THE TEMPORAL PATTERNS OF VENTURE CREATION PROCESS: AN EXPLORATORY STUDY
Jianwen
Liao, Northeastern Illinois University
Harold Welsch, DePaul University
ABSTRACT
INTRODUCTION
LITERATURE REVIEW
RESEARCH METHODOLOGY
RESULTS AND PROPOSITIONS
CONCLUSIONS AND IMPLICATIONS
CONTACT
REFERENCES
TABLE 1
FIGURE 1
FIGURE 2
This study took a grounded theorizing approach in exploring the temporal patterns of venture creation process. Data mining technique was employed to analyze a data set of 669 nascent entrepreneurs. Our results suggested that firm gestation is a complex, non-linear process, more than simple, unitary accumulation of sequential events, in which the developmental stages were hardly identifiable. Firm gestation also appears to be a time-based pacing process in which activities are playing differentiating roles in affecting. Implications and future research directions were provided.
The creation of new organizations has increasingly become the focus of entrepreneurship research (Oviatt & McDougall, 1994; Gartner, 1990). Studies have attempted to explain the creation of new ventures from an array of theoretical perspectives, such as economics (Brenner, 1987), psychology (Katz, 1992; Bird; 1992) and population ecology (Aldrich, 1990), to name just a few. However, despite the growing literature on this area, there were few studies exploring the venture creation process. Empirical studies of the process are even rare, with the exceptions of Bhave (1994), Reynolds and Miller (1992), Carter, Gartner and Reynolds (1996). Scant attention has been paid to the temporal patterns of the process.
The venture creation process refers to the temporal sequence of events or activities that occur as entrepreneurs create a new business. Other terms with the same connotation include firm gestation (Reynolds & Miller, 1992), organizational emergence (Gartner, Bird & Starr, 1992), preorganization (Katz & Gartner, 1988), and start-up (Vesper, 1990). Existing research that explores the venture creation process has assumed a linear, unitary process, which begins with the recognition of a business opportunity and culminates with first sales and first hires. The linear model implies that an additive combination of events will lead to the creation of a new firm (i.e., Reynolds & Miller, 1992; Carter, Gartner & Reynolds, 1996). Yet, there is little empirical evidence to validate or falsify the linear model.
This current study, using an inductive rather than deductive approach in theory building, aims to generate a grounded process model of firm gestation process by linking conceptual categories and sub-processes in the venture creation process identified from PSED data. Glaser and Strauss (1967: 6) defined a grounded theory approach as “generating a theory from data means that most hypotheses and concepts not only come from the data, but are systematically worked out in relation to the data during the course of the research.” For example, Carter, Gartner & Reynolds (1996) did not impose a priori theory onto the data, but analyzed the presence, sequence, and time frame associated with major activities and events in the venture creation process of nascent entrepreneurs. The study followed this approach by specifically addressing this research questions: what are the association and temporal patterns of the events and activities that occur in the venture creation process. Here, association is referred to as the occurrence of one event associated with the presence of another one, regardless of timing. The temporal pattern is referred to as the sequences of closed associated events—what occur first, and what follows.
Consistent with Van de Ven & Poole’s (1989) approach in developing a grounded theory of entrepreneurial process, we defined constructs, found indicators, developed propositions about the venture creation processes using appropriate methods for sequencing analysis. It represents an initial effort in examining venture creation process empirically using a national survey and interview dataset. This paper was structured as follows. A literature review of major theories in venture creation process was provided and major issues related to the current research were identified. The next section delineated a data mining methodology, including the reports of descriptive statistics, association analysis and sequencing analysis across two groups—both nascent entrepreneurs with first sales and without first sales. Results from data mining were reported in the third section. The paper concluded with discussion, implications as well as future research opportunities.
A number of scholars have offered frameworks for exploring the characteristics of the firm gestation process. However, research in the venture creation process remains descriptive and conceptual. Much of the rest of the literature continues to treat the venture creation process as a linear cause-and-effect phenomenon with an emphasis on several components. For example, Katz and Gartner (1988) suggested four emerging properties that indicate an organization is in the process of coming into existence. These properties include the intention to gather information toward the creation of an organization, efforts in boundary establishment activities that distinguish the venture from the rest of the world (such as incorporation, partnership/management agreements, the establishment of physical offices and a phone line), the acquisition of necessary financial resources that are needed to operate an entity (including paying rent, phone bill, equipment), and finally the exchanges with external suppliers, customers, culminating with initial sales and/or initial hiring. Van de Ven, Angel and Poole (1989) also suggested that researchers explore the business creation process by looking at “1) how a business idea emerges over time, 2) when and how different functional competencies are created to develop and market the first proprietary product, 3) when and how these functional competencies are redeployed, and 4) how these business development efforts both influence and are constrained by organization and industry contexts” (pp. 224–225).
Other theoretical models include a life cycle view of the venture creation process, largely based on observations of high-tech venture founding. For example, Galbraith (1982) proposed a stage of development model. In his model, he described four stages: 1) a proof-of-principle stage in which the hopeful entrepreneurs with an idea face the basic task of developing some proprietary technology; 2) A prototype stage in which entrepreneurs turn proprietary technology into a prototype; 3) a model shop stage in which a number of models are produced and tested; 4) The startup stage in which formal production begins and the firm makes its first sales. In a similar vein, Churchill and Lewis (1983) proposed a more detailed, five-stage model. A number of subsequent empirical studies provided credence to the stage model (i.e., Roure & Keeley, 1990; Kazanjain & Drazin, 1990). For example, Bhave’s (1994) case study divided the process into three stages: definition of business concept, creation and set up of production technology and exchange of product, which provided additional support to the stage model. However, on a large sample study (n > 3000) of the sequence and timing of four key events (commitment, first hire, first sale, first financing), Reynolds and Miller (1992) found substantial variations in the number of events reported, the sequence of events, and the amount of time between events, providing little evidence to support a stage model theory.
Incidentally, most of these theories on venture creation process have used linear models, which imply that an additive combination of specific events will culminate in the start-up of a new business (Reynolds & Miller, 1992; Carter, Gartner, Reynolds, 1996). Muck like adding weight to a scale, a linear additive model suggests that the eventual accumulation of enough activities would eventually tip the scale in favor of the creation of a new organization.
A common practice in empirically testing the stage model has been to identify the activities during the venture creation process and then dividing the events into stages based on a priori stage model, while ignoring the timing when the events occur. There are two potential drawbacks to this method. First, there may exist a high probability that two events occur in completely different time frames but are assigned into the same stage due to conceptual similarities. Second, the use of a priori theory may create a self-fulfilling prophecy in testing a stage model. Empirically grounded, sequential models of development process are noticeably absent (Van de Ven, 1992).
More recently, complexity theory is offered new ways to conceptualizing nonlinear processes in dynamic systems like nascent start-ups. For example, Bygrave (1993) suggests the use of deterministic chaos theory to understand organizational emergence. Stevenson and Harmeling (1990: 2) argued that a more “chaotic” theory of entrepreneurship would more accurately “illuminate the dynamism and the complexity of real organizations creating and adapting to change.” Similarly, in his 27 interviews with entrepreneurs, Bhave (1994) described the venture creation as an iterative, nonlinear, feedback driven process.
Despite the growing literature on process-related entrepreneurship research, there exist few empirical studies exploring and testing these linear and especially nonlinear models. Empirical studies of the Katz and Gartner (1988) framework by Reynolds and Miller (1992), and Reynolds (1994) failed to uncover a common pattern or sequence of events for all emerging organizations. Most of research studies so far have been descriptive accounts of the venture creation process, as well as qualitative case studies. Several major factors may contribute to the lack of empirical studies in this area. First, process-related research requires large-scale, longitudinal data, which is usually time consuming and costly to create. Second, traditional statistic modeling techniques are not well equipped to detect temporal patterns from time-based as well as event-based variables. Third, traditional methodologies emphasize “variance theories” which examine interrelationships among variables measured more or less at one point in time (Mohr, 1982).
Dataset Description
The data for this study was obtained from the Panel Study of Entrepreneurial Dynamics (PSED). The PSED is a longitudinal data set of individuals in the process of starting businesses who were identified from a random digit dialing telephone survey of 64,622 adults in the United States who are 18 years of age, or older. A nascent entrepreneur is identified if he/she answered yes to the following two questions: 1) Are you, alone or with others, now trying to start a new business or starting a new business for your employer? 2) Are you, alone or with others, now starting a new business or new venture for your employer? Is the effort a part of your job assignment? All of these individuals were considered candidates for the nascent entrepreneur interview if they met three additional criteria. First, they expected to be owners or part owners of the new firm. Second, they had been active in trying to start the new firm in the past 12 months. Third, the effort was still in the startup or gestation phase and was NOT an infant firm. Follow-up surveys were conducted at 12 months intervals to evaluate the status of the start-up effort. Data related to nascent entrepreneurs were collected using a combination of survey and phone interviews. Survey questionnaires included items related to opportunity recognition, entrepreneurial climate, start-up problems, start-up context, reasons for starting a new venture as well as nascent entrepreneurs’ demographics, background and personal dispositions. Phone interview questions were concerned with the nature of start-up, start-up activities, start-up team, start-up funding requirements, future expectations for the new business, personal decision making style, market and competition assessment as well as nascent entrepreneurs’ social network. The data set consists of 669 nascent entrepreneurs. A more detailed description on the background and methodology of the PESD data set can be found in Reynolds (2000).
Measures
Start up Activities. We followed Van de Ven’s and Poole’s (1989) methodological approach. One method we used is to code an entrepreneur’s chronological listing of qualitative events with dichotomous indicators. Data related to startup activities came from the interview dataset. Nascent entrepreneurs interviewed were asked to respond to a list of 26 startup activities—1) if an activity was present or not? 2) In which year and month/season did the activity/event occur? A sample question was: has a business plan been prepared? In what year and month did work on business plan begin? A list of startup activities with coding names can be requested from the authors.
Therefore, a dichotomous indicator was used, with 1 to represent the presence and 0 the absence of a certain informative features of the qualitative event in the venture creation process. For each event, there was a timestamp including years and months when the event occurred. For those who couldn’t specifically remember the exact months of the event occurrence, the choices of spring, summer, winter and fall were offered. We then recoded spring as March, summer as June, fall as September, and winter as December.
Gestation Period. In their study of firm gestation in Minnesota and Pennsylvania, Reynolds and Miller (1992) found that not all events occurred in the process and every possible sequence of events was present. In the present study, there is substantial variation in length of the gestation period. Following Reynolds and Miller (1992), the time that elapses from the first event to the last event was considered the gestation period, regardless of the nature of the events. The duration of the venture creation process was calculated by the following steps: 1) the earliest year and the latest year among all the activities engaged by an entrepreneur during the venture creation process were identified; 2) we converted the year differences between the earliest and the latest into months for each informant; 3) We combined the converted months with the month when the event occurred, which yielded the duration for the gestation period in months.
Number of Events. Similarly to the approached employed by Reynolds and Miller (1992), we counted the number activities/events engaged by entrepreneurs during the start-up process.
Timing of Each Startup Event. For each informant, we subtracted the earliest years (starting years) from each times tamp. We then converted the years into months, adding the months when the activity started. Therefore, the final dataset had all the activities engaged by entrepreneurs coded in bitmap format, with the timestamp in the form of months.
Completion of Venture Creation Process. According to Reynolds & Miller (1992), there are multiple criteria related to the concept of emergence of new firms. One is to assume that a new firm is not part of the economy until sales begin or first hire begins. From a legal perspective, new firms come to exist when they register with government agencies and receive tax identification. They suggested that a firm is considered fully established, when 1) Personal commitment: when did members of the start-up team first begin to make major investments (personal time, personal resource) in the new firm; 2) Financial support—when was the first outside financial support obtained; 3) Sales—when did the firm receive its first sales income? 4) Hiring—when did the firm hire anybody—full or part time? Hansen & Wortman (1989) and Hansen (1991) also have proposed that the first commercial sale of a product or service marks the end of the pre-organization stage and signals the emergence of the new organization. Researchers have argued that the first sale is a significant and last-step event in the physical creation of a new venture creation (Bhave, 1994; Block and MacMillan, 1985) for the following reasons: 1) first sales is the materialization of an business opportunity. It vindicates the business concept; 2) the first sales is the starting point of customer feedback, which would become feedback into the venture’s future direction. Based on these rationales, we chose first sales (Q162) as our key indicator of firm birth and the completion of venturing process. Within our sample of 668 entrepreneurs, we divided the sample into two sub-samples—“started” group consisting of 292 entrepreneurs with first sales, and “trying” group consisting of 376 entrepreneurs with no sales.
Modeling
Two data mining modeling nodes, directed web node and sequencing node within Clementineâ were employed.
First, we used directed web node to identify the strength of relationships among different venture creation activities using the “started” group—a dataset of 292 entrepreneurs with first sales. We were particularly interested in finding out what activities, regardless their timing of occurrence, that were closely connected with first sales (Q162). Directed web graphs would show unidirectional connections from all other start up activities to the first sales (Q162). We set up the directed web node in the following ways: 1) Weak connections below 20%—the threshold for weak connections (dotted lines); 2) Strong connections above 40%—the threshold for strong connections (heavy lines); 3) Show only connections of and above 3.
Second, we employed the sequence node to discover the sequential patterns of startup activities for the “started” entrepreneurs. We set up the sequence node by the following parameters:
Compared with traditional variance based methodology, data mining has a few unique advantages in analyzing a venture creation process. First, it does not impose a theory and presume a unitary sequence or stage model. It examines sequencing patterns in a detailed and systematic way. Second, it does not assume the events are always organized. In reality, nascent behaviors are disorganized. The current data mining technique provide an alternative to detect the patterns and non-coherency, if any, within the seemingly chaotic activities related to firm gestation.
Descriptive Statistics of Gestation Period: “Started”
Results from descriptive statistics indicated that the median of the duration of firm gestation period is 32, which is close to 2 years and 8 months, and the mean is 75.57 months, which is slightly above 6 years. One in five (20%) report that it took about 12 months (one year), two in five (40%) within 24 months (2 years), three in five (60%) in 48 months (4 years) for the startup duration. Overall, the start up duration reported here appears to be much longer than reported by Reynolds and Miller (1992). In their analysis of entrepreneurs in the states of Minnesota and Pennsylvania, they found that four of five (80%) of all new firms go through gestation within two years (24 months), nine of ten (90%) within three years.
Number of Start-up Events: Started vs Trying
On average, nascent entrepreneurs engaged in 10 activities. In terms of percentiles, 40% of nascent entrepreneurs engaged in 9 activities, 60% with 11 activities, and 80% of them with 14 activities. The distribution of the number of activities by nascent entrepreneurs is close to a normal distribution, with Skewness of .376 and Kurtosis of -.142.
We employed Analysis of Variance (ANOVA) to test the mean difference in the number of activities between nascent entrepreneurs with first sales and those without first sales. In our sample of 668 nascent entrepreneurs, 292 have first sales and the rest 376 do not have first sales. The averaged number of activities for the first sale group is 13.101, and 8.66 for nascent entrepreneurs with no sales. ANOVA test for the two groups indicated the differences between the two groups in the number of start-up activities were statistically significant (F = 267.715, p < .001).
These results suggested that “started” entrepreneurs with first sale as compared with “trying” entrepreneurs with no sales were more active and engaged in significantly more activities during the firm gestation process. The accumulation of activities seems to lead to a more successful completion of firm gestation, which culminated in first sales. However, the results should be interpreted with caution since we have limited information of the “trying” entrepreneurs. The low number of activities may well be due to the fact that many of them were at the starting point of venture creation. Therefore, it may not necessarily be an indicator of inactivity, or lack of effort. The inconclusiveness of these findings warrant further empirical testing. Based on these results and discussions, we proposed:
Proposition 1: “Started” entrepreneurs would engage in more activities than “trying” entrepreneurs.
Association Patterns of “Started” Entrepreneurs
What activities were related to first sales? Results from data mining web node indicated that the presence of six startup activities was directly associated with first sales (Q162). However, we failed to uncover any strong connections and all the connections were weak (<20%). In the sequence of strength of connections, these activities included investing your own money in this business (Q143), defining market opportunities (Q134), purchasing raw materials and, inventory and suppliers (Q128), starting marketing and promotional efforts (Q122), spending time on thinking about business ideas (Q109), and developing models and procedures (Q120). These activities were clearly fallen within the two of the properties for firm emergence by Katz and Gartner (1988), intentionality and resource acquisition. The association patterns between first sales and other start-up activities were summarized in Figure 1.
Also noted in Figure 1, the absence of five start-up activities was also associated with the presence of first sales (Q162). These activities included application for a patent, copyright and/or trademark (Q124), hiring employees and/or managers (Q155), paying state unemployment insurance tax (Q1735), paying federal social security taxes (Q177), as well as listing with Dun & Bradstreet. These activities appear to be more closely related to defining boundary as defined by Katz and Gartner (1988).
These findings imply differentiating roles played by different activities in relating to firm emergence. Notably, activities related to intention to create a business and acquisition of resources seems to be more closely associated with first sales (exchanges) than defining boundary. Based on our findings and in relating to Katz and Gartner’s (1988) four properties of firm emergence, intentionality, boundary, resources and exchange, we proposed:
Proposition 2: In a venture creation process, internationality and resources are more closely related to exchanges than boundary.
Sequencing Patterns of Start-up Activities
Results from sequencing analysis were reported in Table 1. Overall, our sequencing analysis yielded only single-activity sets, instead of sets with multiple startup activities. Since we set a minimum support level of 15% and timestamp tolerance of 6 months, it meant that there were no more than 15% of nascent entrepreneurs who engaged in two or more startup activities within a six month time frame. When considering that entrepreneurs in our sample, on average, engaged in 10 activities and took 32 months to complete firm gestation, it implied a deliberate pacing-based venture creation process. Bird (1992) referred pacing to “the speed with which the organizing events occur and reflects the time between bracketed events as well as the timing within those brackets (p. 16).”
Researchers have speculated a few reasons for the occurrence of pacing. First, entrepreneurs use pacing to shift attention from being externally focused on sales, markets, or technology to being internally focused on costs, employees or production processes (Gersick, 1994). Second, pacing serves to manage uncertainty and risk by bounding the period of waiting for results and feedback (Bird, 1992). Third, pacing also serves as a structure for organizational learning based on trial and error (Bird, 1992). Gersick (1994) found that time-based pacing occurs when leaders 1) have a serious deadline, 2) have control over their actions, and are uncertain about the outcomes to reach for and how to reach those outcomes. The last two conditions were applicable to nascent entrepreneurs who were in the process of creating a venture. Based on the results and discussion, we therefore proposed:
Propositions 3: Firm gestation is a time-based pacing process.
Lack of sets with multiple activities provided little support for the stage-based theory. A stage (which is called item sets in Clementine) is referred to as a meaningful set of co-occurring activities (items) serves in the gestation process. A clearly identifiable phase would follow the structure as: (a, b. c. d) è (e, f, g, h, i) è (j. k, l, m. n, o) è (p, q, r, s, t). Activities within parentheses means they are closely related and co-occurring. We therefore used the following two criteria to detect the existence of stages in our data: 1) a cluster of events/activities occurring in reasonably close time frame to satisfy the definition of co-occurring. 2) Interpretable sets of events/activities, that point to unique theoretical constructs in the venture development process. Holding supporting level (15%) constant, we varied the time frame of co-occurring from 6 months, 8 months to 12 months. But none of these variations yielded identifiable stages. We therefore argued:
Proposition 4: Firm gestation is a process where developmental stages are hardly identifiable.
As indicated in Table 1 and also illustrated in Figure 2, we found the following sequencing patterns in our data set:
As indicated in Figure 2, we identified some temporal patterns. These sequences illustrate the developmental trajectories of particular paths of venture creation. In light of with Ven de Ven and Poole (1989), several unique temporal patterns emerged:
Overall, our results suggested that venture creation processes are exceedingly more complex than we presumed. It is more than an orderly, unitary, and progressive path that consists of the addition of a series of events that culminate in a firm’s gestation. The convergent and divergent paths imply a non-linear dynamic theory of venture creation process. Previous research always assumes a simple, unitary progression model of venture creation that characterizes most of the stage models of venture creation. However, the results from data mining indicate a much more complex and fluid process. Based on the these discussion, we proposed:
Proposition 5. Firm gestation is a complex process that includes more than simple, unitary progressive paths.
Using a grounded theorizing approach, this study found that firm gestation a complex, non-linear process, more than simple, unitary accumulation of sequential events, in which the developmental stages were hardly identifiable. It also appears to be a time-based pacing process in which activities are playing differentiating roles in affecting firm formation.
Implications and Directions for Research
The study made several theoretical and raised methodological contributions to the field of entrepreneurial process research, and also has important implications for nascent entrepreneurs. From a theoretical standpoint, the study falsified traditional stage models and other additive, sequential models of venture creation. It examined different attributes associated with a process, and laid the foundation for building a more integrated, parsimonious process theory. Additionally, the project is among the first studies to explore a pattern recognition methodology—data mining, which has significant applications in process-related research. From a practical standpoint, findings from this study have significant relevance for government policy makers as well as nascent entrepreneurs. For policy makers, they would consider designing programs, policies, incentives that would have the greatest impact on the processes that would lead to the creation of entrepreneurial firms. For nascent entrepreneurs, they should probably invest their own money and spend time in defining market opportunities—two activities that were closely associated with first sales—a key milestone in venture creation process.
This study also generated several future research opportunities, especially in the area of how the contexts in which the entrepreneurial process takes place influence the firm gestation process. These contexts may include entrepreneurs’ growth aspiration and the nature of startups, to name just a few. Specifically future research efforts may be directed to: entrepreneurial growth aspiration, for example. Some nascent entrepreneurs may expect to start firms that grow rapidly compared with other nascent entrepreneurs. As a result, we may expect process difference in terms of sequence, time, as well as activities involved. Second, focus may be directed to types of new ventures, i.e., tech and non-tech-based ventures, as well as industry characteristics may also have impact on the venture creation process. Third, McCarthy, Krueger and Schoenecker (1990) found that entrepreneurs spend their time differently in earliest stages than in later stages of venture development. Therefore, an interesting question that is of considerable interest is designed to identify the specifics of time allocation during organization creation. Fourth, a number of start-up behaviors were not explored in the PSED data, for example, seeking advice from mentors and advisors. It may be appropriate to develop measures of these activities. The addition of these activities may have influence on the results of current study. Fifth, the sequencing analysis we employed in this study may be particularly useful in comparing those nascent entrepreneurs who in fact complete the creation of new business with those who “die off.” These should be the central research questions for future entrepreneurial processes. The trajectories of the processes associated with the two groups would shed light on what processes lead to differences in results. Sixth, additional efforts should be devoted to exploring the sub-processes, such as addition, substitution, modification, inclusion, mediation (Ven de Van & Poole, 1989). These five types of sub-processes may serve as useful building blocks to examine the developmental sequence and progression of venturing process over time.
Limitations of Current Study
The current study has a few caveats and the results need to be interpreted with caution. The current study of examining entrepreneurial process took a retrospective approach, conducted after activities had occurred. Nascent entrepreneurs in the PSED were asked to recall the time when the events took place—a typical retrospective case histories conducted after the outcome, or part of outcome, of venture creation process were known. However, it is widely recognized that prior knowledge of the success or failure of an entrepreneurial process invariably biases the recall of events and the subsequent research findings. Though concerned efforts can be taken to minimize the bias, it is generally preferable to record the events within a reasonable time frame when the events occur—especially before the outcomes of the events become known. Currently, PSED is following up with nascent entrepreneurs and tracking down changes taking place every year. Therefore the bias issue in the dataset will dwindle down.
CONTACT: Jianwen Liao, Department of Management and Marketing, Northeastern Illinois University, 5500 St. Louis Ave, Chicago, IL 60625; (T) 773-442-6136; (F) 773-442-4490; j-liao@neiu.edu
Aldrich, H. E. (1990) Using an Ecological Perspective to Study Organizational Founding Rates. Entrepreneurship Theory and Practices. 14(3): 7–24.
Bhave, M. (1994) A Process Model of Entrepreneurial Venture Creation. Journal of Business Venturing. 9: 223–242.
Bird, B. (1992) The Operation of Intention in Time: the Emergence of the New Venture. Entrepreneurship Theory and Practice. 17(1): 11–20.
Bird, B. (1992) The Operation of Intentions in Time: The Emergence of the New Venture. Entrepreneurship Theory and Practice. 17(1): 11–20.
Block, Z. and MacMillan, I.C. (1985) Milestones for Successful Venture Planning. Harvard Business Review. 14(4): 496–515.
Carter, N., Gartner, W. & Reynolds, P. (1996) Exploring Start-up Event Sequences. Journal of Business Venturing. 11: 151–166.
Churchill, N.C. and Lewis, V.L (1983) the Five Stages of Small Business Growth. Harvard Business Review. May–June: 30–50.
Galbraith, J. (1982) the Stage of Growth. Journal of Business Strategy. 3(1): 70–79.
Gartner, W. (1990)What Are We Talking When We Talking about Entrepreneurship? Journal of Business Venturing. 5(1): 15–28.
Gersick, C. (1994) Pacing Strategic Change: the Case of a New Venture. Academy of Management Journal. 37: 9–45.
Hansen, E. & Wortman, M. (1989) Entrepreneurial Networks: the Organization in Vitro: Academy of Management Best Paper Proceedings. 49th Annual Meeting. Washington DC. 69–73.
Hansen. E. (1991) Structure and Process in Entrepreneurial Networks as Partial Determinants of Initial Venture Growth. In R. Ronsad, N. Churchill, W. Bygrave, D. Sexton, D. Slevin, K. Vesper and W. Wetzel (eds) Frontier of Entrepreneurship Research. (pp. 320–334), Wellesley, MA: Babson College.
Katz, J. & Gartner, W. (1988) Properties of Emerging Organizations. Academy of Management Review. 13(3): 429–442.
Katz, J. (1992) A Psychological Cognitive Model of Employment Status Choice. Entrepreneurship Theory and Practice. 17(1): 29–37.
Kazanjain, R. & Drazin, R. (1990) A Stage Contingent Model of Design and Growth for Technology Based New Ventures. Journal of Business Venturing 5: 137–150.
Kazanjan, R. K. (1988) Relation of Dominant Problems to Stages of Growth in Technology-based New Venture. Academy of Management Journal. 31(2).
Larson, A and Starr, J. A.(1993) A Network Model of Organization Formation. Entrepreneurship: Theory and Practice. 17(2): 5–16.
McCarthy, A. M., Krueger, D.A. and Schoenecker, T.S. (1990) Changes in Time Allocation Patterns of Entrepreneurs. Entrepreneurship: Theory and Practice. 15(2): 7–18.
Mohr, L. B. (1982) Explaining Organizational Behavior: the Limits and Possibilities of Theory and Research. San Francisco: Jossey-Bass.
Oviatt, B. & McDougall, P. (1994) Toward a Theory of International New Ventures. Journal of International Business Studies. 25(1): 45–64.
Pettigrew, A. (1985) the Awakening Giant: Continuity and Change in ICI. Oxford: Basil Blackwell.
Reynolds, P. (2000) National Panel Study of U.S. Business Startups: Background and Methodology. In Databases for the Study of Entrepreneurship. Vol 4: 153–227. Greenwich, CT: JAI Press/Elsevier.
Reynolds, P. & Miller, B. (1992) New Firm Gestation: Conception, Birth, and Implications for Research. Journal of Business Venturing. 7. 405–417.
Roure, J. & Keeley, R. (1990) Predictors of Success in New Technology Based Ventures. Journal of Business Venturing. 5: 201–220.
Van de Ven & Poole (1989) Methods for Studying Innovation Processes. In A.Van De Ven, H. Angle, M. S. Poole, (ed) Research on the Management of Innovation: The Minnesota Studies, New York: Oxford University Press.
Van de Ven A. & Poole, M.S. (1988) Paradoxical Requirements for a Theory of Organizational Change. In R. Quinn and K. Cameron (eds) Paradox and Transformation: Toward a Theory of Change in Organization and Management. Cambridge, MA: Ballinger.
Van de Ven, A. (1992) Longitudinal Methods for Studying the Process of Entrepreneurship. Journal of Business Venturing. 8: 211–230 .
Van de Ven, A., Angle, H.L. & Poole, M. (1989) Research on the Management of Innovation. New York: Harper and Row.
Vesper, (1990) New Venture Strategies. (2nd edition) Englewood Cliffs, NJ: Prentice-Hall.
2002 Babson College. All Rights Reserved. Last Updated March 2003.