The affect of social capital on the success of crowdfunding project

MSc. Business Administration – Entrepreneurship & Innovation

THE EFFECT OF SOCIAL CAPITAL ON THE SUCCESS

OF CROWDFUNDING PROJECT

Course: Master Thesis (6314M0254) Supervisor: Dr. T. Vinig

Author: Gabriela Suruceanu (10296808) Final Draft: August 21st, 2015

Statement of Originality

This document is written by Student Gabriela Suruceanu 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.

ABSTRACT

Crowdfunding has become a popular method of financing creative entrepreneurial ventures by tapping into a crowd of investors through the use of online platforms. Accordingly, the role of social capital is of relevant importance. Crowdfunding literature is in its nascent state and the existing studies are mainly focused on general dynamics of successful projects. In addition, most of these studies reflect characteristics of U.S. based projects. Hence, to compensate for the literature gap, this paper investigates the impact of social capital on the projects’ success in reward-based model by focusing on non-U.S. projects. The impact of social capital is assessed in terms of internal and external social ties, as well as entrepreneur’s reputation. In addition, intensity of social media is used to determine its influence on the outcome, potentially explained by number of backers which serves as a moderator. The research draws on a dataset of 322 successful projects on Kickstarter that ended by the time the data had been collected. The empirical results suggest that external social capital does not influence the success of the project; a finding in contradiction with published works. In addition, the findings highlight that the role of internal social capital ends once the project reaches its goal. However, entrepreneur’s reputation continues to exert its power even after the successfully achievement of its goal leading to an overfunded project. Lastly, the role of social media intensity is significant. However, the moderator does not explain this relationship. Overall, the present findings indicate that the role of social capital in non-U.S. crowdfunding initiatives is significant to the extent of achieving the goal, but that other factors such as the entrepreneur’s reputation is by far more important in surpassing this goal.

Table of Contents

4.3.2!Profile!Page!...!33! 4.4#Measures#...#34! 4.5#Validity#and#Reliability#...#36! 5.#RESULTS#...#37! 5.1#Descriptive#Statistics#...#37! 5.2#Preliminary#Data#Analysis#...#39! 5.3#Correlation#Analysis#...#41! 5.4#Assumptions#...#43! 5.5#Hierarchical#Regression#Analysis#...#45! 6.#DISCUSSION#...#48! 7.#CONCLUSION#...#54! REFERENCES#...#57! APPENDICES#...#61! Appendix#1:#Project#Page#...#61! Appendix#2:#Profile#Page#...#62! Appendix#3:#Overall#summary#of#missing#values#...#63! Appendix#4:#Descriptive#statistics#(before#the#transformation)#...#64! Appendix#5:#Descriptive#statistics#(after#the#transformation)#...#64! Appendix#6:#Histograms#and#Q[Q#Plots#...#65! Appendix#7:#Multicollinearity#statistics#...#67! Appendix#8:#Normality#residuals#and#P[P#plots#...#68!

1. INTRODUCTION

"The new economy is not just about the exchange of information, it's about the exchange of relationships. To break through the noise, to get your message out, count on personal networks… Relationships are the most powerful form of media… To build trust, invest in your

relationships constantly.”

Pam Alexander, CEO Alexander Ogilvy Public Relations Worldwide (Fast Company Magazine, March 2001)

This quote is in a perfect alignment with the topic of the current research, the importance of social capital for entrepreneurial success. It embodies the fact that in order to be successful it is crucial to develop and maintain a personal network in order to activate the crowd. It is no longer unusual for people to develop network relationships within the digital space. This is facilitated by the use on a large scale of social networking sites such as Facebook, Twitter, LinkedIn, Blogs, etc. In particular entrepreneurs rely on their social capital on their path towards success. In this context, social capital is accrued through direct and indirect social ties.

One of the main challenges faced by entrepreneurs is lack of capital. This issue is prominent especially for start-ups that face difficulties in attracting external financing during the seed-stage, either through bank loans or equity capital (Belleflamme et al., 2014). As a result, creative founders established a new evolutionary financing method – crowdfunding. Crowdfunding became a valuable alternative source of funding for those entrepreneurs that are seeking external financing (Belleflamme et al., 2014). In addition, crowdfunding phenomenon has been boosted by technological developments such as Web 2.0 which offers new

opportunities for consumers to use, create and modify content, and interact with other users through social networks (Ordanini et al., 2011).

As a result, crowdfunding has been going viral around the world (Thorpe, 2014) and it became a transformation tool for anyone aiming to raise money to start a business (Johnson, 2014). According to the Crowdfunding Industry Report, global crowdfunding experienced accelerated growth in 2014, expanding by 167% and raising $16.2 billions (Crowdsourcing.org, 2015). The report predicts a double growth of the industry in 2015, which is an estimated $34.4 billion raised in initiatives. Hence, the success of crowdfunding is explained by the fact that entrepreneurs seeking external seed financing are “tapping the crowd” instead of specialized investors by raising investment money through the use of online social networks (Belleflamme et al., 2014).

Entrepreneurs seek funding from the crowd by using specialized online platforms. Those that find the initiative interesting can back the project. The types of crowdfunding models differ based on what backers receive in exchange for their support (Colombo et al., 2015). Accordingly, these models are: donation, lending, reward-based, and equity (Mollick, 2014). The scope of this research is to investigate the reward-based model which entails that entrepreneurs recompense backers for pledging money to support their projects with rewards such as an actual product, limited editions, or collaborations and experiences (Kickstarter.com, 2015). Since crowdfunding provides a platform to project owners to engage in entrepreneurial activities with a community of investors, this paper will refer to project creators as “entrepreneurs” and people that fund the projects “investors” (Agrawal et al., 2011).

Initially, crowdfunding was developed in creative industries and had been dominated by the music-only platform SellaBand which was founded in 2006 (Agrawal et al., 2013). With the

establishment of Kickstarter in United States (U.S.) in 2009, this reward-based platform, which operates globally, became the largest existing crowdfunding provider in terms of money raised and projects financed (Colombo et al., 2015).

Crowdfunding projects can greatly range in both goal and magnitude, spanning from small projects to entrepreneurs seeking large amounts in seed capital (Mollick, 2014). One such example is the Pebble Time smartwatch funded on March 2015, raising $20.3 million from 78,471 backers having an initial goal of $500,000 (Zipkin, 2015). This campaign broke the record in raising its first million in less than an hour and it also became the most funded project ever on Kickstarter (King, 2015). This example illustrates the magnitude of the project itself and the important participation of the crowd in the path towards success.

Despite the escalating popularity of crowdfunding phenomenon, the underlying literature is still in its infancy. The existing studies are often working papers rather then peer reviewed articles. However, their contribution is significant to this paper since it facilitates the understanding of this fairly new phenomenon. The dynamics of crowdfunding have been investigated in order to determine the importance of the geographic distance between the project creator and the investor (Agrawal et al., 2011). The importance of quality project, social networks and social capital was highlighted (Mollick, 2014; Kuppuswamy & Bayus, 2014; Colombo et al., 2015). Furthermore, the role of family and friends in the seed-stage of the venture was analyzed (Ordanini et al., 2011). And the role of information asymmetry within crowdfunding environment determined (Agrawal et al., 2011,2013; Colombo et al., 2015).

The existent literature points to the important role of social networks within online crowdfunding environment which needs to be mobilized in order to increase the chances of success. One of the most important contributions in this direction is the study conducted by

Colombo et al. (2015) which focuses entirely on the role of social capital in reward-based crowdfunding model and hence, is used as the main reference in this paper.

However, the existent studies dedicated to reward-based crowdfunding model are focused on the analysis of crowdfunding patterns of U.S. based projects. In his study, Mollick (2014), excluded from his research all “foreign projects” since these might by atypical when compared to projects inside United States. Whereas, Colombo et al. (2015) point that the results of their study should be approached with caution since crowdfunding communities might be culturally mediated.

Consequently, it becomes evident that currently the dynamics of non-U.S. crowdfunding initiatives are unexplored. The need to fill in this literature gap is especially accentuated in a global online environment where traditional borders seem to disappear and support and social connections are established with a click of a button. Accordingly, the scope of this paper is to analyze the role of social capital in online reward-based crowdfunding environment that includes non-U.S. initiatives. To achieve the proposed, this paper will partially replicate some of the existing studies with a special attention to the study conducted by Colombo et al. (2015). Therefore, the main research question that will be answered in this paper is:

To what extent does social capital impact the success of crowdfunding project?

The aim of this study is to test empirically the effect of social capital on the success of crowdfunding project. In line with the reviewed literature, this study proposes that social capital is significantly associated with the project success. This explanatory study will test this

relationship by employing social media research strategy and conducting a subsequent quantitative analysis.

This study can contribute to the existing literature on crowdfunding by providing with new insights into the role of social capital on the success of crowdfunding initiatives. Particularly, it can shed some light on the success factors for non-U.S. initiatives in their race to achieve and surpass the proposed goal. Furthermore, this study can provide fundamental knowledge for entrepreneurs that intend to use crowdfunding as means of fundraising by suggesting methods to develop, maintain and expand their potential important social capital. In addition, this paper can provide useful knowledge to crowdfunding platform managers that want to attract and foster successful projects, by highlighting the fundamentals of social interactions on their platforms needed for their users.

The paper will continue with literature review in which existing literature on crowdfunding, entrepreneurial finance, business and marketing will be presented in order to determine the current knowledge in the explored fields and to define concepts relevant to the study. Next, the conceptual model is provided along with relevant hypotheses. The relationship between relevant concepts is empirically tested in the following chapter by means of social media research. Next, the research method and data collection technique are explained in detail. In addition, key terms found on Kickstarter website are explained and variable measurements used for testing the conceptual model are presented. The results chapter will present relevant statistics and the underlying findings. These results are discussed in detail in the following chapter by assessing the impact of social capital on the success of crowdfunding projects. In addition, the chapter provides academic and practical implications for entrepreneurs, highlights

the limitations of the study and recommends directions for future research. The paper ends with overall conclusions wherein the research question of the study is answered.

2. LITERATURE REVIEW

This chapter begins with an overview of crowdfunding including a general definition. Next, relevant crowdfunding literature is presented which highlights the characteristics of investors and the dynamics of successful crowdfunding projects. In addition, the roles of community and social capital are emphasized as well. Due to the nascent nature of crowdfunding literature, other academic sources focused on social network, social capital and social network sites are reviewed. Each section elaborates on appropriate aspects embedded in the study.

2.1 Defining Crowdfunding

Crowdfunding is rooted in the concept of crowdsourcing, which emphasizes using the crowd to obtain ideas, feedback and solutions in order to develop corporate activities (Belleflamme et al., 2014). The development of Web 2.0 led to the expansion of online crowdsourcing which allowed users to generate content, to communicate and interact with a wide network of people (Macht & Weatherston, 2015). However, none of these activities refer to fundraising (Macht & Weatherston, 2015), which is the main distinction from crowdfunding.

In an entrepreneurial context, crowdfunding is defined by Mollick (2014) as “the efforts

by entrepreneurial individuals and groups – cultural, social, and for-profit – to fund their ventures by drawing on relatively small contributions from relatively large number of individuals

its own unique category of fundraising, facilitated by the development of Web 2.0 (Ordanini et al., 2011; Macht & Weatherston, 2015) and a growing number websites specialized in this field (Mollick, 2014).

2.2 Crowdfunding Literature

Crowdfunding is a relatively new phenomenon and hence, the academic literature in this field is only nascent (Belleflamme et al., 2014). Consequently, there is a lack of peer-reviewed works on this topic (Mollick, 2014). Despite the fact that much of published literature are working papers; these offer valuable contributions to this subject (Mollick, 2014). Nevertheless, the existent literature adopts perspectives of entrepreneurial finance, social capital and social networks.

In the following sections, significant academic works dealt with crowdfunding are discussed.

2.2.1 Investors

Ordanini et al. (2011) conduct a qualitative study on three different types of crowdfunding platforms: SellaBand (i.e., music-only), Trampoline (i.e., financial services) and Kapipal (i.e., non-profit services). The goal of the study is to determine how and why customers turn into crowdfunding participants and the service providers set up a crowdfunding initiatives. The authors highlight that one of the unique characteristics of crowdfunding is that customers use their evaluative decisions and provide monetary support, and hence, are key players in this process by making the generation of the offering possible. The findings suggest that the main drivers for customers to become investors vary depending on the type of the platform used.

Accordingly, these are project involvement, social participation, and monetary payoff. Nevertheless, all investors share a common purpose – they like engaging in innovative behavior (Ordanini et al., 2011). Besides the underlying interest in the content of the crowdfunding initiative, investors are attracted by the novel way to use the platform especially in the context of social networking (Ordanini et al., 2011).

Further more, Ordanini and colleagues (2011) determined that the investment process of crowdfunding initiative consists of three phases: 1) rapid accumulation of funds at the initial stage due to the support of family and friends, 2) slowdown phase where ‘getting the crowd’ through word-of-mouth communications is crucial, and, if successful, leads to the final stage 3) the “engagement moment” during which those interested and with no prior connection to the project, decide to invest.

The findings determined by Ordanini et al. (2011) emphasize the role of social networking in crowdfunding, which is relevant for this paper. Nevertheless, these results suffer from limitations due to its qualitative nature. A further quantitative analysis is needed in order to investigate drivers of success of crowdfunding initiatives (Ordanini et al., 2011).

2.2.2 Dynamics

One of the characteristics of crowdfunding is that investors are geographically dispersed (Agrawal et al., 2013). In order to test whether empirical evidence that investors in early-stage entrepreneurial ventures tend to be local applies to crowdfunding as well, Agrawal et al. (2011) conduct a study in order to assess the role of geographic distance between the entrepreneur and investors on early stage funding of entrepreneurial projects. The quantitative study on SellaBand

result of the study emphasizes the role of family and friends in the early-stage funding in entrepreneurial venture, which serves as a signal of entrepreneurial commitment for later investors (Agrawal et al., 2011). This result is in line with the investment process established by Ordanini et al. (2011).

According to Agrawal et al. (2011), the role of geographic distance fades over time since distant investors rely on the investment decisions revealed by others (i.e., family and friends) at an early-stage of funding. This is explained by the fact that information asymmetry between family and friends and other investors decreases after the initial funding phase since other investors are able to monitor the entrepreneur’s progress directly on the crowdfunding platform (Agrawal et al., 2013). These results suggest that despite the fact that the crowdfunding platform eliminates most distance related frictions (i.e., acquiring information, monitoring progress), it does not eliminate social-related frictions associated with the type of information about the entrepreneur held by family and friends (Agrawal et al., 2011). The role of information asymmetry in seed-stage financing is also emphasized in the study conducted by Shane and Cable (2002), and will be discussed later on.

Mollick (2014) expands the study conducted by Agrawal et al. (2011) by investigating not only the role of geography but also other crowdfunding dynamics that might lead to the success of a project. The empirical study focuses on the reward-based crowdfunding platform – Kickstarter. It analyzes 48,526 U.S. based projects in order to depict the dynamics of crowdfunding across a wide variety of projects from the perspective of entrepreneurship (i.e., project creators). The results of the study indicate that social capital, measured as Facebook friends of project creators, and quality signals (i.e., preparedness) are associated with increased chance of project success (Mollick, 2014). Particularly, the research determined that the size of

social capital plays an important role, such as large networks are associated with project success and that having no Facebook account is better than having a few Facebook friends. The study could not determine the exact role of geography on the success of crowdfunding efforts besides that it has strong geographic component to the nature of the projects; hence, further research is required (Mollick, 2014). The results of the study are relevant for this paper since it highlights the importance of social network ties in crowdfunding, which is in line with the study conducted by Agrawal et al. (2011). Nevertheless, the limitation of the study conducted by Mollick (2014) is the exclusion of foreign projects from the sample due to the atypical character such projects might have.

Kuppuswamy and Bayus (2014) study the dynamics of project backers over the funding cycle by analyzing data from Kickstarter. The authors determined that other investor’s funding decision is a key factor in the success of a project. It appears that project creators on Kickstarter attract most of their funding at early and late stages of the project, by activating both their personal social network (i.e., family and friends), and followers who directly know the creator from social media connections (Kuppuswamy & Bayus, 2014). The importance of family at early stage of funding is in line with the finding emerged from the study conducted by Ordanini et al. (2011). However it extends the finding by highlighting the importance of family at a late stage as well. In addition, according to Kuppuswamy and Bayus (2014), backers are not influenced by the herding behavior of other backers in their decisions to support a project. But rather, backers are influenced by the estimation of how much of the goal has already been pledged.

These findings emerged from Kuppuswamy and Bayus (2014) study are relevant to this paper since the role of creator’s personal network composed of family, friends and followers is highlighted as being important in reward-based crowdfunding settings. The limitation of these

findings is that the research was done only for U.S. based projects.

2.2.3 The Community

Belleflamme et al. (2014) conduct a study in order to determine the entrepreneur’s choice between the two types of crowdfunding models: pre-ordering and profit sharing. The main finding of the study is that regardless of the chosen crowdfunding model, in order for crowdfunding initiatives to succeed, it is crucial for entrepreneurs to build a community that supports their efforts. The authors specify that the process of building a crowdfunding community requires integrating online social networks into the managerial process as a mean to interact with the crowd. From the entrepreneur’s point of view, these ties are formed with the crowd for the strategic purpose of raising money (Belleflamme et al., 2014). From the crowd’s perspective, these ties are formed as a result of new social interactions that motivate the crowd to participate in the initiative due to feelings of connectedness to a community with similar ideals and interests (Belleflamme et al., 2014). The importance of community participation for investors is also highlighted in the preliminary exploratory research conducted by Agrawal et al. (2013) and in the earlier mentioned study conducted by Ordanini et al. (2011).

2.2.4 Social Capital

Colombo et al. (2015) investigate the role of social capital in crowdfunding community in seed-financing stage. The quantitative study analyzes 669 U.S. based Kickstarter projects in four categories: design, technology, film and video, and video games. The study distinguishes between internal and external social capital. According to Colombo et al. (2015), internal social capital relates to social contacts developed from within the crowdfunding platform, whereas

external social capital is the one created outside crowdfunding platform (i.e., family and friends, Facebook friends). The internal social capital emerges as a result of online interactions between project creators and backers and is represented by the number of Kickstarter projects than the entrepreneur had backed at the time of launching its own campaign (Colombo et al., 2015). This information is publicly accessible on the project creator’s profile page.

Further more, Colombo et al. (2015) determined that in reward-based crowdfunding model the role of information asymmetry is reduced as a result of contributions during the seed-stage of the campaign. Initially, both the financial support received from early backers and the number of backers is publicly visible on Kickstarter, which leads to “observational learning” for other potential backers. This in turn, generates “word-of-mouth” around the project since early backers talk about their experience with their friends, which in turn might become backers themselves. This is facilitated by the presence of “Facebook Share” and “Tweet” buttons accessible from the project page (Colombo et al., 2015). Finally, early backers offer suggestions and feedback to project creators on a continuous basis which can be used to improve the underlying product or prevent a product failure once it reaches the market. The decreasing role of information asymmetry is in line with the findings of Agrawal et al. (2011).

The main finding of this study is that internal social capital has a stronger impact on early contributions compared to external social capital. This indicates that social contacts developed within crowdfunding communities may be a vehicle to attract seed financing and requires further investigation (Colombo et al., 2015). These results are particularly relevant to this paper since the role of social capital accrued both within and outside crowdfunding communities is stressed. The limitation of this study is that it focuses on four categories and that the analysis includes U.S. projects. Nevertheless, the results emerged from the study conducted by Colombo et al.

(2015) will be used as the main reference for this paper due to the detailed research conducted on the role of social capital which is in line with this paper.

The above-presented academic works highlight the community belonging aspect of crowdfunding. The underlying community is built by exploiting the emerged online social networks and subsequent social capital by tapping into the right crowd of investors. This in turn stresses the need to further elaborate on social networks and social networking sites. Since crowdfunding literature is scarce, additional sources in the field of entrepreneurial finance and marketing are employed and discussed in the following sections.

2.3 Social Network

The role of social network on the diffusion of new information is assessed in the study conducted by Granovetter (1973). The author differentiates between weak and strong ties that form individual’s social network and reflect both the quality and the strength of these relationships. In this context, strong ties relate to close friends of an individual, whereas weak ties relate to acquaintances (Granovetter, 1973). The study emphasizes the strength of weak ties in the diffusion of more novel information. This is explained by the fact that close friends share same information since they move in the same circles, whereas acquaintances have in their network people from different clusters (Granovetter, 1973).

In another research conducted by Granovetter (2005), the effect of strong ties is highlighted once more in the perspective that these ties are of high importance in transmitting unique information across largely disconnected segments of social networks. However, it is not

the quality of a particular tie that is important but rather the way different parts of network are bridged (Granovetter, 2005).

Shane and Cable (2002) investigate the role of social ties in seed-stage investments in order to determine the influence of social relationships in overcoming information asymmetry characteristic for seed-stage venture financing. The information asymmetry between an entrepreneur and potential investor arises since the former is reluctant to disclose all the information about the product, or the opportunity, whereas the latter lacks the necessary information in order to make an investment decision. According to the study, the reason why venture finance market does not fail is due to the role of social relationships among the parties. Shane and Cable (2002) test the effects of direct and indirect ties across investor-entrepreneur relationship by conducting both qualitative and quantitative study that incorporates U.S. investments. The difference between direct and indirect relationships is that the latter refers to the situation when the entrepreneur and the potential investor are not directly connected, but have someone in their network through whom the connection can be made (Shane & Cable, 2002).

The results suggest that while direct ties may encourage investments, the information received from indirect ties is more influential. However, overall, social relationships influence investor’s decisions, an effect mediated by entrepreneur’s reputation. The role of reputation in this study indicates that investors exploit their social ties in order to gather private information on entrepreneur, which once publicly available, mitigates the role of social relationships (Shane & Cable, 2002). The results of this study are relevant to the current research since it stresses the role of social ties which serves as a mechanism for obtaining information about the entrepreneur. In addition, it emphasizes the role of publicly available information about entrepreneur’s past

experience. Nevertheless, the limitation of the study conducted by Shane and Cable (2002) is the focus on U.S. investments and hence, need to be further explored in different economic settings.

2.4 Social Network Sites

Increasingly, people are using content sharing sites, blogs and engage in social networking with the scope of creating, modifying, sharing and discussing internet content (Kietzmann et al., 2011). Social network sites (SNSs) such as Facebook or LinkedIn allow individuals to shape their social networks and establish or maintain connections with others (Ellison et al., 2007).

As a consequence, the emerged use of social media has changed the communication practices not only between individuals but also between organizations and individuals. From marketing point of view, social media is often used as a promotion tool through which consumers communicate with each other in real-time (Kietzmann et al., 2011) by means of SNSs, thus, activating word-of mouth mechanism (Mangold & Faulds, 2009). In addition, a consumer can turn into “influencer” in its network if the portfolio of its relationships is dense and large (Kietzmann et al., 2011). Further more, consumers like networking with people who have similar interest and desires as their own (Mangold & Faulds, 2009). SNSs facilitates this interaction, and hence, reaching hundreds and even more similar types of people dispersed globally becomes increasingly easy. In addition, the reputation of a user in social media environment can be measured as the amount of followers on Twitter, or the number of ‘Facebook Likes’ (Kietzmann et al., 2011).

The role of Facebook on the formation and maintenance of social capital has been assessed by Ellison and colleagues (2007). The study is conducted when Facebook have been

used exclusively by college students, which is the early days of its existence. The main finding of the study is that internet-based linkages facilitate the creation of weak ties in which Facebook plays an important role wherein individuals form and maintain social capital. This finding is relevant to this paper since it emphasizes the role of Facebook on the creation of weak ties. Intuitively, this finding should be treated with caution taking into account the evolvement of Facebook from college to global social network site. Accordingly, weak ties in this setting are strong ties in today’s use of Facebook. Nonetheless, reviewed crowdfunding literature points to the relevance of Facebook use within its community, and hence, is embedded in the further concepts.

3. CONCEPTUAL MODEL

The conceptual model tested in this paper mainly incorporates the information accumulated from the reviewed academic works. In addition, due to the nascent nature of the analyzed phenomenon, an extra source of information is used in the development of the conceptual model – a video presented by B and H (2013, April 24) in which Stephanie Pereira,

the art program director of Kickstarter, informs the audience on how to bring a project to life. The role of social capital within online crowdfunding environment has been stressed by academics. It becomes evident that this particular type of fundraising revolves around the principle of being part of a community in which social networks and social interactions play important roles. However, the source of these interactions can be different. In order to test the effect of each type of social capital, it is important to differentiate amongst these. In line with the study conducted by Colombo et al. (2015), the conceptual model distinguishes amongst external

and internal social capital.

Additionally, the proposed model encompasses the potential role of entrepreneur’s previous experience on Kickstarter, as well as the impact of social media intensity. The outcome of the conceptual model is the success rate of the project. The visual representation of the conceptual model is presented at the end of this chapter (figure 1).

3.1 External Social Capital

In crowdfunding environment social network plays an important role. In general terms, this network is measured as the number of Facebook friends (Mollick, 2014; Kuppuswamy & Bayus, 2014) and hence, represents the external social capital of an entrepreneur (Colombo et al., 2015). Furthermore, the role of family and friends is proved important not only at the early stage of the fundraising (Ordanini et al., 2011; Agrawal et al., 2011) but also at the late stage (Kuppuswamy & Bayus, 2014). Some entrepreneur’s decide to link their Facebook page to their Kickstarter project page, while others prefer to abstain. At this stage it is not clear what guides these decisions. According to Mollick (2014), the role of social capital increases with the size of Facebook network and it negatively influences the outcome when the amount of Facebook friends of the entrepreneur is small.

Consequently, it becomes interesting to investigate whether the size of entrepreneur’s external social capital boosts the successful outcome. This relationship is studied in general terms throughout the entire project time span. Hence, the following hypothesis is proposed:

H1: The larger the size of entrepreneur’s external social network, the higher is the success rate of the project.

3.2 Internal Social Capital

The role of crowdfunding community has been highlighted by literature. Both entrepreneurs and investors feel the need of being part of a community. Entrepreneurs engage in internal social interactions in order to attract the necessary funds (Belleflamme et al., 2014). Investors on the other hand, are attracted to this type of fundraising since it provides with innovative ways of using social networking (Ordanini et al., 2011). In addition, investors enjoy the idea of being part of a community that shares the same interests (Belleflamme et al., 2014). It can also be argued, that some investors might eventually turn into entrepreneurs themselves after experiencing the support of the community.

Accordingly, it is evident that the community of crowdfunding is built on a principle of establishing and maintaining new social ties. These ties emerge as a result of crowdfunding activities. Such ties are traditionally known as weak ties since are established among people initially not closely related to entrepreneur. Literature highlights the strength of weak ties as being of crucial importance for diffusing new information (Granovetter, 1973). Despite using different definition of weak ties, Shane and Cable (2002) suggest that indirect ties are more influential than direct ties in gathering the necessary information about an entrepreneur in the context of seed-stage venture financing. In the context of crowdfunding the role of weak/indirect ties in early stage of fundraising is assessed by Colombo et al. (2015). The authors stress that entrepreneurs develop social contacts within the community as a result of supporting projects of their peers. Accordingly, the internal social capital accumulated facilitates the fundraising at early days of the campaign.

With this in mind, it becomes interesting to determine the importance of the size of entrepreneur’s internal social capital on the success of a crowdfunding initiative by accounting

its total overall effect, and not just at the seed-stage. Hence, the following hypothesis is suggested:

H2: The larger the size of entrepreneur’s internal social network, the higher is the success rate of the project.

3.3 Reputation

Although the reputation of an entrepreneur within crowdfunding community is indirectly related to the aspect of social capital, it can be argued that its previous experience on the platform can lead to an increase in social capital, either external or internal. Also, reputation in online setting might entail different aspects compared to traditional venture financing environment. Traditionally, the reputation of an entrepreneur is assessed as how successful he or she has been in its previous ventures, which in turn determines whether the entrepreneur will receive the necessary funding (Shane & Cable, 2002). However, in online setting, the reputation can be measured as the amount of followers on Twitter, or the number of ‘Facebook Likes’ (Kietzmann et al., 2011).

Accordingly, it can be argued that in crowdfunding environment, previously established projects might lead to an increase in social capital. Nevertheless, it becomes interesting to determine in general terms whether the experience of an entrepreneur, counted as its reputation, impacts the outcome. Or is the outcome determined entirely by the presence of direct and indirect social ties? According to traditional views of seed-venture financing, the role of social ties in investment decisions is mitigated by the publicly available information on entrepreneur’s reputation (Shane & Cable, 2002). Consequently, determining the exact role of reputation for

fundraising process is of particular interest in crowdfunding setting where the information about entrepreneur’s experience is publicly available on the platform.

Furthermore, according to Stephanie Pereira, the art program director of Kickstarter, B&H (2013, April 24), entrepreneurs on Kickstarter might fail a project once and then return with a new project and successfully raise funds. She explains this by the fact that entrepreneurs learn from each project and apply the gained knowledge and experience in their following ventures. This suggests, that supporters might not base their investment decisions on whether the reputation of entrepreneur is positive or negative. Accordingly, this study will not differentiate between previous successful and unsuccessful established projects since the goal is to determine the overall impact of reputation on successful outcome. Hence, the following hypothesis is proposed:

H3: The reputation of the entrepreneur positively affects the success of the project.

3.4 Social Media Intensity

Being part of crowdfunding community entails communicating online with direct and indirect social ties. Entrepreneurs motivate their supporters to share their projects through the use of social media. Investors can communicate with other potential supporters by making use of Facebook Share button (Colombo et al., 2015) and hence initially activate their direct social ties. This in turn increases the chances of diffusing the information further towards the social network cluster of weak ties (Granovetter, 1973). Accordingly, the important bridge between weak and strong ties might emerge (Granovetter, 2005), leading to the diffusion of information throughout

second phase of fundraising determined by Ordanini et al. (2011), which emphasizes the importance of activating the crowd in order to increase the chances of success.

Social networking sites such as Facebook facilitate not only the creation and the maintenance of social ties (Ellison et al., 2007); it also activates the mechanism of word-of-mouth (Mangold & Faulds, 2009). This impact is magnified by the fact that people like networking with individuals who have similar interest (Mangold & Faulds, 2009). Consequently, increasing the likelihood of becoming a potential influencer in its own network (Kietzmann et al., 2011). Hence, the intensity of using Facebook share button might positively impact the project outcome.

It is argued that backers who want to share their activities with their direct social network use Facebook share button. In addition, since crowdfunding investors are not influenced by herding behavior but rather by the estimation of how much goal has been reached (Kuppuswamy & Bayus, 2014), it is expected that they would make use of the “share” button to accelerate the fundraising process. Accordingly, the strength of the relationship between social media intensity and the success of the project might dependent on the amount of backers/investors that make use of the “share” button. Hence the following hypotheses are proposed:

H4a: The social media intensity positively affects the success of the project.

H4b: The relationship between social media intensity and the success of the project is

moderated by the number of investors, so that this relationship is stronger for projects that have a larger number of backers.

Figure 1: Conceptual model – Impact of social capital on the project success

4. DATA AND METHOD

This research is entirely based on the data extracted from the crowdfunding platform Kickstarter.com. Accordingly, this chapter will focus on a brief description of Kickstarter and the officially published statistics. Next, data collection technique is presented. A short explanation of terminology found on the platform follows. Subsequently, variable measurements are determined.

External Social Capital

Internal Social Capital

Reputation

Social Media Intensity

4.1 The Platform

Kickstarter was launched in April 2009 with a mission to help bring creative projects to life (Kickstarter.com, 2015). This new way of funding projects has shortly gained its recognition and became the most well-known and leading reward-based crowdfunding website (Outlaw, 2013; Barnett, 2014). Kickstarter, which publishes overview statistics on their webpage, lists a total of 247,867 launched projects, out of which, 90,135 were successful, accounting for $1.61 billions in pledges (Kickstarter Statistics, August 5th 2015).

The reward-based crowdfunding model of Kickstarter is “All or Nothing” which means that if a project succeeds in reaching its funding goal within a chosen timeframe, the amount pledged is cashed in by the project creator. If otherwise the goal is not reached, the campaign is considered unsuccessful and all the pledges are voided and returned back to backers. For their intermediary services, Kickstarter charges a 5% fee of total collected funds and 3-5% in payment processing fees. For their support, backers receive rewards depending on the pledged amount. Furthermore, project creators keep 100% ownership of their work. Projects are classified in 15 categories: Art, Comics, Crafts, Dance, Design, Fashion, Film & Video, Food, Games, Journalism, Music, Photography, Publishing, Technology, and Theater. According to Kickstarter statistics up to August 5th_{, 2015, Film & Video is the category with the highest amount of} launched projects, however it is not the most successful category. Please see figure 2 below for detailed statistics presented in descending order based on the most launched category.

Figure 2: Kickstarter Statistics August 5th, 2015 Project Category Launched Projects Success Rate

All 247,867 36%

4.2 Data Collection

Data could be extracted only with regards to successful projects since the company’s policy does not permit generating a complete overview of both successful and unsuccessful projects through their search option. In order to discover unsuccessful campaigns, exact information on the title of a project or creator’s name is required. (Kickstarter.com, 2015). With regards to external search, Kickstarter de-indexes projects whose funding was unsuccessful

49,211 40,876 26,585 19,026 18,775 16,365 15,774 15,707 12,643 8,454 7,805 6,263 4,693 2,871 2,819 Film & Video

Music Publishing Art Games Technology Design Food Fashion Theater Photography Comics Crafts Journalism Dance 38% 52% 30% 42% 33% 21% 34% 27% 25% 61% 30% 50% 24% 23% 64%

conducted by Colombo et al. (2015), only successful projects in four categories: Film & Video, Technology, Games, and Design were looked into.

The underlying data was collected directly from Kickstarter website using a freely accessible data extraction tool “import.io”. Initially, projects were discovered using “advance search” option. By selecting successful projects within a relevant category, an overview of all relevant projects was generated regardless of their location. The difficulty in extracting the necessary information came from the fact that Kickstarter uses infinite scroll pagination. This means that in order to discover projects beyond the first page, it is necessary to click on “load more” button each time the bottom of the page is reached. On one hand, the extractor tool can be used only when exact links from which data should be extracted are provided. On the other hand, the infinite scroll format creates difficulties in generating these links since the link in the address bar does not change once more projects are loaded, and hence, the extractor does not recognize the newly generated page. To overcome this difficulty, Google Chrome tools were used in order to clearly determine the exact uniform resource locators (URLs). Furthermore, another encountered difficulty was that an infinite scroll format does not indicate the overall amount of pages hidden behind “load more” button. Using trial and error method, it was determined that only first 200 pages were possible to access which is less than the total amount of projects initially discovered. This can be explained by the fact that Kickstarter has changed their website version and hence, further links are connected to an older platform format.

Nevertheless, it is argued that this will not impact the underlying sample size. First, this study is cross-sectional in nature that requires taking a snapshot of the population at a certain time. Second, due to the dynamic nature of researched variables, the sample included only projects that ended by the time the sample was collected. Hence, the sample incorporates only

those projects that have ended on June 1st _{and on July 31}st_{. Accordingly, reflecting the maximum} campaign length allowed by Kickstarter (Kickstarter.com, 2015).

The extraction tool was trained to extract data from certain website content such as project page and creator’s page. The generated data was organized into spreadsheets in which data has been cleaned and organized for further analysis.

Data on a total of 15,154 successful projects was extracted. Due to computer automated data collection technique, 5.7% of data was lost due to recording double records or due to inconsistency in data provided on the website. It is believed that double records are a result of software updates or platform updates. From the remaining 14,284 projects, foreign-based projects were 3,052. These projects were narrowed down to 325 successful projects which will be included in the subsequent analysis in order to account for the desired time frame.

4.3 Terminology

The terminology used by Kickstarter on the platform reflects both project related and creator related terms. These are explained below in order to facilitate the understanding of adopted and will be discussed in this particular order.

4.3.1 Project Page

The project page is the place where the content of the project can be found containing all relevant details about the project itself. The visual representation of the project page layout can be found in appendix 1.

The funding goal is the amount of money the creator needs in order to complete the project (Kickstarter.com, 2015). Once someone likes the project and wants to support it financially, the person can pledge money. This entails that the amount pledged is the total amount raised from supporters. This amount can be more than the funding goal itself, and hence the project becomes overfunded (Kickstarter.com, 2015).

Backers

The supporters of a project who pledge money are called backers. Backers can pledge an amount equal to one of the funding tiers (i.e., pledge levels) determined by the creator itself. The funding tiers are accompanied with rewards, precisely, the higher the pledge the higher the reward.

Funding Period

The platform indicates the dates for when the project has been launched, when it has ended, and for how long it has lasted. Initially Kickstarter allowed creators to raise funds for as long as 90 days (Mollick, 2014), it now limits the duration to 60 days, while recommending 30 days or less (Kickstarter.com, 2015).

Facebook Share Button

Kickstarter provides with certain social media tools for those interested in sharing the project within their network. Anyone that finds the project worth of sharing can do so by clicking on Facebook Share, Tweeter, Tumblr, Pin it buttons; or it can embed it on its own website. Both backers and platform visitors can perform these actions. For the scope of this research, only Facebook Share intensity is relevant and hence, will be used accordingly. Kickstarter platform

records the amount of Facebook Shares done via Kickstarter. A project can be shared both during the project and after its culmination.

4.3.2 Profile Page

A separate page is designated to each creator on which personal information can be found such as: general information, past experience, whether or not the creator has linked its Facebook account. This type of information is of particular interest to this study. The visual representation of the profile page can be found in appendix 2.

Facebook Account

Creators can choose whether to connect their Facebook account or not. When Facebook account is linked to the Kickstarter project, the amount of Facebook friends of the creator is visible. Some project creators decide not to link their Facebook account and hence this type of information is missing.

Previously Created Projects

The information about creator’s past experience on Kickstarter is visible and it includes both successful and unsuccessful previously created projects. Some creators can be new to the platform and hence, the project can be their first, others can be more experienced and have more than one previously established project. For the scope of this paper, it is irrelevant whether or not these projects were successful.

The platform also provides information on how many Kickstarter projects the creator has previously supported. This information is of particular interest to this study.

4.4 Measures

This section will look into the variables used in this study accompanied by a brief description and indication of applied measurements.

Project Success

Project success is the dependent variable in this study (Success) and is defined as the achievement of the targeted goal within the designated timeframe. The adopted measurement is the ratio between the total amount pledged and the funding goal at the closure of the campaign. The ratio value of 1 indicates a successful project that reached its exact goal, whereas values above 1 point to overfunded projects.

External Social Capital

For those creators that have linked their Facebook account, the external social ties are measured by the amount of Facebook friends of founders (FbF) visible to everyone. The measurement method is adopted from the study conducted by Mollick (2014), who records FbF as of the time of data collection rather than at the time of project initiation. According to the author, FbF are for the project initiator only and are less likely to increase as the result of projects, since founders tend to separate their project from their personal pages. The amount of FbF provide with an insight into the size of the founder’s social network (Mollick, 2014). It is

important to mention that for those founders that did not connect their Facebook accounts, the amount of friends is not considered zero, but rather missing values.

Internal Social Capital

The internal social capital is measured as the amount of projects that the entrepreneur had previously backed registered at the time of data collection (PrBacked). This variable indicates the degree to which the creator has been supportive and has established social contacts within Kickstarter community (Colombo et al., 2015). It also emphasizes the community belonging aspect.

Reputation

In this study the reputation of the entrepreneur is measured by the amount of previously established Kickstarter projects regardless of their outcome (Reputation). The success factor in this aspect is not important since the study adopts the view that within the crowdfunding community, the important role of previous experience is the learning curve through which an entrepreneur has passed, as mentioned in the video by B and H (2013, April 24).

Social Media Intensity

To quantify the usage intensity of social media, the amount of Facebook shares

(FbShares) is used. The moderation effect between social media intensity and the amount of

backers (Backers) is measured as the product term between the two variables

Controls

Previous research in the field of crowdfunding has determined certain factors to be important in explaining the success of crowdfunding campaigns. Hence, in order to provide more accurate estimate of hypothesized variables some of these factors directly extracted from the platform are used as controls: project category (Category) and fundraising duration (Duration) (Mollick, 2014). Categories are dummy coded in order to perform the necessary multiple regression analysis. Hence, for each category the same pattern was applied. For example dummy equals 1 if project category is Design, 0 otherwise.

4.5 Validity and Reliability

In order to ensure the trustworthiness of the research findings, validity and reliability are considered. Validity relates to the assessment of whether the research instrument measures what is intended to measure (Field, 2009). Accordingly, the measurements used in this study are primarily derived from the existing academic works with the exception of the social media usage intensity; which is self-explicit in this context and hence should not impact the validity of results especially taking into account the nascent nature of crowdfunding literature. In addition, in order to prevent certain influences on the analyzed relationships, controls are used which are derived from literature as well. The large sample size allows generalizing the findings to other research settings.

Reliability relates to the ability of the measure to produce the same results under the same conditions (Field, 2009). Since the constructs used in this study are derived from prior academic studies, it is expected that once applied using the same research method and analysis, consistent results would be generated.

5. RESULTS

The objective of this chapter is to report the results derived from the conducted empirical analysis. Initially, descriptive statistics on all variables are presented. Following, data is preliminarily analyzed. Further on, in order to test the underlying hypotheses, the relationships between the dependent and independent variables are assessed. Consequently, the emerged results from the hierarchical regression analysis are presented.

5.1 Descriptive Statistics

The presented results include information about 322 projects. Foreign currencies were converted to USD using an average exchange rate at the time of the campaign. On average projects raised (Amount Pledged) $101,460 (SD = $517,255) and had an average goal (Goal) of $29,727 (SD= $118,666). The average success rate (Success) is 3 times (SD= 3). Among these projects, the largest pledged project raised $6,333,295 which is 3 times its goal of $2,000,000. In addition, on average, project owners have 633 Facebook friends (FbF) (SD=582). Project owners previously backed (PrBacked) on average 7 Kickstarter projects (SD=16) and established 2 projects on Kickstarter (Reputation) (SD=2). Projects were shared on Facebook (FbShares) 1662 times on average (SD=7292) with the highest frequency of 112,518 times. Furthermore, projects were backed by 1226 backers (Backers) (SD=6717) and had an average duration (Duration) of 32 days (SD=9). A summary with descriptive statistics can be found in table 1 below.

Table 1: Descriptive Statistics

Variable N Minimum Maximum Mean SD

Amount Pledged 322 56.36 6333295 101460.45 517254.925 Goal 322 22 2000000 29727.16 118666.019 Success 322 1 26 2.69 3.165 FbF 135 0 3782 632.50 581.517 PrBacked 322 0 165 6.76 15.579 Reputation 322 1 14 1.56 1.598 FbShares 322 0 112518 1662.19 7292.331 Backers 322 4 73206 1226.20 6717.070 Duration 322 6 60 31.53 8.843

In addition, on average the most successful projects were within Games and Design category surpassing its goal on average 3 times, followed by Technology. The lowest performer was Film & Video category in which projects on average just reach their goal. Detailed overview can be found in figure 3 below.

These results are in contradiction with the statistics published by Kickstarter presented earlier (figure 2) which denote that the most successful projects in these four categories are within Film & Video and least successful are in Technology. Surely, these results are just a rough benchmark since the statistics presented by Kickstarter span over longer period of time compared to data used in this analysis. In addition, official statistics reflect all projects regardless of their origin.

5.2 Preliminary Data Analysis

In order to conduct the proposed analysis, the data was initially checked for missing values, normality and outliers.

Missing Values

The results of missing value analysis indicate that the only missing values are related to the variable FbF. This is explained by the fact that not all project owners have linked their Facebook account to Kickstarter platform. From 322 observations, only 42% have complete values (see appendix 3). Since the scope of this research is to test the relationship between the external capital and the success of the project, these missing values cannot be excluded from the analysis since these do not reflect the fact that the underlying capital is missing, but rather that these external ties are not observed. In addition, eliminating the cases with missing data would considerably reduce the sample size, which might have a detrimental effect on the underlying analysis.

being a fairly new technique, MI is considered a reliable model in handling missing data since it has been shown to perform favorably by providing unbiased parameter estimates (Wayman, 2003).

The particular technique applied for MI is Markov Chain Monte Carlo (MCMC) which generates multiple imputations by simulating random draws. The amount of imputations used in this study is 5 since it is considered enough for this percentage of missing values (Schunk, 2008). Accordingly, SPSS presents the generated results individually for each 5 imputations, and for certain statistic tests it also provides pooled values. In those instances when pooled values are missing, relevant statistics are calculated manually as averages of 5 imputed data sets.

Normality

In order to determine whether data is normally distributed, both skewness values and Q-Q plots for each variable are used. According to Hair et al. (2010), skewness values outside the range of -1 to +1 indicate substantially skewed distribution. As results show, excluding the variable Duration and FbF, these values indicate that all of the remaining variables are substantially positively skewed (see appendix 4). In addition, Q-Q plots used to visually assess normality, confirm the same. In order to address these issues, variables have been transformed by applying natural logarithm (Hair et al., 2010). Since logarithm of zero is undefined and some variables might have values of zero, a constant of 1 was added to the equation in such cases which is common practice: Ln(x+1). Consequently, as both dependent and independent variables are transformed, the further tested relationships in the regression analysis are interpreted as elastic “the expected percentage change in the dependent variable when independent variable increases by some percentage” (Benoit, 2011).

After the transformation, the same line of testing was conducted again to depict improvements. Accordingly, the distribution has improved significantly (see appendix 5). Q-Q normality plots and histograms indicate that some variables are relatively non-normal (see appendix 6). However, this is not surprising since larger sample sizes tend to reduce the detrimental effects of non-normality, and hence, the effect is negligible (Hair et al., 2010). Accordingly, the non-normality distribution of these variables is disregarded.

Outliers

Outliers are observations with unique identifiable characteristics distinctly different from other observations, represented by unusually high or low values (Hair et al., 2010). Despite the fact that outliers might bias the results, these unique values might as well point to unique characteristics of the population which otherwise would not have been discovered once eliminated from the analysis (Hair et al., 2010). Hence, for the scope of this analysis, which deems to estimate the effect of social capital on the success outcome, outliers are retained at this stage due to their estimated legitimate character (Orr et al., 1991) as well as potential explanatory effect. In addition, natural logarithm transformations tend to reduce significantly extreme values (Field, 2009).

5.3 Correlation Analysis

As a preliminary estimation of the underlying conceptual model, correlation analysis was conducted. This serves as an initial rough check of the relationships relevant to the study. According to the correlation matrix it becomes evident that most of the variables are significantly

The results show that there is a positive correlation between success (Ln_Success) and the following variables: reputation (Ln_Reputation) (r= .279, p<.01); number of Facebook shares (Ln_FbShares) (r= .119, p<.05), and number of backers (Ln_Backers) (r= .457, p<.01). Further more, a strong correlation is present between number of previously backed projects and the following variables: reputation (Ln_Reputation) (r= .352, p<.01), number of Facebook shares (Ln_FbShares) (r= .153, p<.01), and number of backers (Ln_Backers) (r= .258, p<.01). Reputation is negatively correlated with amount of Facebook shares (Ln_FbShares) with Pearson correlation coefficient of r= -.198 at the significance value less than .01. Furthermore, there is a strong positive correlation between amount of Facebook shares (Ln_FbShares) and number of backers (Ln_Backers) (r= .590, p<.01).

With respect to controls, duration (Duration) is positive and significantly correlated with success (Ln_Success) (r= 136, p<.05), number of Facebook shares (Ln_FbShares) (r= 240, p<.01) and number of backers (Ln_Backers) (r= 224, p<.01); whilst negatively and significantly associated with reputation (Ln_Reputation) (r= -.174, p<.01). Further more, control variables for four categories are negatively associated amongst each other, whilst being significantly associated with some of the predictors. Conversely, among all variables, number of Facebook friends (FbF) is only associated with the control Games (r= -1.59, p<.05).

A detailed summary for means, standard deviations (SDs) and correlations for all variables can be found in table 2 below.

Table 2: Means, SDs, correlations Variables Mean SD 1 2 3 4 5 6 7 8 9 10 11 1.Ln_Success .64 .731 -2.FbF 715.7 522.636 .013 -3.Ln_PrBacked 1.15 1.222 .105 -.009 -4.Ln_Reputation .25 .51327 .279** -.017 .352** -5.Ln_FbShares 5.53 2.160 .119* .082 .153** -.198** -6.Ln_Backers 5.33 1.600 .457** .010 .258** .015 .590** -7.Duration 31.53 8.843 .136* .154 .028 -.174** .240** .224** -8.Design .25 .43 .083 .036 -.094 -.034 .042 -.040 .116* -9.Games .34 .47 .115* -.159* .280** .176** .072 .270** -.140* -.418** -10.Technology .31 .46 -.057 .057 -.180** -.069 -.084 -.152** .058 -.392** -.487** -11.Film&Video .09 .29 -.220** .116 -.031 -.126* -.045 -.139* -.038 -.186** -.231** -.217** -** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed).

5.4 Assumptions

Initially, the regression model was checked to ensure it satisfies the underlying assumptions of proper variable types, no multicollinearity, normality of residuals and linearity among variables (Hair et al., 2010; Field, 2009).

Variable Types

In order to run a regression analysis, all predictor variables must be quantitative or categorical with two categories, whereas the outcome variable (i.e., dependent) must be continuous. In this study, all variables used are in line with this requirement. Accordingly, the predictor variables are scale and the categorical variable is dummy coded, whilst the dependent variable is a scale/ratio type of variable.

Multicollinearity

The assumption of no multicollinearity was tested by screening values for tolerance, variance inflation factor (VIF) and correlations between dependent variables. Accordingly,