Abstract
To assess the progress of the burgeoning entrepreneurship research literature and whether the entrepreneurship research converges or diverges since Shane and Venkataraman (2000, 2001), we conducted a systematic review of dependent variables in the entrepreneurship literature across the US and Europe. We find that the entrepreneurship literature becomes relatively stable since 2005 and that the US and the European literature converge on seven core categories of dependent variables. We then develop an integrative model of dependent variables to summarise previous research. We conclude our study with mentioning the implications for the existing literature and future studies in the field of entrepreneurship.
Shane and Venkataraman (2000) propose three central questions of entrepreneurship: ‘(1) why, when and how opportunities for the creation of goods and services come into existence; (2) why, when and how some people and not others discover and exploit these opportunities; and (3) why, when and how different modes of action are used to exploit entrepreneurial opportunities’ (p. 218). Later, Shane and Venkataraman (2001) further argue that the outcomes of exploiting entrepreneurial opportunities (e.g., industry and society development) should be also considered. Shane and Venkataraman (2000, 2001) and other such articles helped establish entrepreneurship as a field of study to include the existence, discovery, pursuit and consequence of entrepreneurial opportunities.
Later studies also helped establish additional boundaries for the field by distinguishing entrepreneurship from closely related areas such as strategy and organisational theory and behaviour (Busenitz, West, Shepherd, Nelson, Chandler & Zacharakis, 2003). However, while the entrepreneurial opportunity remains one of the main focuses in the entrepreneurship literature, how the field has progressed is less well known. Because entrepreneurship is still a new academic field, we argue that it is time to take inventory of the research produced since Shane and Venkataraman (2000, 2001) and examine the progress of the field.
VanderWerf and Brush (1989) argue that an emerging field such as entrepreneurship may first converge on a few distinct topics, and then diverge again. As an emerging field progresses, there is often a tension between convergence and divergence (Brush, Manolova & Edelman, 2008). For the field of entrepreneurship, convergence can help researchers focus on those phenomena distinctive about entrepreneurship (Shane & Venkataraman, 2000), and allow researchers to compare their results. On the other hand, there are still many debates within the field of entrepreneurship in terms of what phenomena are important, and being too convergent may limit the breadth of phenomena examined (Brush et al., 2008).
The issue of convergence and divergence remains a hot topic in entrepreneurship (Davidsson, 2003; Erikson, 2001; Grégoire, Noël, Déry & Béchard, 2006; Shane & Venkataraman, 2000, 2001; Singh, 2001; Ucbasaran, Westhead & Wright, 2001; Zahra & Dess, 2001). While many researchers believe that the field of entrepreneurship remains highly divergent (Aldrich, 1992, 2000; Aldrich & Baker, 1997; Davidsson, Low & Wright, 2001; Low, 2001; Low & MacMillan, 1988; Wortman, 1987), other research begins to help establish the convergence of the field (Grégoire et al., 2006). Nevertheless, the debate about the progress of the field (convergence versus divergence) remains important and is likely to continue.
In this study, we try to assess the progress of the field since Shane and Venkataraman (2000, 2001) by examining the dependent variables in the entrepreneurship literature. Brush et al. (2008) argue that the dependent variable is critically important in research, because this variable is a leading indicator of the cumulative nature in the field. A better understanding of the dependent variables can tell us what we are studying, and categories of dependent variables can be helpful in showing the progress of the field (Brush et al., 2008; Grégoire, Meyer & De Castro, 2002). Specifically, the trend of categories of dependent variables may inform us whether the field is converging on a few categories or diverging towards more categories over time (Brush et al., 2008).
While the question of what we are studying remains central, most of these reviews focus on documenting the methodological practices of entrepreneurship (Aldrich, 1992; Aldrich & Baker, 1997; Chandler & Lyon, 2001; Churchill & Lewis, 1986; Paulin, Coffey & Spaulding, 1982; Wang, Jessup & Clay, 2013). Although some studies focus on dependent variables, these studies are not without limitations. For example, Brush et al. (2008) reviewed previous research and classified dependent variables. However, some of Brush et al.’s (2008) categories cannot be differentiated from each other. For example, it is not quite clear if their innovation and alliance/network categories are completely differentiated, given that many alliances are developed to facilitate innovation (Tiessen, 1997). Furthermore, developing an integrative model to capture categories of dependent variables can provide even more insights. Such an integrative model helps understand how these categories are related and how addressing one category can help address and understand another.
To summarise, the entrepreneurship literature has become even more popular since the publication of Shane and Venkataraman (2000, 2001)and it would be helpful to examine the progress of the field and assess if there is any convergence or divergence on what we are studying. In this study, we review dependent variables between 2002 and 2012 from four top journals. We also compare the European literature to the US literature. Because the field of entrepreneurship has become more internationalised and the European literature has expanded rapidly (Brush et al., 2008), it can be very insightful to assess if there is any difference across different regions as the field progresses. We then develop an integrative model to capture the dependent variables in the entrepreneurship literature. Our study addresses four broad research questions:
What are different categories of dependent variables in the entrepreneurship research? What are the trends of dependent variables in the entrepreneurship research? Is there any difference in the categories of dependent variables across origin of journals? Can an integrative framework be developed to capture the essence of entrepreneurship research to date?
The rest of this article is organised as follows. First, we review four top research journals in the field from 2002 to 2012. Then, we present the findings of the classification as well as the trend of the dependent variables. Next, we compare the difference between the US and the European literature. Based on those findings, we present the integrative model that captures these dependent variables. Finally, the article concludes with the implications of our study.
Method
To assess the progress of the entrepreneurship literature, we conducted a systematic review, the purpose of which is to identify, appraise and synthesise the existing research literature with a replicable, scientific and transparent process (Petticrew & Roberts, 2006; Tranfield, Denyer & Smart, 2003). In this section, we explain our sample, the protocols for article admittance and exclusion, following Tranfield et al. (2003).
Planning the Review
To plan the review, it is essential to create a review protocol, including ‘the specific questions addressed by the study, the population (or sample) that is the focus of the study, the search strategy for identification of relevant studies, and the criteria for inclusion and exclusion of studies in the review’ (Tranfield et al., 2003, p. 215). Because our study tries to examine the whole entrepreneurship literature, we do not specify any search questions. Our sample is from published articles in four top journals: two published in the US and two published in Europe. Specifically, we reviewed Entrepreneurship Theory and Practice (ETP), Journal of Business Venturing (JBV), International Small Business Journal (ISBJ) and Small Business Economics (SBE). These four journals are selected because they are well-respected outlets for entrepreneurship research (Brush et al., 2008), and have relatively high impact factors (ETP: 2.542; JBV: 3.849; ISBJ: 1.492 and SBE: 1.549). We admit that our results can be limited by the journals selected.
Our criteria for including and excluding studies followed the approach of Brush et al. (2008). We included empirical and conceptual studies, and excluded articles in other scholarship formats (e.g., editorials, commentaries and teaching case studies) because these studies do not have clearly defined dependent variables. Specifically, for conceptual studies, we included those that clearly define their dependent variables or present their conceptual models; for empirical studies, we included quantitative studies, as well as qualitative studies clearly articulating their dependent variables based on their qualitative data.
Conducting the Review
We systematically reviewed every article published between 2002 and 2012 in four journals. We selected 2002 as the starting year to assess the progress of the entrepreneurship literature since Shane and Venkataraman (2000, 2001). To reduce human error and bias, we created a data-extraction form for each included article (Tranfield et al., 2003), and each data record included the general information for that article (author, title, publication details, brief description, method, sample, dependent variables). The output of this step was a full listing of articles with relevant information, which served as the data repository on which the data analysis was based (Tranfield et al., 2003).
For each study, we included outcomes and mediators. A dependent variable refers to an ‘output’ factor in a relationship modelled as regressed, explained or outcome variable (Dodge, 2003); a mediator refers to a variable transmitting the effects of another variable and explains a relation between other variables (Vogt, 1999). In other words, a mediator functions as a dependent variable in one relationship and as an independent variable in another relationship. Therefore, a mediator can be seen as a specific type of dependent variable. For example, in Edelman, Brush and Manolova (2005) study, resource is the independent variable, strategy is the mediator and performance is the dependent variable. Our review included both strategy and performance.
Coding Dependent Variables
After we identified all of the relevant articles, we then coded dependent variables. The iterative process of identifying categories and checking for consistency and validity was based on Jones, Coviello and Tang (2011) and Grégoire, Corbett and McMullen (2011). Fifty papers were first randomly selected and the dependent variables were coded independently. In this stage, categories and stakeholders for each dependent variable were identified. The results were then compared, and any differences were resolved. Next, the rest of the papers were coded independently. In this stage, new stakeholders and categories were added whenever necessary. The coding resulted in more than 90 per cent consistency. Then any difference of coding was discussed. In this process, Trauth and Jessup’s (2000) process was followed.
Our coding process was guided by the literature and an inductive approach of category identification (Lee, 1991), which is different from previous studies examined dependent variables (Brush et al., 2008; Grégoire et al., 2002). Specifically, we follow a partially grounded means to develop meaningful categories through the iterative process of examination, connection and re-examination by focusing on conceptual meanings of dependent variables (Glaser & Strauss, 1967). When the categories from the literature were suitable for our study, we used them directly; when the categories from the literature could be combined, we used the combined categories; and when no appropriate categories were available to classify certain dependent variables, we let the coding categories emerge as our interpretive understanding and engagement with the dependent variables. In this context, we began with the inductive development of provisional categories, engaged in an ongoing examination of categories and comparison of new categories with the dependent variables that had already been coded, and subsequently altered existing categories as others were created or eliminated (Strauss, 1987). Our coding resulted in 17 finely grained categories (see Table 1). The interpretative account of each coding was confirmed by returning to the papers and employing pattern-matching against category fit.
During the coding, we focus on dependent variable’s stakeholder rather than level of analysis. If we simply divided the dependent variables based on level of analysis (e.g., individual, firm and societal level), there is no way to tell, for example, if firm performance in one study refers to ventures’ performance or investors’ investment outcomes, making our developed categories confusing.
A necessary (but insufficient) requirement for the reliability of interpretive content analysis is the detailed documentation of procedures (Kirk & Miller, 1986; Yin, 2013). It is also necessary to follow approaches that can demonstrate how the categories are consistent with the dependent variables. This happens when readers, after having gone through the explanations of each category, are able to see how the clarification is meaningful. In other words, the objective of reliability in interpretive analysis considers the extent to which the observational process yields observably consistent findings.
One method to support validity is replication. Through replication across multiple dependent variables, the findings (categories, in this context) are shown to be generalisable beyond the immediate case (Yin, 2013). Here the objective of validity in interpretive studies is not to verify a correct answer but rather to convince the reader that a believable story is being told.
Coding Schema
A certain degree of arbitrariness may occur with any attempt to organise past research. During the coding, some dependent variables stand alone and may not seem to fit any category, while others may fit into several categories. In these cases, we try to determine the best match to keep the categories succinct. One example is alliance capital from Coombs, Mudambi and Deeds (2006). While the name of the dependent variable may make it fit into ‘venture resources’, Coombs et al. (2006) actually assign alliance capital to the practices of alliance development. Therefore, we decided to put this variable into ‘strategies and practices’.
Results
The number of studies included is shown in Table 2. Our study is not to examine the methods used in previous studies. Therefore, we did not further break the number based on the methods.
Research Question 1: What are different categories of dependent variables in the entrepreneurship research?
Our review identified a total of 1,585 dependent variables for four different stakeholders. The four different stakeholders are entrepreneurs (also including employees), investors (including venture capitalists, business angels and bankers), other agencies that provide various supports to entrepreneurs, and environment. Based on the different stakeholders, our interpretation of the literature yielded 17 specific categories to group the literature. Table 3 summarises the results, and Table 4 presents examples of previous studies in each category.
Our analysis revealed that different stakeholders received different levels of attention. For example, 90.91 per cent of the dependent variables examine entrepreneurs, while only 1.01 per cent of the dependent variables focus on other agencies. For specific categories, the percentages range from 35.90 (start-up outcomes) to 0.06 (investors’ characteristics).
Entrepreneurship Studies Includeda
Number of Dependent Variables from Each Categorya
We admit that our categorisation method is not without limitations, and there may be other possible ways to classify dependent variables. However, the method for classifying the dependent variables should be consistent with the research objectives. For example, we could have followed Carpenter, Geletkanycz and Sanders (2004) and classified the dependent variables into the categories of individual/team, performance and strategic categorisation, in which case the attitude, intention and individual characteristics would go into the entrepreneur/team category. While this categorisation might have been adequate in many contexts, we found it less useful for our study. Readers may ask what the entrepreneur/team category means exactly. We believe that our categorisation method can preserve the conceptual meanings between these categories, and is appropriate for our study.
Examples of Dependent Variables
Research Question 2: What are the trends of dependent variables in the entrepreneurship research?
To assess the trends of dependent variables in the entrepreneurship research, we first look at how the number of dependent variables’ categories change between 2002 and 2012. Table 5 shows the number of dependent variables examined each year and if any study examines a certain category of dependent variables in a certain year, and Figure 1 shows how the number of categories changes.
Based on the results, the trend of the dependent variables in the entrepreneurship research between 2002 and 2012 can be divided into two stages. First, between 2002 and 2004, the number of categories decreased to the lowest level of seven. Second, between 2005 and 2012, the number of categories increased again and remained stable between 12 and 14. These results show that immediately after Shane and Venkataraman (2000, 2001), researchers seemed to first focus on distinctive phenomena about entrepreneurship and the field appeared to become more convergent. Then after three years, researchers began to examine more topics related to entrepreneurship, and the field became a little more divergent and then remained relatively stable.
We then further look at the trend of specific categories as shown in Table 5. Based on the results, 17 categories can be divided into three groups. Categories from the first group receive stable attention from the field. There is at least one study dealing with categories of pre-launch attitude, pre-launch intention, entrepreneurial entry, venture resources, strategies and practices, start-up outcomes and industry and region outcomes in each year. Further, all these categories focus on entrepreneurs. Therefore, there is a convergence regarding to the important phenomena of entrepreneurs for the field. Categories from the second group receive relatively high attention. There is at least one study examining categories of individual characteristics, pre-investment strategies and practices, decisions and investment, investment outcomes and environmental characteristics for more than half of the period between 2002 and 2012. These results may indicate that these five categories can be important for the field, but researchers paid mixed attention to these categories. Categories from the third group receive relatively little attention. Less than half of the period has any study focusing on pre-investment resources, post-investment practices, investors’ characteristics, other agencies’ practices and supports. These results suggest that some researchers believe that these categories are important, but the field as a whole has not yet embraced these categories.
Number of Dependent Variables from Each Category per Year

To summarise, our results show that the field first became more convergent and focused immediately after 2002, and then became divergent and relatively stable since 2005. Further, the categories of dependent variables can be divided into three groups the core component, the emerging component and the peripheral component of the field (see Figure 2). Overall, the entrepreneurship literature seems to converge on the essential phenomena of entrepreneurs.
Our results also show that the publications of Shane and Venkataraman (2000, 2001) have a significant impact on the field in two aspects. First, seven categories of the core component can be fit into the framework proposed by Shane and Venkataraman (2000, 2001): categories can be grouped into opportunity discovery (pre-launch attitude, intention and entrepreneurial entry), opportunity exploitation (venture resources and strategies and practices) and exploitation outcomes (start-up outcomes and industry and region outcomes). Second, in 2004, when the number of categories became lowest and the field became most convergence, the literature focuses exclusively on these seven categories. These results show that the work of Shane and Venkataraman (2000, 2001) helps lay out the foundation and the core of the field, and researchers agree on the importance of the phenomena proposed by Shane and Venkataraman (2000, 2001). On the other hand, as the field progress, some researchers realise that there are other important phenomena relevant to the field and worth pursuing. Thus, the field became divergent again.

Research Question 3: Is there any difference in the categories of dependent variables across origin of journals?
In the US literature, the results of ETP and JBV are quite similar: both journals examine dependent variables from a variety of categories such as pre-launch attitude, pre-launch intention, entry, venture resources, strategies and practices, start-up outcomes, industry and region outcomes and individual characteristics. Both journals also examine investors’ activities (strategies and practices, decision and investment and investment outcomes) as well as other agencies’ support to entrepreneurs. On the other hand, there are two differences between ETP and JBV: first, EPT has examined employees’ outcomes while JBV does not; second, JBV has assessed pre-investment resources while ETP does not. As a whole, the US literature deals with the whole process of entrepreneurship from pre-launch to launch and post-launch (Baron, 2002), examines the beginning (pre-investment) and the outcomes (post-investment) of the investment process, as well as tries to understand how other agencies (e.g., universities) support various stages of the entrepreneurship process.
In the European literature, the difference between ISBJ and SBE is more obvious. Both journals examined dependent variables from categories of pre-launch attitude, intention, entry, venture resources, strategies and practices, start-up outcomes, industry and region outcomes, individual characteristics, other agencies’ practices and support and environmental characteristics. On the other hand, no studies from ISBJ focus on investors, while SBE examines categories of pre-investment strategies and practices, resources, decisions and investment, post-investment practices, investment outcomes and investors’ characteristics. SBE also examines employees’ practices during venture development and their outcomes, while ISBJ does not. As a whole, the European literature deals with the whole process of entrepreneurship and investment, other agencies’ practices and supports, as well as environmental characteristics.
While the US literature is consistent with the European literature in many aspects, there are several interesting differences. First, the US literature deals with fewer stakeholders than the European literature. For example, no studies in the US literature examine the environment surrounding ventures. Second, the European literature examines additional categories of dependent variables such as post-investment practices, investors’ characteristics, other agencies’ practices and environmental characteristics, while no dependent variables from the US literature deal with these categories. Third, when dealing with investment outcomes, the European literature focuses only on financial outcomes and ignores other aspects of investment outcomes. On the other hand, the US literature examines not only financial outcomes but also other aspects of investment outcomes.
Research Question 4: Can an integrative framework be developed to capture the essence of entrepreneurship research to date?
Previous literature has proposed several conceptual models or frameworks, and we provide a few examples in Table 6. Overall, each work has some missing parts and cannot fully capture our results. Therefore, we propose a new model integrating different categories of dependent variables, which is thought to be ‘an important part of the reporting process’ (Tranfield et al., 2003, p. 219; see Figure 3). Note that our presentation of previous literature is not intended to be a critique, but to highlight the limitations of previous literature and how our model can address these limitations. Because most relationships in the model have been examined in the literature, we provide only a brief description of our integrative model. We choose those relationships that seem to be the most supportable based on the literature. We fully admit that other relationships may exist in the current literature or will emerge in future studies, and we encourage other discussions or perspectives to complement our model.
Examples of Conceptual Models and Frameworks from Previous Literature
Entrepreneurs
Pre-launch Decisions
Shane (2003, p. 10) argues that an entrepreneurial process ‘begins with the perception of the existence of opportunities, or situations in which resources can be recombined at a potential profit’. After entrepreneurs identify entrepreneurial opportunities and form their attitudes towards these opportunities, they decide whether to exploit the opportunities found. According to Theory of Reasoned Action, individuals’ attitudes towards certain activities influence their intention to conduct these activities, which in turn impact their actual behaviours (Fishbein & Ajzen, 1975). Therefore, the more positive entrepreneurs feel towards certain entrepreneurial opportunities, the more likely the entrepreneurs intend to exploit these opportunities and engage in actual entrepreneurial activities.

Launch Activities
In order to exploit certain entrepreneurial opportunities, entrepreneurs first need various resources. Shane (2003) summarises that financial resources are important for ventures because abundant financial resources overcome the liquidity constraints. In addition, ventures need other kinds of resources, such as human resource value, to progress (Andrews & Welbourne, 2000). Therefore, when entrepreneurs acquire sufficient resources, they are more likely to adopt appropriate strategies to exploit opportunities.
Post-launch Outcomes
We should not only examine the outcomes for entrepreneurs or ventures but also consider industry and society (Cohen, Smith & Mitchell, 2008; Shane & Venkataraman, 2001). When ventures survive, increase the number of employees they hire, and generate profits, industry and society as a whole may enjoy such benefits as a lower level of unemployment. Therefore, the success of ventures can have many positive impacts on industry and society.
Entrepreneurs’ Individual Characteristics
Shane (2003) emphasises the importance of individual characteristics on various steps of the entrepreneurial process. Entrepreneurs are not randomly determined, and certain individual characteristics influence entrepreneurs’ decisions to exploit entrepreneurial opportunities (Shane, 2003). Further, individual characteristics also influence entrepreneurs’ subsequent activities and performance. For example, Kent, Van Auken and Young (1982) find that education provides entrepreneurs with useful skills for opportunity exploitation. Entrepreneurs with higher education levels will probably make better decisions and follow appropriate strategies that ultimately improve venture performance.
Other Agencies
Other agencies can provide various supports in different stages of the entrepreneurial process. For example, the Small Business Development Centre (SBDC) programme in the US is to assist the start-up and growth of entrepreneurial ventures (Chrisman, Gatewood & Donlevy, 2002). The SBDC’s services include assisting entrepreneurs with such things as financial aspects, marketing, production, organisation and other related problems. Support from agencies such as the SBDC can help entrepreneurs acquire useful skills to better evaluate entrepreneurial opportunities, choose more suitable strategies and enjoy enhanced performance.
Investors
Investors’ Investment Decisions
Investors invest in ventures so that they can get returns in the future. The strategies that investors follow, therefore, influence the way in which they make investments. For example, if certain investors tend not to emphasise early-stage investment, and a venture in its early stages tries to get financed via these investors, then these investors will probably decide not to invest in that venture. Furthermore, to make investments, investors first need financial resources. Sufficient resources can provide investors the ability to make investments in the ways desired. Otherwise, with limited financial resources, investors might not be able to invest in the way that they would like, which in turn influences their decisions to invest.
Ventures’ strategies and practices can also influence investors’ decision to invest. Before making decisions to invest, investors may first evaluate ventures’ likely profitability and survival based on ventures’ strategies and practices (Bruton, Fried & Manigart, 2005). When investors perceive that ventures adopt appropriate strategies and practices and are more likely make much profit in the future, they are more likely to invest in these ventures.
Once investors decide to invest in certain ventures, the financial resources of those ventures increase and those ventures can better put their efforts into obtaining their desired value from opportunity exploitation. However, if investors decide not to make investments, then those ventures might not have sufficient financial resources to operate and deliver planned products or services.
Investors’ Individual Characteristics
Investors also have individual differences, which may influence the process of investment. For example, investors with more experiences may be more easily obtain financial resources to invest and may be more likely to make accurate decisions. In addition, during the investment, those investors with more experience are more likely to know how to interact with entrepreneurs and how to facilitate venture development, which in turn leads to higher returns from the investment.
Environment
Different characteristics of environment may also influence the entrepreneurial process. For example, with lower barriers of internationalisation, entrepreneurs are more likely to conduct international-oriented strategies and try to expand their ventures internationally. These strategies and practices may in turn influence ventures’ subsequent sales and revenue performance.
Overall, our model has several attributes unique to previous literature. First, our model highlights start-up outcomes, the final stage of the entrepreneurial process. Thus, our integrative model shows the final stage of the entrepreneurial process as well as highlights the importance of other parts of the model (e.g., strategies and practices) in achieving outcomes. Second, our integrative model highlights that in order to exploit opportunities; launch activities are executed not only by entrepreneurs but also by employees.
Third, our integrative model also includes investors, other agencies and environment. Explicating dependent variables focused on investors is consistent with the literature in recognising the importance of the relationships between entrepreneurs and resource suppliers (Venkataraman, 1997). Further, one must think more broadly about the field of entrepreneurship highlighting other agencies and environmental causes.
Fourth, our model does not view post-launch outcomes and post-investment outcomes as the end. Rather, as entrepreneurs finish a complete progress of opportunities exploitation, they may continue to pursue other opportunities in the future; similarly, as investors complete a specific investment, they turn to other ventures and invest.
We think it is appropriate to regard this integrative model as a dynamic, digital snapshot in time rather than as a complete, rare painting. The development of an integrative model of the dependent variable is an ongoing process, given that the knowledge within the literature changes over time. As Low and MacMillan (1988, p. 139) suggest, ‘As a body of literature develops, it is useful to stop occasionally, take inventory of the work that has been done, and identify new directions and challenges for the future.’ Therefore, our aim here is to take inventory of what has been studied. While we hope our model can reasonably represent the literature as well as guide future studies, we do not intend to suggest that it is the model for entrepreneurship literature and will never change.
Discussions
Implications for the Field of Entrepreneurship
Our study has important implications for the field of entrepreneurship. Our study identified seven core categories of dependent variables and researchers agree that these seven categories are important for the field of entrepreneurship. The number of categories remains relatively stable since 2005. These results are consistent with the argument of Vander Werf and Brush (1989) in that an emerging field may first converge and then diverge again. Specially, the field first became convergent between 2002 and 2004, and then became divergent and relatively stable since 2005.
Davidsson et al. (2001) argue that the challenge for the progress of the field is to examine a set of phenomena which is neither too broad nor too narrow. Our results may help solve this issue and guide the future progress of the field. First, we identify seven core categories of the field, and researchers can focus on these categories so that the entrepreneurship literature can be distinctive from other related fields. Second, we also identify emerging as well as peripheral categories of the field, and these categories can make sure that the field has a certain level of divergence and does not become too narrow. Further, these categories can be good candidates when the field decides to include additional important and distinctive phenomena as core components.
Implications for Future Studies
Our study also has important implications for future studies. First, our study shows that employees’ attitude and intention have not received much attention in previous literature. Because entrepreneurs cannot conduct every action by themselves and may need employees to facilitate the progress of the ventures, the importance of employees is also recognised in the literature. As seen from our results and integrative model, studies have examined employees’ activities (during launch) and outcomes (post-launch). However, we do not find studies trying to understand how they decide to engage in opportunity exploitation with entrepreneurs. Given the importance of start-up teams to the success of start-up ventures, we believe that understanding these questions can be quite useful. After all, before employees make any effort to help entrepreneurs develop ventures and create value, employees first need to form a positive attitude about the exploitation of entrepreneurial opportunities. The factors influencing employees’ attitude and intention could be different from those influencing entrepreneurs’, and it can be important and interesting to compare and understand these differences.
Second, our integrative model highlights that entrepreneurs’ individual characteristics are important during the entire entrepreneurial process. For example, start-up experience not only can increase the likelihood of the exploitation of entrepreneurial opportunities but also results in better venture performance (Shane, 2003). However, while unemployment increases the likelihood of entrepreneurial activities, ventures founded by people after becoming unemployed are less likely to survive than those founded due to other reasons (e.g., a sound business plan; Shane, 2003). Therefore, it is very important to adopt a holistic perspective on entrepreneurs’ individual characteristics, and it is not enough to exclusively focus on the effect of those characteristics on the pre-launch stage. Future studies can examine the effect of more individual characteristics of entrepreneurs on various stages of the entrepreneurial process. If one characteristic can facilitate the exploitation of opportunities but decrease venture performance, that characteristic may be less desired.
Third, our integrative model also highlights that we should adopt a holistic perspective on support from other agencies. This support should not only aim to increase people’s likelihood of exploiting entrepreneurial opportunities but also help increase venture performance. After all, the ultimate purpose of these programs is to assist the launch and growth of entrepreneurial ventures (Chrisman et al., 2002). Therefore, future studies can assess the role of different programs in various stages of the entrepreneurial process, to examine if the support from these programs indeed helps ventures grow and survive. Programs founded to simply increase the likelihood of exploiting opportunities but not help ventures grow may need to be revised and changed accordingly.
Fourth, having recognised the importance of entrepreneurs’ individual characteristics in the entire entrepreneurial process, we also wonder if investors’ individual characteristics also have an important role in the whole investment process. In other words, do any individual characteristics differentiate investors who get higher returns from the investment from those who do not? If so, do these characteristics influence the whole process of investment from pre-investment decisions to post-investment activities and post-investment outcomes, or do the characteristics just impact one of these stages? If some individual characteristic can influence investors’ investment decisions, but does not influence their returns, then that individual characteristic may be less interesting than one that influences investment decisions as well as increasing investment returns. If we can find additional relevant investors’ characteristics, then we may in turn use them as dependent variables and understand what factors (e.g., environment) influence these relevant individual characteristics.
Fifth, most dependent variables focusing on investors deal with venture capitalists, and few examine other types of investors, such as ‘friends and relatives’, angel investors or bankers. While our results can be limited by the outlets reviewed, those results nevertheless indicate that the entrepreneurship literature has paid less attention to other forms of investors. Venkataraman (1997) has recognised that the investment strategies are quite different between venture capitalists and business angels, for example. While venture capitalists are portfolio investors looking for projects that can enhance potentially higher portfolio returns, business angels are not portfolio investors and typically have a lower capacity to undertake highly risky projects (Venkataraman, 1997). Therefore, the process of decision making for business angels is quite different from that for venture capitalists. Parallel to the studies focusing on venture capitalists, future studies may also want to examine the whole investment process of business angels and/or angel groups. Further, it can also be interesting to examine venture capitalists, business angels, as well as other kinds of investors in other countries, to compare the differences across countries.
Limitations and Future Studies
In addition to other limitations already noted, one might also ask why we did not choose a meta-analysis approach. Because the phenomena and research questions are quite diverse across the entrepreneurship literature, it is unlikely that meta-analysis will be appropriate for examining the entrepreneurship literature in totality (Tranfield et al., 2003). On the other hand, meta-analysis can be quite helpful to answer the question of whether a relationship exists (e.g., between a certain strategy and venture performance). Future studies may want to examine other parts of our integrative model by focusing on one or two categories of dependent variables and conducting a meta-analysis. For example, given that the strategy of supply chain integration is an important factor in influencing the success of ventures, researchers can examine what the factors influencing ventures to adopt supply chain integration are (i.e., focusing on one category: strategy and practices), or what kinds of resources influencing ventures to adopt supply chain integration are (i.e., focusing on two categories: resource and strategy and practices). When a sufficient number of studies have accumulated, meta-analysis can be quite helpful to answer these kinds of research questions. Here our integrative model can help researchers choose their focus and answer important questions for the entrepreneurship literature. Besides, future studies can use other statistical tools (e.g., Latent Semantic Analysis) to assess the progress and the core components of the field.
Conclusions
In this article, we classify dependent variables and develop an integrative model of the entrepreneurship literature following an interpretative approach to review previous entrepreneurship research. The taxonomy and model developed in this article can be useful in guiding future research in entrepreneurship. Our study makes several important contributions. First, our model provides a comprehensive picture of the field as a whole. Second, we take a rich but unwieldy body of work and attempt to make it more understandable. Third, our results can help researchers identify the areas where researchers have paid much attention so that latter studies can build upon previous work, which encourages the creation of a cumulative tradition in the field. Finally, our work can also help researchers identify the gaps where more effort is needed. While our work has its limitations and is not intended to be the definitive literature review on entrepreneurship, we hope that it at least offers a useful piece to the entrepreneurship puzzle.
