing to start a business each time the survey was posted, only half of the respondents (7%) falls into the target group.
According to the statistics of the survey, 192 respondents want to start or already started his/her business of which 162 people were allowed to fill in the survey, 121 answered until the end of second block, 105 answered the third and 104 the fourth block, which means 58 respondents quit the questionnaire, resulting a 35% dropout rate. The overwhelming season, the low interest and the length of the survey could be possible explanations for this high percentage of dropouts. After controlling for the respondents who answered the questions of all variables (by the 4th block) 104 responses are kept for conducting the analysis. It is important to note that respondents who did not answered all variables were deleted listwise.
After cleaning the sample, 100 respondents remained for analysis of which 58% are male and 42% female. While 43% of the respondents probably want to start their businesses, 39% definitely want to start and 18% have their business running less than 1 years. The average business size they aspire to reach in the fifth year of operation is on average 31 person including the founder, however if we exclude one outstanding response with 500, it decreases to 26 people. The demographic section is answered by 94 respondents, of whom average age is 28 years. 87% of them are currently involved or were involved in higher education before and 91.5% of them are from Hungary and from the Netherlands (their ratio is 50-50%).
3.4 VARIABLES AND MEASURES
As it is shown in the conceptual model, there are four variables included which are related based on the literature. The coming section intends to explain their roles in the model and how these variables are measured within the online survey. It is important to note, that all variables are adopted from original scales tested and used in peer reviewed A* and A level journals in order to ensure the reliability of these scales.
3.4.1 Independent Variable
In the model gender serves as the independent variable which are converted into dummy variables (0 = Male, 1 = Female), males being the base category. As it was explained earlier in the literature review, many researchers found evidence on differences in the intentions, motivations and perceived confidence is their abilities based on a simple variable, which is gender. This study tries to prove or confute those findings on the sample of nascent entrepreneurs. The independent variable is a categorical / nominal variable in the model.
3.4.2 Mediator 1: Self-efficacy
The five factors of Self-efficacy serve as one of the mediators (M1) in the model, as in McGee et al. (2009) outlined its effect to nascent entrepreneurs’ intention. It is hypothesised from the literature review that women and man entrepreneurs perceive their capabilities differently, former often seeing themselves less capable of strating and running business (Noguera, Alvarez, & Urbano, 2013; Verheul, Thurik, Grilo, & van der Zwan, 2012). Therefore i assume that self-efficacy mediates the relationship of gender and growth aspiration of nascent entrepreneurs.
Table 1. and Table 2. summarize the factors of the mediator and the items they are measured
by. The reliability of the original scales are extremely high (≥ .80), therefore those measures can fully support the goal of the research. One of the original factors is ommitted from the questionnaire, as it intends to measure the attitude toward venturing, which is not relevant in the sample, as respondents have already decided to start their businesses, therefore it is assumed, that they find it valuable, positive and rewarding, otherwise they would look for more rewarding and valuable ways of employment with positive associations.
Table 1. – Scales to measure the factors of self-efficacy
Factor of self-efficacy Author Year of publication Originality of scale Type of scale Number of items Original Cronbach's alpha Searching McGee et al. 2009 Original 5 point likert 3 .84 Planning 4 .84 Marshalling 3 .80 Implementing people 6 .91 Implementing financials 3 .84
Table 2. – Items to measure the factors of self-efficacy
Factor of
self-efficacy Items
"How much confidence do you have in your ability to … ?"
Searching
1. Brainstorm (come up with) a new idea for a product or a service 2. Identify the need for a new product or service
3. Design a product or a service that will satisfy customer needs and wants
Planning
1. Estimate customer demand for a new product or a service 2. Determine a competitive price for a new product or a service
3. Estimate the amount of start-up funds and working capital necessary to start my business 4. Design an effective marketing/advertising campaign for a new product or service
Marshalling
1. Get others to identify with and believe in my vision and plans for a new business 2. Network – i.e. make contact with and exchange information with others
3. Clearly and concisely explain verbally/in writing my business idea in everyday term
Implementing people
1. Supervise employees 2. Recruit and hire employees
3. Delegate tasks and responsibilities to my employees in my business 4. Inspire, encourage, and motivate my employees
5. Train employees
6. Deal effectively with day-to-day problems and crises Implementing
financials
1. Organize and maintain the financial records of my business 2. Manage the financial assets of my business
3. Read and interpret financial statements
3.4.3 Mediator 2: Motivation to start a business
The second mediator (M2) is motivation to start a business containing 6 factors which were designed to measure the career reasons of nascent entrepreneurs (Carter, Gartner, Shaver, & Gatewood, 2003). As it was discussed in the literature, male and female entrepreneurs often start their businesses for different reasons mainly owing to their social roles. These different reasons could be the explanation of different business size and business performance in the future. Based on the following logic, this measure could be a mediator in the relationship of gender and growth intentions of nascent entrepreneurs. Table 3. and Table 4. summarize the factors of this variable and the items by which those factors were measured. Compared to self-efficacy (M1), it is immediately observable that the reliability of the scales are not high, in three cases it is not reaching the acceptance (>.70) level. The reason why the study adopted this scale is that it is building on many previously validated scales to create an even more appealing one (Carter, Gartner, Shaver, & Gatewood, 2003). Furthermore the authors emphasize that is always harder to reach high reliability when the scale consists of only two items, but they kept them even though their marginal reliability as it can help to explain differences. In addition, the research were published in an A* journal, Journal of Business Venturing, which can ensure that the authors argument is legit.
Table 3. - Scales to measure the factors of motivation to start a business
Factor of motivation
to start a business Authors
Year of publication Originality of scale Type of scale Number of items Original Cronbach's alpha Innovation Carter et al. 2003 Original 5 point likert 3 .63 Independence 2 .58 Recognition 3 .60 Roles 2 .73 Financial success 3 .76 Self-realization 4 .78
Table 4. - Items to measure the factors of motivation to start a business
Factor of motivation to
start a business Items
"To what extent are the following reasons important to you in establishing this
new business?"
Innovation
1. Innovative, forefront of technology 2. Grow and learn as a person 3. To develop an idea for a product Independence 1. Free to adapt my approach to work
2. Get greater flexibility for personal life
Recognition
1. Gain a higher position for myself 2. Achieve something, get recognition
3. To be respected by my friends Roles 1. Follow example of a person I admire
2. To continue a family tradition
Financial success
1. Financial security
2. Earn a larger personal income 3. Build great wealth, high income
Self-realization
1. To challenge myself 2. To fulfil a personal vision 3. To lead and motivate others 4. Power to influence an organization
3.4.4 Dependent variable
The dependent variable in the model is growth aspiration, which is measured by a number based on how many employees certain prospective business owner imagine in his/her business in the 5th year of operation. Within the literature business performance is a common occurring dependent variable, measuring sales growth or profitability, however in the current study different measure should be applied to be able to capture the aspirations of nascent entrepreneurs and see whether they differ based on gender. This is the reason that another commonly applied measure, growth aspiration or often referred as growth motivation is
introduced. According to the study of Delmar and Wiklund (2008), growth motivation in the number of employees is significantly and positively related to employment growth, giving a reason to believe that it can differentiate nascent entrepreneurs’ aspirations even in this early phase of starting a business. Growth aspiration is originally an index constructed of four items, however researchers took items out of the index which they considered applicabale in their cases (Delmar & Wiklund, 2008; Wiklund, Davidsson, & Delmar, 2003; Wiklund & Shepherd, 2003). Although the authors used sales growth in their studies as well, it is omitted from this rearch as the sample is international. Therefore sales is not found to be a sufficient indicator. The simple explanation is that economies differ not only in currencies but also in GDP and other economic indicators, while the size of the business is comparable through economies, hence it provides a good basis to distinguish between aspirations. The other two possible measures are also omitted, because they intend to measure growth by % from a baseline, therefore it can be applicable in the case operating ventures, but not in planned businesses. In the e-survey, respondents were asked to report the ideal size of the business in five years in terms the number of employees (Appendix 1, block 4). This is a single number, which were the categorized based on the European Commission (2016)’s report on SMEs, accordingly micro (>10 employees), small ( 10 ≤ x < 50 employees), medium ( 50 ≤ x < 250 employess) and large (≥ 250 employees) entreprises in the dataset.
3.5 STATISTICAL PROCEDURE
The statistical analysis is supported by the SPSS Statistics software of IBM. First, some preliminary steps are taken such as dealing with missing values, cleaning the database, coding dummy variable and categorizing the dependent variable in order to make the dataset ready for carrying out the analysis. Next, the descriptive statistics are performed including normality check, reliability analysis (Cronbach’s alpha), computing the scale means of the mediator
variables and finally compiling the correlation matrix. Afterwards multiple regression analysis is carried out by an additional PROCESS macro developed by Andrew F. Hayes. This adds-on enables users to carry out simple and multiple mediation analysis by model number four. It is also suitable for testing the direct effect of the independent variable to the dependent variable. By doing so, the statistical significance of the model is going to be tested in order to investigate the effect of the independent variable and mediations to the dependent variable. Thereafter some additional tests are carried out in order to identify further possible connections within the dataset.
4 RESULTS
4.1 PRELIMINARY STEPS
Firstly, dealing with missing values is carried out. All answers which were not completed by the fourth block, asking on the dependent variable are deleted list wise from the sample. Since the dropout rate is 38%, this means 58 answers were deleted leaving 104 responses. Then the extreme outliers were deleted based on their answer to the dependent variable, leaving 99 responses, 61% of the targeted respondents for the final analysis. This carries the risk of reducing the effective sample size and power, however pairwise deletion can result mathematical inconsistency in the sample while substituting the means also carries distortion in the results. Therefore I find list wise deletion the most effective way to clean the dataset even though the sample size is much smaller this way.
After taking those steps, the independent variable is coded into dummy variables, having male as the base (male=0, female=1). Scale means were calculated for the mediator variables and categorizing the continuous values of the dependent variable is also carried out. As it was previously described, it is done on the basis of the SME categorization, therefore responses
ranging from 1-9 employees fall into category 1 (micro enterprise), from 10-49 employees fall into category 2 (small enterprise) and responses from 50-249 employees fall into category 3 (medium enterprise) . All other answers up to 250 employees would fall into category 4 (large enterprise), however there were no responses above 250 in the cleaned dataset.
4.2 DESCRIPTIVE STATISTICS
It is important to note that in order to keep the consistency within the tables, the results are reported in APA style, accordingly all variables are used in their abbreviated forms in which they are labelled in SPSS (see table 5.).
4.2.1 Normality and outliers check
For all the variables normality check is run, except for the dummy. The Kurtosis exceeds the +1 barrier in two cases SelfReal and ImplPeople, which means the distribution is leptokurtic, meaning that many scores are in the tails and the shape is pointy. As skewness is between +/- 1, no transformation is needed on the data. The variables are skewed close to normal distribution, however Independence and SelfReal are over the -0.5 score, meaning that the distribution of scores are skewed right, while Roles is over the 0.5 score meaning that its distribution is skewed to the left, but they still remain within the acceptable interval enabling to run the regression analysis on each variables.
Table 5. - Skewness and Kurtosis
Variables Denomination Skewness Kurtosis
Growth_ Growth aspiration (Categorized) .125 -.791 Innovation Motivation: Innovation (Scale Mean) -.047 -.200 Independence Motivation: Independence (Scale Mean) -.582 -.400 Recognition Motivation: Recognition (Scale Mean) -.089 -.692 Roles Motivation: Roles (Scale Mean) .619 -.173 FinSuccess Motivation: Finance (Scale Mean) -.240 -.454 SelfReal Motivation: Self-realization (Scale Mean) -.734 1.179 Searching Self-efficacy: Searching (Scale Mean) .016 .230 Planning Self-efficacy: Planning (Scale Mean) .493 .145
Marshalling Self-efficacy: Marshalling (Scale Mean) .101 -.143 ImplPeople Self-efficacy: Implementing people (Scale Mean) -.497 1.333 ImplFinancial Self-efficacy: Implementing financials (Scale Mean) .161 -.555
4.2.2 Reliability testing
Reliability testing is conducted on the variables. Even though previously validates scales were used to collect data, reliability found to be unacceptable in all cases. All variables, which has the Cronbach’s alpha ≥.70 are kept unchanged, and all variables below this reliability criteria are compared to the Cronbach’s alpha of the original, because in case of the variables of Motivation to start a business, some scales are not reaching the sufficient reliability criteria originally, however the authors kept them despite the marginal reliability. I follow their guideline and consider the variables in the analysis, which have a Δ ≤ .10 to the original scale or to the .70 criteria (*)(**). Therefore Innovation, Roles, SelfReal, Searching and Planning are deleted from the variables based on their marginal reliability and Independence is kept. In case of Marshalling, one item (Clearly and concisely explain verbally/in writing my business idea in
everyday term) is deleted (2 items remained), to reach higher reliability .660. New scale mean
Table 6. - Reliability of variables Variable Number of items Cronbach's alpha Original Cronbach’s alpha Stays in analysis Innovation 3 .481** .63 - Independence 2 .558* .58 + Recognition 3 .730 .60 + Roles 2 .526** .73 - FinSuccess 3 .813 .76 + SelfReal 4 .469** .78 - Searching 3 .496** .84 - Planning 4 .459** .84 - Marshalling 2 .66* .80 + ImplPeople 6 .783 .91 + ImplFinancial 3 .841 .84 +
* Δ ≤ .10, therefore they not excluded from analysis ** Δ > .10, therefore excluded from analysis
4.3 CORRELATIONS
After removing scales which does not have a sufficient level of reliability, the correlation matrix is assembled including means and standard deviations. In case of Gender, which is a dummy variable, there is no sense to calculate mean and standard deviation, also reliability of scales is not included as well as for growth, as for a categorical variable a reliability score is not interpreted.
Interestingly, the highest and a really strong significant correlation (.600) is found between FinSuccess and Recognition. Despite the number of significant correlations between the mediators, the conceptual model does not examine these relationships. However, many correlations are found between the examined variables as well. In the dataset, the independent variable, Gender correlates significantly with Independence and with ImplFinancial both with a medium effect. The dependent variable, Growth correlates significantly with Recognition but with a small effect, while Marshalling and Independence it has a medium effect. These results are casting important impressions for the hypothesis testing.
Table 7. - Correlation matrix Variable M SD 1 2 3 4 5 6 7 8 1. Gender - - - 2. Independence 3.96 0.79 .302** (.558) 3. Recognition 3.13 0.87 -.071 -.017 (.730) 4. FinSuccess 3.54 0.88 .001 -.059 .600** (.813) 5. Marshalling 3.69 0.69 -.190 -.122 .390** .198* (.66) 6. ImplPeople 3.49 0.63 .095 .185 .165 .063 .256* (.783) 7. ImplFinancial 3.41 0.83 -.364** -.142 .316** .194 .450** .324** (.841) 8. Growth_ 1.90 0.68 -.114 -.349** .243* .144 .304** .197 .188 -
**Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).
4.4 HYPOTHESIS TESTING
Having carried out the preliminary steps to prepare and check the variables, simple mediation analysis is run on the mediational model (Figure 2.), which is done by installing the PROCESS macro into IBM SPSS Statistics, developed by Andrew F. Hayes. It gives several tools to test 2 and 3 way interactions and for the mediation models it creates bootstrap confidence intervals of the conditional and unconditional effects which are bias corrected (Hayes, 2012). Model 4 is used, which enables to run all the mediation within one test (up to ten mediators) which does not affect each other, even though the independent variable is a categorical, dummy variable. Using this method, the indirect (a+b), the direct (c’) and the total (c) effects can be tested. The analysis is run in one sum, but presented in three sections in order to be able to report the results
X M Y 0: Male 1: Female 1: Micro 2: Small 3: Medium Independence Recognition FinSuccess Marshalling ImplPeople ImplFinancial a b c’
more transparently. First, self-efficacy is introduced, then motivation to start a business follows and eventually a summary on the effects is presented.
It is important to note, that the analysis is not carried out by the causal approach, rather it is run by a tool, as according to the author of PROCESS, “modern thinking about mediation analysis
does not require evidence of a total effect prior to the estimation of direct and indirect effects”
(Hayes, 2012 p. 24) contradicting the causal method of Baron and Kenny (1986), which suggest testing the total effect and having it as a gatekeeper for further testing the mediation.
4.4.1 Mediator: Self-efficacy
In the case of self-efficacy variables (Marshalling, ImplPeople, ImplFinancial), the only significant effect is found between Gender and ImplFinancials. This means that females estimated to differ by a6 = -.611 units on ImplFinancials (p<.001). The sign of a6 is negative, meaning that women estimated to be lower on their self-efficacy regarding implementing financials. This effect is statistically different from zero t=-3.85, p=.000, with a 95% confidence interval from -.926 to -.296. Despite the significance of the a6 effect, the b6 effect is not significant (p>.001), meaning that the indirect effect, therefore the mediation of self-efficacy is not significant in any cases of the variables. By this result, Hypothesis 2 must be rejected.
Table 8. - Regression analysis on self-efficacy variables
Consequent
Antecedent Marshalling (m4) ImplPeople (m5) ImplFinancials (m6) Growth_ (Y)
Coeff SE p Coeff SE p Coeff SE p Coeff SE p
Gender (X) a4 -.263 .138 >.001 a5 .120 .128 >.001 a6 -.611 .159 <.001 c1' .001 .141 >.001 Marshalling (m4) - - - b4 .168 .106 >.001 ImplPeople (m5) - - - b5 .233 .110 >.001 ImplFinancials (m6) - - - b6 -.049 .093 >.001 constant i1 3.798 .090 <.001 3.436 .083 <.001 3.667 .103 <.001 i2 1.536 .565 >.001 R²=.036 R²=.009 R²=.133 R²=.251 F(1,97)= 3,6227 F(1,97)= ,8814 F(1,97)= 14,8241 F(7,91)=4,3520 p>.001 p>.001 p<.001 p<.001
4.4.2 Mediator: Motivation to start a business
In the case of motivation to start a business variables (Independence, Recognition, FinSuccess), the effect of Gender on Independence is really close to be significant (p=.0023), meaning that females estimated to differ by a1=.484 units on Independence. The sign of the a1 is positive, meaning that women are estimated to be higher on the motivation to start a business regarding Independence, than males are. This effect is statistically different from zero t=3.126, p=.002, with a 95% confidence interval from .177 to .791.
Interestingly, the effect of Independence on Growth_ is also found to be significant (p<.001). Because of b1=-.32, this result can be interpreted that those who are higher on Independence are estimated to be .32 units lower on Growth_. This effect is statistically different from zero t=-3.832, p=.000, with a 95% confidence interval from -.485 to -.154.
