Making research alliances between academia and industry in the Triple Helix Model work : the effect of academics’ evaluations of perception of benefits and barriers on the likelihood to engage in university-industry col

(1)

Radboud Dam

0215856

“Making research alliances between academia and industry in the

Triple Helix Model work: The effect of academics’ evaluations of

perception of benefits and barriers on the likelihood to engage in

university-industry collaboration”

Amsterdam, January 2017

Master Thesis

(2)

Statement of Originality

This document is written by Radboud Dam 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.

2 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(3)

“Innovation is the specific instrument of entrepreneurship. The act that endows resources with a new capacity to create wealth.” - Peter Drucker (1985)

(4)

Preface

After much hard work, I would like to present you my masters’ thesis regarding the impact of individual decisions of academic scientists on the process of university-industry collaborations. It has been written to fulfill the graduation requirements of the Executive Programme in Management Studies (EPMS) at the Amsterdam Business School.

Due to some major developments in my personal life, for example I became father to a beautiful son named Lucas, it took me nearly a year to complete it. Therefore some acknowledgements are in order here. First of all, I would like to thank Marten Stienstra for his continuing support, patience, clear feedback, pointers in the right directions and maybe most of all, his flexibility to work with my ‘last-of-the-moment’ timeline. Second, my girlfriend Anne, who pushed and supported me along the way and who took and takes great care for our son; I could not have done this without you. I thank you both. I also wish to thank all of the respondents, without whose cooperation I would not have been able to conduct any data analysis and report it in this thesis.

I hope you enjoy reading my thesis.

Radboud Dam, Amsterdam, January 2017

(5)

Abstract

The Triple Helix model explains how the interaction between government-industry-university drives innovation. In this context the entrepreneurial university emerged with a focus on industry collaboration. However, much is still empirically unknown about university-industry collaborations (UIC), especially about the reasons why academics engage or not. This study enriches the UIC and Triple Helix literature by the examining how academics’ perceptions of the benefits (e.g. access to resources, increased funding) and costs (e.g. loss of academic freedom) impact the likelihood to engage in industry collaboration. Cross-sectional data was collected via an survey from researchers employed at two large universities in Amsterdam. It is argued that academics favor academic over commercial goals, thus incentives beneficial for research and teaching are more influential than monetary gains to increase industry collaboration. Moreover, barriers related to transactional costs hinder future collaboration the most, while prior UIC experience is an important indicator for future industry engagement. General implications for future UIC policies are discussed.

Keywords: university-industry collaborations (UIC), Triple Helix model, entrepreneurial

university, knowledge and technology exchange, cost-benefit analysis, public-private partnerships, higher education industry, strategic alliances, knowledge-based view, resource-based view, motives to collaboration

(7)

1. Introduction

1.1. The quest for innovation in knowledge-based economies

The Western World faces several grand economic challenges in the near future (WRR, 2013). Innovation solutions based on knowledge and technology-exchange are necessary for challenges such as productivity issues, meaning learning to do more with less, and scarcity issues such as the lack of human resources or raw materials. Moreover, as a result of the globalizing world, nations will have to learn how to manage the shocks that may arise from that process: products and processes will be subject to continuous adaption and disruptive innovation, making the ability to absorb and circulate knowledge crucial (WRR, 2013; p. 23). In these highly dynamic knowledge-based economies, firms battle over strategic knowledge resources for (sustained) competitive advantage and survival (e.g. Andrews, 1971; Ansoff, 1965; Barney 2001; Teece, Pisano & Shuen, 1997). As the ‘knowledge-based view’ of the firm (KBV; e.g. Grant, 1996) argues, the key is to integrate knowledge within and across firms. As Grant (1996) postulated, if a “company’s strategically most important resource is knowledge, and if knowledge resides in specialized form among individual organizational members, then the essence of organizational capability is the integration of individual’s specialized knowledge” (p. 375). These core notions of the knowledge-based view can be used to explain competitive behavior between firms, but also to explain the competitive dynamics of knowledge-intensive environments, such as knowledge-based economies, incorporating other stakeholders as well such as governmental (contractual) agencies and universities (Etzkowitz, 2003; 2008).

To describe and explain such macro competitive dynamics of innovation and knowledge-transfer, the framework of the ‘national innovation system’ (NSI) is generally

(8)

adopted (e.g. Etzkowitz & Leydesdorff, 1995; 2000; Etzkowitz, Webster, Gebhardt, & Terra, 2000) which assumes dynamic interactions and bilateral relations between the triade of government-industry-university institutions or spheres. Nations can be compared with each other based on the specific configuration of government-industry-university relations (Etzkowitz et al., 2000; Etzkowitz, 2003). Dependent upon the configuration, NSI’s are more or less successful in acquiring, sharing, exchanging and integrating knowledge (including technology, know-how and organizational capability) across and between institutions (Etzkowitz, 2003; Dyer & Nobeoka, 2000; Grant et al., 2004; Grant, 1996). NSI’s drive innovation and if institutional spheres are configured and aligned properly, they could provide a potential solution to the grand challenges of the future (Etzkowitz, 2003).

Currently, the most promising configuration for successful innovation is the (third) Triple Helix model (Leydesdorff & Etzkowitz, 1996; Etzkowitz, 2003; 2008). In this model, the institutional spheres are viewed as helixes, and a triple helix of university-industry-government relations is configured in such a way that knowledge is able to freely circulate, to be shared and absorbed within and between the spheres, fostering innovation and learning both on a regional and national level (e.g. Adler & Kwon, 2002; Kivleniece & Quelin, 2012; Etzkowitz, 2003). Furthermore, in the Triple Helix model each helix has bilateral relations with the other two helixes with specific roles assigned to each helix. The industrial sphere functions primarily as the production unit, while government serves as the contractual agent and policy maker, and the university functions as the key knowledge-generating institute, all in close interaction (Etkowitz, 2003). Moreover, in this model, the university plays an enhanced role in societal innovation through incubation, knowledge transfer and circulation (Etzkowitz & Leydesdorff, 1995; 2000).

(9)

This development marks a new role for universities and refers to a third mission, besides teaching and research, which is commonly referred to as the ‘entrepreneurial mission’ (D’Este & Perkmann, 2011). This mission entails new functions, for example as an incubator of knowledge- or technology-based firms or for other type of ‘knowledge-into-application’ spin-offs (Etzkowitz, 2003). The ‘entrepreneurial university’ in this context is not a ‘commercialized’ university, but an university that encompasses the conservation and passing on of knowledge, integrating teaching and research, as well as supporting innovation (Etzkowitz, 2003; p. 333). Furthermore, on a regional or national level, the ‘entrepreneurial university’ is an equal and influential partner in the Triple Helix model of university-industry-government relations as it valorizes knowledge into application (p. 295).

1.2. Research gap

One way the entrepreneurial university is driving innovation, is by encouraging (research) collaboration with industry (D’Este & Perkmann, 2011). Collaborations based on joint research between universities and industry, as one of the bilateral relationships in the Triple Helix model, in which both researchers and practitioners actively participate, are referred to as ‘university-industry collaboration’ or UIC (e.g. Bekkers & Bodas Freitas, 2008; Siegel, Waldman & Link, 2003). These collaborations between universities and industry usually take the form of joint research, contract research or consulting (Taktari, Salter & D’Ester, 2012; D’Este & Perkmann, 2011) and can exist on all levels, ranging from academic-practitioner collaborations to governmental funded industry-university consortia (e.g. Murphy, Perrot & Rivera-Santos, 2012; Perkmann & Walsh, 2009). They are most common in the field of biotech, (data-) engineering and natural sciences (Tartari & Breschi, 2012) and the outcomes can be manifold such as reports, products, patents or spin-offs and new ventures. UIC is

(10)

considered an established form of (research) alliance especially in countries like the United States, Japan, Singapore and increasingly in European Union countries (Ankrah & Al-Tabbaa, 2015).

Although the UIC literature encompasses a considerable body of research, mainly from the industry perspective (e.g. to acquire knowledge or patents; Bruneel, D’Este & Salter, 2010; Jelinek & Markham, 2007) and from the last decade (Ankrah & Al-Tabbaa, 2015), there is still much unknown about this form of collaboration. One gap is the lack of empirical and quantitative studies on these type of collaborations from the university perspective, particularly regarding the question of why some academics engage in UIC while others do not (Taktari & Breschi, 2010; Tartari et al., 2012; Killian, Schubert & Bjorn-Andersen, 2015). As one of the most defining features of the university context is autonomy (Aghion, Dewatripoint & Stein, 2008), academic scientists usually have decision rights on what projects they take on and have creative control over the process. Thus it would be interesting to gain insight in academics’ motives for UIC.

The importance of the academic scientists for the commercialization of university inventions has been stressed in some studies (e.g. Zucker & Darby, 1996), but generally there is little known about the processes through which individuals choose to collaborate (Taktari & Breschi, 2010). One implicit assumption in the literature regarding the entrepreneurial university is that academics engage in industry collaboration in order to commercialize their knowledge (D’Este & Perkmann, 2011). Therefore, policy-makers provide monetary incentives to academics to facilitate their commercial involvement (Link & Siegel, 2005). However, as Agion et al. (2008) postulated, given the nature of scientific research and the norms of the academic community, academics might rather pursue academic goals and ultimately engage to place scientific discoveries in the public domain (Merton, 1973). This

(11)

would imply a certain degree of dualism or conflict among scientists in regard to university-industry relations.

Tartari and Breschi (2010), among other researchers, assume therefore that academics’ decisions to engage in UIC activity follows standard cost-benefit analysis. Like D’Este and Perkmann (2011), they point to the importance of the formation of academics’ perceptions of the benefits (e.g. access to resources, increased funding) and costs (e.g. loss of academic freedom) associated with industry collaboration. Bruneel et al. (2010) argue that based on the evaluation of associated benefits and costs with UIC the likelihood or propensity to engage in UIC is either fostered or impeded. Thus, understanding how these perceptions are shaped is crucial to determine academics’ willingness to (actively) collaborate (p. 656). Willingness to participate in knowledge transfer activities is fundamental to university-industry collaboration (Tartari et al., 2012).

It remains unclear however, how the perception of benefits and barriers shape the cost-benefit decision process and impact the propensity to engage in industry collaboration (Taktari et al., 2012). The few studies to date that focused on academic’s cost-benefit analysis in relation to UIC (e.g. Lee, 2000; D’Este & Perkmann, 2010; Taktari & Breschi, 2012) studied primarily the positive side of collaboration (Taktari et. al, 2012) and failed to include the perception of barriers or the negative side of collaboration. A better empirical understanding of these micro-foundations of the entrepreneurial university, might help to explain why some academics engage in UIC and others do not. Moreover, by gaining insight in how academics’ evaluate the impact of cost-benefit perceptions, a deeper understanding of the nature and drivers of UIC is evoked. Ultimately these insights might increase the success ratio of new UIC policies at universities which benefits society and drive innovation in the triple helix of government-industry-university.

(12)

1.3. Research question and potential contributions

The focus of this study is therefore to explore to what extent the likelihood (or decision) to engage in UIC is explained by the individual evaluations of expected benefits and costs of collaboration. Thus, the central research question in this study is: “what is the effect of academics’ perceptions of benefits and costs of university-industry collaboration (UIC) on the likelihood to engage in UIC?”

This study potentially contributes to the extant literature in a number of ways. First, insight into the motivational foundations of academics’ choice to collaborate with industry might help to explain why some UIC activities or policies are successful and others not. As Rothaermel, Agung and Jiang (2007) state: “This debate (of university-industry engagement) would benefit greatly from a deeper understanding of the involvement of individual researchers” (p. 310). Thus it contributes to the debate of the entrepreneurial university by shedding light on its micro-foundations. Second, much of the conducted research on UIC is theoretical or conceptual of nature (Tartari et. al, 2012). As there is no clear theory about university-industry relationships, especially from the university perspective, there are no validated scales developed yet (Taktari & Breschi, 2012). By empirically validating some of the theoretical assumptions regarding industry collaboration from the university perspective, a modest methodological contribution to the UIC literature and a deeper understanding of UIC is evoked. This can foster further research. Third and final, examining the relations between perceived benefits and barriers and the propensity to engage in UIC simultaneously in one study, has not been done before or at least not been found in the literature.

(13)

1.4. Thesis overview

This thesis is structured as followed. First, the relevant literature on the dynamics of innovation and the knowledge-based view of the firm is reviewed. As it gives context to the notions of the Triple Helix model and the entrepreneurial university. Second, the relevant literature on UIC and academic cost-benefit analysis is discussed. In this section distinct type of benefits and barriers are identified and explored, and hypotheses articulated. Subsequently, a quantitative, cross sectional survey study is constructed to test the various hypotheses regarding the effects of perceived benefits and barriers on the propensity to engage in UIC, controlling for a number of control variables, and exploring a potential moderating factor. Results will be thoroughly analyzed before interpreted and placed in a larger context. Both scholarly and managerial implications will be discussed and directions for future research explored, before answering the main research question.

(14)

2. Theory and hypotheses

In this section the theoretical framework is constructed. To gain insight in the reasons why researchers engage in UIC, it is pivotal to first understand the context – both the organization and environment – they work in. Therefore the first section addresses the current market dynamics and pursuit of innovation in knowledge-based economies by explaining the Triple Helix model of innovation and how differentials in Triple Helix outcomes across nations or regions can be understood by adopting a ‘resource-based view’ (RBV) or ‘knowledge-based view’ (KBV) of the firm. Second, the concept of the entrepreneurial university in relation to university-industry collaboration will be explained more in detail. Third, it is argued that based on the UIC literature, clarification is needed as to why and for what reasons academic scientists choose to engage or not in industry collaboration. Subsequently, a number of hypotheses regarding the impact of perceived benefits and barriers on the likelihood to collaborate are articulated.

2.1. The Triple Helix thesis: a national system of innovation

In the face of increasing dynamism of competition, grand challenges, uncertain markets and rapidly changing conditions, nations pursue (technological) innovation in knowledge-based economies to increase social welfare (Etzkowitz & Leydesdorff, 2000). Although the literature on innovation models discriminate between a number of different models, for example ‘Mode 2’ knowledge production (Gibbons, Limoges, Nowotny, Schwartzmann, Scott & Trow, 1994), a ‘system of innovation’ (Edquist, 1997) or ‘post academic science’ (Ziman, 1996), the general consensus is to refer to the Triple Helix model to describe and explain national systems of innovation (e.g. Etzkowitz 2003; Leydesdorff & Etzkowitz, 1996; Etzkowitz & Leydesdorff, 1995; 2000). The Triple Helxi model describes the specific configurations of institutional 14 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(15)

spheres (or helixes) and specific arrangements of ‘university-industry-governments relations’ (Etzkowitz, 2008). The Triple Helix model postulates that the knowledge-exchange interaction between each helix or between university, industry and government on a regional or national level, is key to improving conditions for innovation in a knowledge-based society (Etzkowitz, 2003; p.295).

Over time the configurations of the helixes and thus the relations between university, industry and government evolved, resulting in different national innovation systems which can be placed in a historical context. Etzkowitz (2003; 2008) and Etzkowitz and Leydesdorff (1995; 2000) provide detailed overviews over these historical developments. In the first model, which

Fig. 1. Triple Helix Model I Fig. 2. Triple Helix Model II Fig. 3. Triple Helix Model III can be labeled ‘Triple Helix I’, the nation state encompasses academia and industry and manage the relations between them (see Figure 1.) This model could be found in the former Sovjet Union and Eastern European countries under “existing socialism” (Etzkowitz, 2003; p. 111). A second model (see Figure 2.) is comprised of separate institutional spheres with strong borders between them and highly circumscribed relations among them; often dubbed a “laissez-faire” model of university-industry-government relations (Etzkowitz, 2008). This model was typically found in the US and some Western European countries such as Sweden. Finally, the third and current Triple Helix model (see Figure 3.) is comprised out of overlapping

(16)

institutional spheres, with each taking the role of the other and with hybrid organization emerging at the interfaces (Etzkowitz, 2000; p. 112).

The first Triple Helix model is widely perceived as a failed developmental model in which innovation is more discouraged than encouraged (Etzowitz & Leydesdorff, 1995). The second, laissey-faire model still generates interests especially in countries with strong deregulation and a subsiding governmental role. The third model is the contemporary form that most countries and regions are pursuing (Etzkowitz, 2003; p. 112). The aim here is to build an innovative environment in which university spin-offs firms, tri-lateral initiatives for knowledge-based economic development, strategic alliances among firms, research groups and governmental laboratories can emerge. Such arrangements are often advocated and encouraged by the government, but not controlled as in model I (p.112). In this constellation industry operates as the locus of production, government as the source of contractual relations that guarantee stable interactions and exchange, and the university as a source of new knowledge and technology (Etzkowitz, 2003; p. 295). This means that, as all institutional spheres are equal and increasingly take on the role of the other, the traditional match between institution and function is superseded (p. 296). The dynamics of innovation in the third Triple Helix model thus work fundamentally different than prior versions, as interaction and networks among the institutional spheres provide the source of innovation and explain differentials in performance, rather than any single driver. (p. 300).

Thus, in modern knowledge-based economies, in which (tacit) knowledge has become the strategic most important resource (Grant, 1996), innovation is not determined anymore by how well knowledge is integrated within the firm, but by how well knowledge is exchanged and integrated across institutional boundaries. The question is how these developmental stages of the Triple Helix model might be best understood?

(17)

2.2. The Resource-based view of the firm: knowledge as the ultimate resource

Firms, like nations, battle for (sustained) competitive advantage in order to secure their survival (e.g. Andrews, 1971; Ansoff, 1965; Barney 2001; Teece, Pisano & Shuen, 1997). In today’s dynamic environments and uncertain market conditions, firms need to constantly adapt, renew, reconfigure, and re-create their resources and capabilities in line with their environment (Wang & Ahmed, 2007). One theoretical important and dominant perspective on the genesis of competitive advantage in these type of environments is the ‘resource-based view of the firm’ (RBV; Wernerfelt, 1984; Barney, 1991), which emphasizes the strategic importance of resources and capabilities as the base for differential firm performance (Wang & Ahmed, 2007). The core notion of the early, more static RBV is that VRIN resources (valuable, rare, inimitable, and non-substitutable), which are heterogeneously distributed across firms and imperfectly mobile, explain firms’ positions across markets, its potential profits and competitive advantage over its competition (Barney, 1991; Wernerfelt, 1989; Peteraf, 1993). Furthermore, firms also need to create and obtain distinctive (core) capabilities to make use of those VRIN resources (e.g. Penrose, 1959; Prahalad & Hamel, 1990).

Entering the 1990’s, the world changed rapidly and the original propositions of the early RBV were deemed too static in the face of increasing market dynamism (Wang & Ahmed, 2007). As the early RBV predominantly explained that a combination of VRIN resources and certain capabilities leads to (sustained) competitive advantage, it failed to clarify how the firm accomplish this precisely (p.32). Several critiques of the early RBV, what is now known as the ‘high church’ of the RBV, emerged. For example, Priem and Butler (2001) summarized that the high church of the RBV failed to clarify what resources precisely are, how value is created by the firm, that it adopted a too static view of the firm, that the assumptions of its boundary 17 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(18)

conditions remained unclear, and finally that its managerial usefulness was doubtful as it missed clear guidelines (Priem & Butler, 2001). Furthermore, Stoelhorst (2008) argued that the high church confused value creation and value appropriation. Thus an extension of the RBV was needed to address the increased dynamism by further opening the ‘black box’ of the firm (Prahalad & Hamel, 1990).

This extension became known as the ‘low church’ of the RBV and adopted a more evolutionary perspective on resource based competition (Makadok, 2001) and provided possible solutions by introducing new concepts such as dynamic capabilities (e.g. Teece et. al, 1997; Eisenhardt and Martin, 2000) and by stressing the importance of organizational capabilities and tacit knowledge as the ultimate strategic resource in knowledge-based economies (Grant, 1996).

Grant was the first to postulate a ‘knowledge-based view of the firm’ (KBV), in which he argued that if a “company’s strategically most important resource is knowledge, and if knowledge resides in specialized form among individual organizational members, then the essence of organizational capability is the integration of individual’s specialized knowledge” (p. 375). The KBV adopts an organizational-learning perspective and assumes that firms engage in activities in order to acquire, share and exchange knowledge (including technology, know-how and organizational capability) as the primary objective (Dyer & Nobeoka, 2000; Grant et al., 2004; Grant, 1996).

Furthermore, the KBV distinguishes between two conceptually distinct dimensions of knowledge management (e.g. Hamel, 1989; 1991). The first dimension is what March (1991) calls knowledge ‘exploration’ and Spender (1996) knowledge ‘generation’. It refers to those activities that increase the knowledge stock of the firm, and in which ‘learning’ is key. The second dimension is what is referred to as knowledge ‘exploitation’ (March, 1991) or

(19)

knowledge ‘application’ (Spender, 1996) and are those activities that deploy existing knowledge to create value, in which ‘accessing’ is key (Grant et al., 2004). The ability to be able to exploit and explore knowledge at the same time is perceived as a dynamic capability (e.g. Eriksson, 2014) as it drives innovation (exploration) and creates value (exploitation) simultaneously (Raisch, Birkinshaw, Probst, & Tushman, 2009). This ambidextrous ability has become today’s crucial competitive advantage and key for survival (Raisch et al., 2009). This is true for firms and more in general for knowledge-generating organizations such as universities as well (Etzkowitz, 2003). The dynamics of the Triple Helix model can thus be explained by the knowledge-based view of the firm. The evolving role of universities into entrepreneurial organizations as a new source of both social and economic (regional) development, can also be placed in this context. Universities seeking new knowledge while simultaneously exploiting their knowledge into (commercial) application, are referred to as ‘entrepreneurial universities’ in the Triple Helix model (Etzkowitz, 2003; p. 298).

2.3. The entrepreneurial university and university-industry collaboration

One of the organizing principles of the Triple Helix is the expectation that the university will play a greater role in society as an entrepreneur (Etzkowitz, 2003; p. 300). In contrast to previous innovation systems in which each institution operates on a single axis, in the Triple Helix model institutions operate on two axis, an x axis, in which they play their traditional roles, and a y axis, in which they play new roles. The underlying assumption is that institutional spheres increasingly obtain ambidextrous abilities or competences, playing multiple roles without the original role being degraded or harmed (p. 317). In case of the university there has been a clear development in its mission, moving from teaching college to research university (the first academic revolution) and then to entrepreneurial university (the second academic 19 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(20)

revolution), embracing ‘knowledge exploitation’ besides ‘exploration’ (March, 1991) by focusing on economic and social development through industry alliances and the formation of government-industry-university consortia amongst other initiatives (Barringer & Harrison, 2000; D’Este & Perkmann, 2010; Etzkowitz, 2003). This ambidexterity of universities is illustrated by the active engagement in technology development and the production of outputs, while it produces scientific knowledge as well (Ambos, Mäkelä, Birkinshaw & D’Este, 2008).

Today, the entrepreneurial university is in vogue (D’Este & Perkmann, 2010). Universities are increasingly being called upon to contribute to economic development and competitiveness. Initiatives advocated by policy-makers, such as the formation of science parks, venture labs, technology transfer offices and patenting, are aimed mainly to encourage knowledge and technology exchange with industry and increase commercial outputs (e.g. Agrawal, 2001; Powers, 2003; Bekkers et al., 2008). These exchanges can be tangible (e.g. funds, materials and equipment) and intangible (e.g. R&D output, NPD, technology and data) resources (Perkmann et al., 2013). Although it is uncertain how effective these initiatives are (Mowery, Nelson, Sampat & Ziedonis, 2004), it is evident that there is growing ambition among universities for a greater entrepreneurial role in technological innovation and applied knowledge (Taktari & Breschi, 2012). This is demonstrated by the rapid growth of these initiatives, as there is increased patenting among universities (Nelson, 2001), increased revenues derived from licensing (Thursby, Jensen & Thursby, 2001), increasing numbers of researchers engaging in academic entrepreneurship (Shane, 2005), and the rise of technology transfer offices, support offices for industry collaboration and science parks (D’Este & Perkmann, 2010).

In line with the Triple Helix model, these interactions between the entrepreneurial university and industry are growing and take on multiple forms, with interaction channels 20 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(21)

ranging from inter-organizational relationships (e.g. joint research or contract research) to spin-off companies, to IP transfer including patenting and licensing (D’Este & Perkmann, 2010; p. 319). Collaboration however, as one the interaction channels, is far more frequent than engagement in academic entrepreneurship or patenting for example (D’Este & Patel, 2007; Perkmann, & Walsh, 2007). As of such, interest in university engagement with industry partners, from both a policy and academic perspective, has grown considerably in recent years (Taktari & Breschi, 2012). This type of interaction, forms of collaboration, is referred to in the literature as University-Industry Collaboration (UIC; (e.g. Bekkers & Bodas Freitas, 2008; Siegel, Waldman & Link, 2003).

University-industry collaboration (UIC) can take on many forms and exist on all levels, ranging from academic-practitioner collaborations to governmental funded industry-university consortia (e.g. Murphy, Perrot & Rivera-Santos, 2012; Perkmann & Walsh, 2009). Moreover, similar to cross-sector collaborations (Kivleniece, & Quelin, 2012) it is assumed thatpartners in UIC typically have individual (e.g. academic publishing for universities and technical problem solving for industry) and common objectives (e.g. create impact by providing solutions for society’s problems; Ankrah & Al-Tabbaa, 2015; Bonaccorsi, & Piccaluga, 1994). However, when referring to university-industry ‘collaboration’ or ‘collaborative forms’ in this study, only the three most frequent forms of collaboration with industry are included: (1) joint or collaborative research, (2) contract research and (3) consulting (Taktari et. al, 2012; D’Este & Perkmann, 2011). With joint or collaborative research all formal collaborative arrangements aimed at cooperation on R&D projects, often publically funded or subsidized, are meant (Hall, Link & Scott, 2001). Contract research entails all the research that is directly commissioned by and commercially relevant to firms, and thus ineligible for public support (D’Este & Perkmann, 2011). Moreover, it is usually more applied than joint research

(22)

arrangements (Van Looy, Ranga, Callaert, Debackere & Zimmerman, 2004). The third and final form, consulting, refers to all research or advisory services that individual scientists provide for their industry clients (Perkmann & Walsh, 2008). Generally these type of services are commissioned directly by the firm and the income derived from this service often amounts directly to the personal wealth of the scientist (D’Este & Perkmann, 2011).

Reflecting on these forms of collaboration, two type of motivations to cooperate emerge: for entrepreneurial or commercial reasons (Etzkowitz, 2008), which is in line with the notions of the entrepreneurial university, and for traditional scientific or academic reasons as elaborated by Merton (1973). Thus, the concept of the ‘entrepreneurial university’ in relation to academics’ reasons or motivations for collaboration possibly creates tension between the traditional teaching and research missions, and the third or entrepreneurial mission (Lee; 1996; Etzkowitz, 1998). Or as Bozeman, Fay and Slade (2013) postulate, collaboration has a ‘dark’ side as knowledge exchange conflicts with ‘academic logic’.

A clarification on the nature of academics’ choice to collaborate seems therefore necessary as the current literature seems to contradict itself. For example, as Clark (1998) showed, there are researchers in the US who are eager to commercialize their technologies or science in an entrepreneurial way, thus focusing less on ‘blue skies’ research and more on research exploitation. Zucker and Darby (1996) describe how in biotechnology ‘star scientists’ are emerging who are both researcher as entrepreneur. And finally, Owen-Smith (2003) showed that a convergence is taken place in the US towards a ‘hybrid system’ of scientific publications and technological patents success due to activities of entrepreneurial scientists. These are all examples of the positive or beneficial side of collaboration.

Taktari & Breschi (2010) argue however, that the academic system still retains certain characteristics that sets it apart from other type of organizations. Academics have academic 22 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(23)

freedom and autonomy and value creative control over their projects (Aghion et al., 2008). Furthermore, scientists operate in an open-science community, ruled by norms of universalism, disinterestedness, originality, communalism and a belief that scientific breakthroughs should be placed in the public domain (Merton, 1973). In return, scientists are awarded with peer esteem, promotion, research grants, scientific prizes, and are granted freedom of inquiry (Taktari & Breschi, 2012; p. 6). Studies show how communication among scientists is restricted and publications delayed, if academic patenting is highly favored (e.g. Blumenthal, Campbell, Causino & Louis, 1996; Campbell, Weissman, Causino & Blumenthal, 2000). Other work reviewed in Taktari & Breschi (2010) illustrated that researchers who were forced to withhold data and delay publication, e.g. because firms required secrecy, felt that both the scientific and their career advancement was slowed, as the scientific reward system still favors publications. As a result, they were reluctant to engage in future collaborations with industry. As these examples illustrate the negative side of collaboration, the question remains, why do academics engage with industry? What type of benefit or barrier is most compelling in their decision-making process?

2.4. Why do academics engage? Boundary conditions, hypotheses and conceptual model So why do academic scientists choose to engage in industry collaboration? As discussed, there is generally little known about the processes through which individuals choose to collaborate (Taktari & Breschi, 2010). This applies not only to the field of academic entrepreneurship, but as Felin and Foss (2005) point out, also to the field of strategic management in general. According to Felin and Ross (2005), the majority of the recent research on strategic organization focused primarily on structure, routines, and capabilities and overlooked or failed to consider individuals in the theories proposed. Underlying this phenomenon, is the often

(24)

implicit assumption of homogeneity of individuals (e.g. Dansereau, Yammarino & Kohles, 1999; Henderson & Cockburn, 1994). As individuals form the foundation of organizations and the antecedents of all collective phenomena (Felin & Foss, 2005), they deserve careful theoretical and empirical consideration (Taktari & Breschi, 2010). This is even more true for the context of academic entrepreneurship and the knowledge-based view of the firm, because academics are a source of tacit knowledge and have great autonomy and freedom to form their decisions (Taktari & Breschi, 2010). Thus, the individual academic is likely to have different perceptions of industry-collaboration as well (Owen-Smith & Powell, 2001) and therefore an interesting unit of analysis in future studies (Taktari & Breschi, 2010). Therefore, the underlying assumption and boundary condition in this study is that researchers are heterogeneous in term of their characteristics, aversion to risk, attitudes and underlying motivations to conduct research. Thus they are likely to have different perceptions of industry-collaboration as well (Owen-Smith & Powell, 2001).

In line with Taktari and Breschi (2010) Taktari et al. (2012), and D’Este & Perkmann (2011), a second boundary condition is the assumption that individual academic decisions to engage in UIC activity follows standard cost-benefit analysis. The cost-benefit analysis is based upon the total evaluation of all the perceived benefits of participating in UIC and on all the possible factors that could act as barriers and impede the engagement in UIC (e.g. Jelinek & Markham, 2007; Killian et al., 2015). As Ajzen’s Theory of Planned Behavior (1985; 1991) predicts, depending on the outcome of the evaluation a behavioral intention is manifested to engage in collaboration or not. Thus, if the perceived benefits are higher or outweigh the potential barriers, then the person is likely to be more motivated and more likely to engage or participate in UIC (Taktari et al., 2012). The question is what kind of benefits or costs impact the decision-making process most profoundly? Will academic scientists favor traditional

(25)

academic values as described by Merton (1973) over entrepreneurial outcomes? Or is the attraction of additional funding resources and monetary incentives just too strong for any scientist to ignore? Moreover, what type of barrier or cost will be most inhibiting for academics, given their motivations towards industry collaboration?

In their literature review, Ankrah and Al-Tabbaa (2015) identified many different benefits and barriers as outcomes of UIC. Benefits and barriers can be best categorized according to type (Ankrah & Al-Tabbaa, 2015; Killian et al., 2015). Benefits are classified as either economical or entrepreneurial (Killian et al., 2015), referring to advantages from an economic perspective that encompass benefits for the individual researcher and for his or her research group, or as academic type of benefits. Academic type of benefits are benefits related to improved means for innovative research such as equipment or tools, or to improved teaching. According to Killian et al. (2015) and Taktari et al. (2012), prominent economic benefits range from more resources (such as increased personal income, more academic staff or more funds for the research group), to increased reputation as a scientist and other academic benefits (such as access to relevant research problems, empirical data and real world phenomenon, and input for teaching). Note that these type of benefits refer back to different missions of the entrepreneurial university.

Barriers or costs, on the other hand, can be classified in terms of economic and academic types as well (e.g. Ankrah & Al-Tabbaa, 2015; Taktari et al., 2012; Killian et al., 2015). Economic barriers are costs related to the imposed regulations, policies and procedures by the university of the TTO, possibly frustrating the collaboration process or the establishment of intellectual property rights (IPR’s) and patenting. While academic barriers are costs related to secrecy, reduced academic freedom and limited communication with the academic community. According to Taktari et al. (2012), most prominent barriers are loss of autonomy and freedom 25 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(26)

(due to rigid data protection or enforced secrecy for example), differences in research interests, regulations imposed or policies adopted by the university or the TTO, the short-term orientation of industry and that it can be very time consuming to find the right industry partner and conduct research together.

The literature encompasses only a few studies that empirically validated the effects of perceived benefits (D’Este & Perkmann, 2011; Lee, 2000; Peñuela, Benneworth & Castro-Martinez, 2014; Welsh, Glenna, Lena, Biscotti, 2008) or perceived barriers (Taktari & Breschi, 2010; Taktari et al., 2012), by using multivariate statistics, on the likelihood to engage in industry collaboration. The majority of this work is biased towards the beneficial side of collaboration, as only two studies were found that studied the cost side of collaboration (Taktari et al., 2012). Moreover, no study has been found that entailed both sides, perceived benefits and perceived barriers, and its effects on the likelihood to engage in UIC.

In the reviewed studies, the relationship between type of benefit and future intention to engage in UIC remains unclear (Killian et al., 2015; Ankrah & Al-Tabbaa; 2015). For example, D’Este & Perkmann (2011) highlighted the importance of learning and additional funding for research as the dominant benefits that positively impact the propensity to engage. While Taktari and Breschi (2010) found that access to additional resources or personal income did not have a significant effect on UIC engagement. Lee (2000) showed a positive relation between academic type of benefits (e.g. additional funds for research assistants and equipment, to gain insight into one’s own research, or benefits related to improved teaching and creating jobs for students) and industry collaboration, while economic type of benefits (e.g. looking for business opportunity) were only moderate compelling. Peñuela et al. (2014), and Welsh et al. (2008), found positive effects of academic type of benefits (e.g. benefits related to productivity enhancing such as new knowledge, tools and equipment, and increased network of scientists)

(27)

and UIC activity as well, and also a positive effect of economic type of benefits such as new business opportunity, improved reputation and additional income. Based on these findings, the first hypothesis with regard to the relation between perceived benefits and likelihood to engage in UIC, is:

H1. The likelihood to engage in UIC is positively influenced by perceived benefits, more specifically by perceived economic type (H1a.) and perceived academic type (H1b.) of benefits.

As mentioned, only two empirical studies related to barriers or costs have been identified in the literature. Taktari & Breschi (2010), looking only to academic type of barriers, found that the expected loss of academic freedom was the most important barrier to UIC engagement, while costs related to secrecy or the diffusion of research results seemed less important as a barrier. Furthermore, Taktari et al. (2012) reported a more extensive and elaborate study on perceived barriers and collaboration, and found that the most dominant barrier to UIC was the cost related to finding suitable industry partners and difficulty of industry’s short-term research orientation. Moreover, academic type of barriers (or Mertonian barriers according to Taktari et al. (2012) such as delays in dissemination and publication, were considered as generally most influential on the likelihood to engage in UIC. The effect of perceived economic type of barriers such as IPR agreements, regulations and policies on the decision to engage were more complex and less convincing. However, as only two studies have been identified in the literature, it is difficult to generalize these findings. Thus, it is hypothesized for the effect of perceived barriers on the likelihood to engage in UIC:

(28)

H2. The likelihood to engage in UIC is negatively influenced by the perception of UIC barriers, more specifically by perceived economic type (H2a.) and perceived academic type (H2b.) of barriers.

Although the results of these studies are inconclusive, contradicting to some extent and the number of empirical studies limited, there is cause to expect a stronger effect for academic type of benefits and barriers on industry collaboration. As Taktari et al., (2012) and Taktari & Breschi (2010) argue, the academic community’s most defining characteristics are autonomy, academic freedom and recognition by esteemed peers. Norms, values and personal drivers are essentially based on these characteristics. Furthermore, as the findings of Taktari et al. (2012) implicate, Mertonian or academic type of barriers hinder industry collaboration more profoundly than economic type of barriers. Thus stronger effects can be expected for academic type of benefits and barriers in comparison with economic types. Therefore:

H3. The likelihood to engage in UIC is stronger influenced by academic type than by economic type of arguments, more specifically, perceived academic benefits (H3a.) and perceived academic barriers (H3b.) are stronger related with the likelihood to engage in UIC than economic type of benefits or barriers.

Finally, it is worth exploring how previous experience with industry collaboration interacts with the likelihood to engage again in UIC. For example, there is reason to expect that prior experience moderates the impact of perceived barriers on the decision to engage in collaboration (Tartari et al., 2012). Owen-Smith and Powell (2001) and Shane (2000) argued, that previous collaborative experience with industry has a favorable influence on the tendency 28 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(29)

of academics towards knowledge and technology transfer. Moreover, by closely interacting with industry knowledge about markets, technologies and or consumer needs is gained, forming positive attitudes and skills toward entrepreneurship (Shane, 2000). In line with this reasoning, Takari et al. (2012) postulate that more prior experience with UIC should lower the perception of both economic and academic type of barriers to industry. As the previous experience provides academics with an opportunity to learn how to recruit industry partners to their research and to frame research questions in ways that appeal to industry (p. 661). The importance of learning via past experiences is also stressed in the alliance literature, for example by Draulans, Deman and Volberda (2003). They explain how alliances such as joint collaborations between organizations quite often result in failure, because they lack alliance capabilities as an organization. These capabilities are learned and build upon through prior alliance experiences (Draulans et al., 2003), increasing the chance of future successful alliances (p.154). Based on this discussion it is hypothesized that:

H4. Previous UIC experience positively moderates the relationship between perceived economic type (H4a.) and perceived academic type (H4b.) of barriers and the likelihood to engage in UIC.

The following conceptual model graphically represents the above discussion and hypotheses:

H2a&b - H3b. -

H1a&b. + H3a.+

Barriers to UIC

Benefits of UIC Likelihood to engage _{in UIC}

H4a&b +

Previous UIC experience

(30)

3. Methods

3.1. Research design

The aim of this study is to explore the extent to which the likelihood to engage in UIC can be explained by individual researchers’ evaluations of the perceived benefits and barriers or costs of such collaboration, controlling for personal characteristics, such as age, gender, scientific field and institutional environment. To test the various hypotheses articulated in the previous section, a deductive study was designed to gather cross sectional data using an online survey, constructed with Qualtrics software. A survey design with predetermined categories was constructed in order to provide statistically inferable data on the overall population. Although a face-to-face survey delivers the most representative results (Szolnoki & Hoffmann, 2013), an online survey was chosen as academic researchers generally work in different places on different times, making it hard to question a large sample in person. The survey item list was constructed by building on the measures and scales used in the quantitative studies that specifically focused on the relation between either the perception of benefits or the perception of barriers and the likelihood to engage in UIC, using multivariate statistics (D’Este & Perkmann, 2011; Lee, 2000; Peñuela et al., 2014; Welsch et al., 2008; Taktari & Breschi, 2010; Taktari et al., 2012). As there is no clear theory of university-industry collaboration, there are no validated scales to employ in this study. Thus the survey was based on items used in different studies, but this does not mean that all of these items are connected.

3.2. Population and sample

The Amsterdam area was chosen for this study as it hosts two of the largest research universities of the Netherlands and the largest applied sciences university of the Netherlands. These are: the University of Amsterdam (UvA), the Vrije Universiteit (VU) and the University of Applied 30 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(31)

Sciences of Amsterdam (HvA). As the focal point of this study, Amsterdam is therefore another boundary condition in this study. The combined workforce, or population, of scientific researchers employed at the 3 universities of Amsterdam, based on the online facts and figures of 2015 collected from each academic website, is approximately 5000 (VU 2200; UvA 2500; HvA 300). Thus, given an confidence level of 95% and a margin of error of 5%, the ideal sample size to base any conclusions on, is 357.

To realize this number, a sample was drawn of 450 academic researchers, either employed at the University of Amsterdam (UvA) or at the University of Applied Sciences of Amsterdam (HvA). Although a random probability sampling method is generally considered superior as it enables the calculation of the sampling error, a convenience nonprobability sampling method was selected for this study. The main reason for this is that the researcher is employed at the HvA. Thus, access to a large group of academic researchers working either at the HvA or UvA was relatively easy and therefore preferable as large groups of researchers usually are difficult to approach. For this reason the Vrije Universiteit (VU) was left out of this study. However, as the online survey was completely anonymous and distributed through intermediates, potential biases and reduction of external validity was kept to a minimum as the author had no idea who participated in the study. This is explained in detail in the data collection section.

The sample was comprised of 150 UvA (only a fraction of the research workforce) and 300 HvA academics (the entire research workforce), working in a variety of scientific fields such as economy and business, humanities, life sciences, data and information technology, and biotech. Moreover, researchers greatly differed in terms of age (between 23 and 65), academic positions, (ranging from research intern to professor) and professional experience (junior to senior).

(32)

3.3. Data collection

The 150 UvA academics, employed at many different departments, are all enlisted on a mailing list, which is owned and used for newsletters and direct mailings by the Technology Transfer Office (TTO) of the UvA, called Innovation Exchange Amsterdam (IXA). IXA agreed on forwarding the invitation for the survey, because a better understanding of why academics engage in university-industry collaborations is insightful for them as well. The email invitation briefly explained the subject and goal of the study. The email was written both in Dutch and English with two hyperlinks to the Dutch and English online version of the survey in Qualtrics. Thus participants could choose which was most convenient. To increase commitment and response rate, the email was personally signed by the researcher, as a ‘colleague’ is harder to turn down than someone unfamiliar. Furthermore, IXA endorsed the study via a brief introduction of the email invitation. See Appendix 2 for an overview of all the email invitations. The invitation was sent on the 5th of April 2016 and a reminder email on the 17th of April 2016. This resulted in a relatively low response rate of 10% (15 participants) by the end of April. Not surprisingly, as researchers probably get lots of emails daily with numerous requests it is difficult to stand out.

The 300 researchers employed at the HvA were approached similarly, by email invitation. In this case however, the email was directly send to the heads of the HvA research departments. The email invitation was the same as the one send to the UvA academics (see Appendix 2). It explained the subject and goal of the study, it had an English and Dutch version and it was personally signed by the author and endorsed by the head of the department. However, not all departments complied. In the end, 5 out of 7 departments participated in the study (with a combined workforce of 220 researchers, employed at those 5 departments) and agreed to forward the email invitation to their researchers. In line with the UvA invitation, HvA 32 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(33)

researchers were emailed from April 5th and after. A reminder was send on the 17th of April. This resulted in a slightly better response rate of 13% (39 participants out of 300) or 18% (out of 220 participants), depending the frame of reference.

The survey was thoroughly pretested by the researcher and the thesis supervisor. On average it took participants 12 minutes to complete the survey. The survey consisted in total of 49 items, divided over 23 separate questions with various answer scales (see Measurement of constructs for more details). After completing the survey participants were debriefed and thanked for their corporation. Furthermore, an email address was provided in case participants would like to stay informed about the results of the study.

After April 2016 the study was paused for several months due to other obligations and priorities of the researcher. In November 2016 the data collection continued. After careful consideration, both IXA and the HvA departments agreed to send one last and final reminder email (the same email as the first reminder) to the sample group. This resulted in an UvA response rate of 27% (40 participants) and a HvA response rate of 20% (61 participants out of 300). The combined response rate was 22% (101 out of 450). 7 Cases were not registered correctly by the Qualtrics software and therefore entirely missing. Furthermore, the completion rate was 69% as only 70 out of the 101 participants completed the entire survey. As it was difficult to approach the group once more, no further attempts were made to increase the response or completion rate. Thus 70 complete surveys were used for further analysis.

3.4. Measurement and reliability of constructs

The core constructs of this study were measured using combined and closely related multi-item scales used in previous studies, with no empirical validation as discussed above. Accordingly, the following constructs were measured in this study: likelihood to engage in UIC (dependent 33 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(34)

variable), type of benefit (economic or academic; independent variable), type of barrier (economic or academic; independent variable) and previous experience (independent variable) as a possible moderator on the relation between barriers and engagement. Finally a number of control variables were measures as well.

Dependent variable: Likelihood to engage in UIC

As the dependent variable in this study, the variable ‘Likelihood to engage in UIC’ was constructed to measure future intentions to collaborate. Following Taktari et al. (2012) and Killian et al. (2015), the corresponding item was framed as “In the future it is likely that I will engage in a collaboration with industry”. The Theory of Planned Behavior (Ajzen, 1985; 1991) states that a behavioral intention is a dominant predictor of actual behavior. Thus by asking participants’ intentions to collaborate, a prediction can be made about future behavior (Killian et al., 2015). Furthermore, as people generally either have or don’t have an intention to behave in a certain way, the choice was made to represent the dependent variable as a dichotomous (or binary) variable with a nominal scale, with ‘0’ representing ‘not likely to engage in future UIC’ and ‘1’ representing ‘very likely to engage in future UIC’. As a single item scale, the reliability is both unknown and unknowable (De Vellis & Dancer, 1991) and therefore cannot be calculated.

Independent variables: type of benefits and barriers, and previous UIC experience

In order to test hypotheses 1 and 1a, two variables were constructed to measure the independent variable ‘perceived type of benefit’, Economic benefits and Academic benefits. The variable Economic benefits (α = .559) was constructed by creating one multi-item scale based on the different items used in the work of D’Este & Perkmann (2011), Lee (2000); Peñuela et al. 34 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(35)

(2014) and Welsh et al. (2008) that measured perceived economic benefits of collaboration. D’Este & Perkmann (2011) used a multi-item, 5-point Likert scale containing 11 items that focused on both economic benefits (5 items) and academic benefits (6 items). Participants needed to rank distinct type of benefits according to their importance, with ‘1’ being not important and ‘5’ being extremely important. Examples of economic benefits, are ‘collaboration provides a source of income’, ‘seeking intellectual property rights’, and ‘access to funding’. Similarly, Lee (2000) adopted a multi-item, 5-point Likert scale containing 7 items, questioning both economic (3 items) and academic (4 items) type of benefits. Likewise, participants needed to rank distinct type of benefits according to their importance, with ‘1’ being not important and ‘5’ being extremely important. Examples of economic benefits, are ‘securing funds for research’ and ‘look for business opportunities’. Peñuela et al. (2014), used a multi-item, 4-point Likert scale (ranging from ‘1’ not important to ‘4’ very important), containing only 5 items that focused on economic (3 items) and academic (2 items) benefits. Examples of economic type of benefits are ‘additional funds’, ‘access to equipment or infrastructure’ and ‘to be part of a professional network’. Finally, Welsh et al. (2008) also used a multi-item, 7-point Likert scale (ranging from ‘1’ not characteristic to ‘7’ highly characteristic) containing 7 items that focused on economic (2 items) and academic (5 items) type of benefits. Participants were asked to rate a number of statements regarding collaboration with industry. Examples of economic type of benefits are ‘provides new monetary support for postdocs and graduate students’ or ‘provides new research funds’. Based on these different multi-item scales, one multi-item, 5-point Likert scale (ranging from ‘1’ strongly disagree, to ‘5’ strongly agree) was constructed for economic type of benefits, excluding those items that enquired the same type of benefit (e.g. multiple studies asked about the importance of additional funding). Thus a multi-item scale with 5 items remained, enquiring about the 35 Radboud Dam, 0215856, Master Thesis, Amsterdam Business School, 2016-2017

(36)

importance of increased funds, equipment, reputation, personal income, IPR’s, and network, thus covering all economic type of benefits of industry collaboration articulated in the previous studies (see Appendix 2 for a complete overview of the measures used in the survey).

With a Cronbach’s Alpha = .559 the reliability of this scale is only moderately strong. The corrected item-total correlations indicate that 3 items (the importance of increased IPR’s, network and the reversed item on personal income) correlated poorly with the total score of the scale (< .30). It may be that IPR’s or an improved business network are not important factors for the questioned scientists. As for the reversed or contra-indicative item regarding the importance of personal income, participants might have just misread the statement. However, none of these items would substantially affect the reliability of the scale if they were deleted. Similarly, the variable Academic benefits (α = .769) was constructed using one multi-item scale containing 10 multi-items, based on the different multi-items used in the same work of D’Este & Perkmann (2011), Lee (2000); Peñuela et al. (2014) and Welsh et al. (2008) that besides economic type of benefits also measured perceived academic benefits of collaboration. 6 Questions of the multi-item scale used in the study of D’Este & Perkmann (2011) informed about academic type of benefits. Examples are ‘collaboration fosters learning, because it informs me about industry problems, research issues and / or it provides me with feedback’ and ‘collaboration grants me access to additional equipment and research expertise’. Lee (2000) incorporated 4 questions on academic type of benefits with examples like ‘collaboration helps me to gain insight in my own research’ and ‘it provides me with knowledge that I can use for teaching’. Peñuela et al. (2014), only incorporated 2 items on academic type of benefits: ‘collaboration provides me with access to experience of non-academic professionals’ and ‘it grants job opportunities for students’. Finally, Welsh et al. (2008) also informed about academic benefits via 5 items. Examples are ‘increases access to new knowledge’, ‘accelerates

(37)

new product development’, and ‘increases access to new research tools’. Based on these different multi-item scales, one multi-item, 5-point Likert scale (ranging from ‘1’ strongly disagree, to ‘5’ strongly agree) was constructed for academic type of benefits, excluding those items that informed about the same type of benefit (e.g. multiple studies asked about the importance of increased opportunities for student jobs). Thus a multi-item scale with 10 items remained, informing about the importance of new research ideas, new knowledge, new product development, learning, insight in research, opportunities for students and teaching, and better network among scientists (see Appendix 2 for a complete overview of the measures used in the survey).

With a Cronbach’s Alpha = .769 the reliability of this scale is strong. The corrected item-total correlations indicated that all items have a good correlation with the total score of the scale (> .30). Also, none of these items would substantially affect the reliability of the scale if they were deleted.

In order to test hypotheses 2a and 2b, two variables were constructed to measure the independent variable ‘perceived type of barrier’, Economic barriers and Academic barriers. The variable Economic barriers (α = .678) was constructed by using the same multi-item scale with 5 items, used by Taktari et al. (2012), as Taktari & Breschi (2010) only informed about academic type of barriers. The different economic type of barriers, Taktari et al. informed about were: (1) regulations imposed by university of funding agency; (2) policies adopted by the university’s TTO; (3) potential conflicts with industry regarding intellectual property rights; (4) absence of established procedures for collaboration with industry; (5) low business profile of the university’s TTO. These different type of barriers were presented as 5 statements with a 5-point Likert scale (ranging from ‘1’ strongly disagree, to ‘5’ strongly agree). The same items

(38)

and scales were used for this study (see Appendix 2 for a complete overview of the measures used in the survey).

With a Cronbach’s Alpha = .678 the reliability of this scale is was good to strong. The corrected item-total correlations indicated that 2 items (difficulties with IPR’s and lack of established procedures) correlated poorly with the total score of the scale (< .30). Again, it seems that IPR’s are not a particularly important factor in decision-making process. as or an improved business network are not important factors for the questioned scientists. Furthermore, by removing the item on IPR’s, the reliability of the scale would improve from good to strong (α = .714). Therefore this item was deleted.

The variable Academic barriers (α = .760) was constructed by creating one multi-item scale containing 9 items, based on the different items used in the work of Taktari & Breschi (2010) and Taktari et al. (2012). Taktari & Breschi (2010) used a multi-item, 5-point Likert scale containing different statements about academic type of barriers. They questioned participants about loss of academic freedom and increased secrecy when collaborating with industry via 6 items. Exemples are ‘collaboration endangers basic research’ and ‘industrial partners require secrecy over research results’. Taktari et al. (2012), informed about academic type of barriers using 6 items as well. Examples of their items are ‘barriers due to the short-term orientation of industry’, ‘differing research interests or needs’ and ‘lack of research understanding’. Based on these different multi-item scales, one multi-item, 5-point Likert scale (ranging from ‘1’ strongly disagree, to ‘5’ strongly agree) was constructed for academic type of barriers, excluding those items that informed about the same type of barrier (e.g. delays in dissemination or publication of results). Thus a multi-item scale with 9 items remained, enquiring about the importance of topic choice of research, interests, continuity, dissemination,

Making research alliances between academia and industry in the Triple Helix Model work : the effect of academics’ evaluations of perception of benefits and barriers on the likelihood to engage in university-industry col

Radboud Dam

0215856