Attributions and success in new venture creation among Canadian nascent entrepreneurs.

Diochon, Monica ; Menzies, Teresa V. ; Gasse, Yvon 等

The question of why some new ventures are successfully brought to fruition while others are not is one that continues to elude researchers. Generally, we know that the rate of firm failure is inversely related to age (Thornhill and Amit, 2003; Sapienza, 1991). While Beaver (2003) claims that failure rates among new businesses are ten times as high as those of larger established firms, other estimates of failure range anywhere from 50 to 95%. Not surprisingly, the high rate of failure among small businesses, particularly new ones, has been a concern among policymakers for some time (Storey 1994). Over the past decade, the importance of this concern has increased as policymakers focus their efforts on encouraging new firm formation.

In light of the relationship between a firm's age and failure, nascent ventures--those in the gestation stage of start-up--would be expected to have the highest rate of demise. Yet, gestation is an aspect of the entrepreneurial process about which very little is known (Gartner et al., 2004). One of the major stumbling blocks in acquiring more knowledge of business formation has been the lack of "real time" study of the start-up process. Currently, the Entrepreneurial Research Consortium (ERC)--comprised of researchers in nine countries--is addressing this issue through a research project commonly referred to as the Panel Study of Entrepreneurial Dynamics (PSED) (for further details on the consortium see Gartner et al., 2004).

Given how little is known about the factors influencing firm formation, there is a pressing need for research. Since entrepreneurs have been shown to differ from other people with regard to certain aspects of cognition (Baron, 1998; Palich and Bagby, 1995; Kahneman and Lovallo, 1993), specifically, this paper explores whether individual differences in attributional style are associated with later success in bringing a business to fruition. Drawing on the attributional framework presented by Shaver (2004), we investigate this issue using four years of real time data collected from Canadian nascent entrepreneurs. The results of this investigation should be valuable to policymakers. Without a better understanding of firm formation, how can policy measures aimed at encouraging new firm formation be effective? Prospective entrepreneurs, too, have a vested interest in knowing more about what contributes to "success" and "failure."

The paper begins by reviewing what we know about success and failure in a start-up context. It then explores the usefulness of attribution theory in explaining events and behaviour. Next it assesses the conceptual framework proposed by Shaver (2004) as a means of predicting the success and failure of nascent entrepreneurs in establishing new ventures. Hypotheses are proposed. The method section describes the research design, sampling and data collection process as well as the measures and analysis employed. Results are then presented and discussed.

Literature Review

New Venture "Success" and "Failure": What Does it Mean?

Researchers have been attempting to explain the "success" and "failure" of new ventures for some time. However, investigations have been criticized for focusing largely on post-founding activities, for focusing only on those who succeeded in forming a business, and for hindsight bias (recollections of pre-founding). Definitions, or lack thereof, also have been problematic (see Diochon, Menzies and Gasse, 2005a for further discussion).

"Success" is a term that means different things to different people (Rogoff, Lee and Suh, 2004; Beaver, 2003). Entrepreneurs are likely to define success differently than the venture's external stakeholders, such as lenders or customers. For example, one entrepreneur might define success in terms of whether the business can generate an income of $25,000, another might define it according to whether (s)he achieves personal fulfilment, while a lender might define it according to whether the business can service its debt. "Failure," being the antonym of "success," also would mean different things to different people.

In the literature, many terms are used synonymously with both "success" ("survival" or "continued viability") and "failure" ("closure" or "bankruptcy") even though they do not mean the same thing (Rogoff, Lee and Suh, 2004; Beaver, 2003). Indeed, a business can be abandoned without failing such as when an owner becomes ill. Given the wide range of meanings employed, it is not surprising that the literature has come under attack for being full of misinformation or for making inferences that are unjustified (Stokes and Blackburn, 2002).

Within the context of start-up, a central and unresolved issue is that of defining success in forming a business. Much of the difficulty in determining the point when the transition is made from becoming to being a new business can be traced to the fact that firm founding is a process, rather than a discrete event (Diochon, Menzies and Gasse 2005a; Carter et al. 2004). Yet, without being able to determine that something did or did not happen (ie. that a business is formed), how are explanations possible for why some people succeed (in forming a firm) while others do not?

We contend that what is most problematic about the lack of explicit definition (Carland, Carland and Carland, 1995), is the validity of the measures. For example, while it is common to operationally define a new firm as one that has been incorporated within the past twelve months, this excludes all unincorporated businesses. With the aim of addressing the definitional void and establishing a measurement foundation that is clear and meaningful, we offer the following definition of a new business: a recently formed commercial organization that provides goods and services for sale (Diochon, Menzies and Gasse, 2005a). Arguably, it is important not only for the business to start but to stay in business. Doing so would constitute "success" in forming a firm. We deal with the issue of operationalizing the above definition in the method section.

To address the question of why some people "succeed" in establishing a new firm while others do not, we turn to the psychology literature. People, generally, attempt to develop explanations of their own and others' behaviour (Shaver et al., 2001; Weiner, 1985; Heider, 1958). Broadly, efforts to answer the question "why" are described in the field of social psychology by the term attribution--the processes through which an individual seeks to identify the causes of events, others' behaviour, or practically anything that is encompassed by their experience (see Pittman, 1993).

With the interpretation of a situation often more important as a determinant of future action than the facts (Robbins, Coulter and Stuart-Kotze, 2003; Martocchio and Ford, 1985), the meaning people give to events and behaviour is of heightened significance. Since we know that each person experiences a somewhat different reality, it is important to try and understand what that person's perceived reality is, how it is being interpreted, and the implications for action. One theoretical framework that enables us to do that is attribution theory. In the next section, we review the usefulness of this theory in a new venture context.

Attribution Theory in a New Venture Context

Attribution theory was originally proposed by Heider (1958). With regard to task success, he argued that three things were required: an intention to perform the task, exertion in the direction of the intention, and a personal ability that exceeded the difficulty of the task. Weiner (1985) later represented these elements by two dimensions: locus of causality and stability of the cause. As shown in Table 1, the locus of causality refers to whether the cause(s) of events or behaviour is considered to be internal or external (to the person). The stability of the cause refers to whether the cause is capable of rather immediate change. For example, a person's ability is deemed internal and stable since it does not really change quickly. However, effort is considered internal and variable since the extent to which a person tries can change from one situation to another. Among the major external causes, task difficulty is the relatively stable factor whereas luck is, by definition, variable.

Given attribution theory's usefulness in explaining intentional behaviour (Anderson, 1991), Shaver (2004) has argued it is particularly appropriate in understanding firm formation which involves domain-specific intentional action that is focused on achieving a desired outcome. The activities engaged in during the start-up process reflect a series of choices (Shaver and Scott, 1991). Arguably, the reasons entrepreneurs give for their choices are likely to provide insight into eventual success (or failure) in forming a firm, among other things.

In examining his/her behaviour or an event, a person will try to determine if its "successful" accomplishment (or "failure") was due to personal factors, such as ability or external factors, such as luck; and whether the cause was stable or unstable (Weiner, 1985). Ultimately, the outcome could be attributed to ability, task difficulty, effort, or luck (as illustrated in the table). Whether "success" or "failure" is attributed to internal causes or external causes will have implications for the future (Shaver et al., 1996; Weiner, 1986).

Attribution theory is said to be one of the few psychological theories that can deal with a key characteristic of entrepreneurs--persistence after setbacks or failure (Shaver, 2004). If an entrepreneur attributes the cause of setbacks or failure to external factors (such as size of the market or the number of competitors), there is no reason not to try again. However, if the cause is attributed to internal factors--specifically stable ones such as ability--then the entrepreneur would be unlikely to view starting a second venture as appealing (Cardon and McGrath, 1999). Alternatively, if the cause is an internal variable one, such as insufficient effort or poor strategy, a failure to establish a business need not imply that another attempt will also fail. Rather than feeling helpless in the face of the problem, the entrepreneur is motivated to solve it. In this case, the individual wants to learn from his/her mistakes, and to apply that learning so future problems can be avoided.

According to Shaver (2004), the utility of an attributional approach in explaining current (as opposed to past) activity lies in the fact that there are reliable individual differences in explanatory style (Peterson and Villanova, 1988). Since events like establishing a new venture would have more than one cause, it would be reasonable to expect that people would disagree about what might be the cause. "In such a case, people whose predilections lead them to look for causes internal to people may concentrate on the enduring personal characteristics. Alternatively, people whose predilections lead them to look for changeable external causes are more likely to concentrate on factors outside the person" (Shaver, 2004: 207).

Shaver (2004) argues that individual differences in explanatory style, coupled with the notion that attributions may be related to entrepreneurial persistence, suggest that an entrepreneur's reasons for entering business might be related to later success (or lack thereof) in establishing a new venture.

Like every process, business start-up takes place over time, involving both positive and negative situations. Indeed, it is well known that entrepreneurs face high levels of uncertainty, novelty and time pressure--factors which tend to overload information-processing capacity and increase entrepreneurs' susceptibility to a number of cognitive biases (Forbes, 2005; Baron, 1998). In particular, entrepreneurs may be more susceptible to the self-serving bias than others (Baron, 1998). The self-serving bias refers to a tendency for individuals to offer internal attributions for positive events and provide external attributions for negative events. Indeed, Gartner and Shaver (2002) report evidence of a self-serving bias among US nascent entrepreneurs in the PSED. The evidence showed that nascent entrepreneurs described opportunities as internal and stable while problems were described as external and variable.

This suggests that nascent entrepreneurs would have a tendency to perceive starting a business positively and would be more likely to offer internal (and stable) attributions in describing their reasons for starting a business. On the other hand, problems or setbacks would be perceived negatively, and would be afforded attributions that are external (and variable). Many researchers refer to the "contrasting patterns of attribution (the typical 'self-serving' pattern shown by most people and its opposite) as discrete attributional styles" (Baron, 1998: 284).

Baron (1998) suggests that among entrepreneurs, one factor that might differentiate those who are successful from those who are not is that "successful" entrepreneurs are less susceptible to the self-serving bias. Accordingly, because the thinking of successful entrepreneurs is influenced less by the self-serving bias, these individuals are better able to establish positive relationships necessary to the survival or growth of their business (such as with investors or suppliers). In presenting his argument, Baron cites an extensive body of literature indicating the self-serving bias can be a major source of interpersonal conflict. This conflict arises when individuals who work together become aware of each other's tendency to take credit for positive outcomes and blame the other person for negative results. Baron goes on to argue that successful entrepreneurs have been shown to be more adept than unsuccessful ones at forging positive relationships with their stakeholders. Using this reasoning in a start-up context, we would argue that nascent entrepreneurs who are successful in starting (and sustaining) a business would be less susceptible to the self-serving bias and, therefore, better able to establish the key interpersonal relationships necessary to bring their venture to fruition.

Accepting the notion that a self-serving bias would lead people to offer external attributions for negative events, then those nascent entrepreneurs who report abandoning their efforts to start and stay in business would be expected to provide external attributions for doing so.

Based on the preceding discussion, we propose the following hypotheses:

H1: Nascent entrepreneurs will predominantly offer internal, stable attributions in describing positive situations.

H2: Nascent entrepreneurs will predominantly offer external, variable attributions in describing negative situations.

H3: Nascent entrepreneurs who are successful in starting a business will be less susceptible to the self-serving bias than nascent entrepreneurs who have given up on their efforts to start a business.

H4: Among nascent entrepreneurs who do not persist in their efforts to start a business, the least likely explanation they will provide for giving up will be internal stable reasons.

H5: Nascent entrepreneurs who provide internal stable reasons for giving up will be less likely to view the prospect of starting another venture as appealing than nascent entrepreneurs who provide other reasons for giving up.

Method

Research Design, Sampling, and Data Collection

The data for this article were obtained from a national longitudinal study of business startups being conducted by the Canadian members of the ERC. While the primary objective of the Consortium is to provide systematic, reliable data on the basic features of the startup process, a second objective is to provide reliable data on the factors that would further our understanding of how and why new firms emerge, grow, decline, die or remain stable. To achieve these objectives, the research design had to fulfil a number of criteria including: standardization of procedures (for selecting a representative sample and for developing systematic descriptions of the entrepreneurial process); incorporation of all major perspectives in data collection (so that the direct and interactive influences can be identified); longitudinal design (to track the process); and the inclusion of a wide range of scholarly input for research design and analysis (to deal with the complexity inherent in the project) (see Reynolds, 2000).

Given the scope of the start-up process and the factors influencing it, the questionnaires that were developed consisted largely of fixed-response questions, although a few open-ended questions were asked. The questions dealing with attributions were developed by a fifteen-member team of Consortium scholars. This team was responsible for developing the person variables portion of the research instrument which included nine topics, of which "attributions for success and failure" is one. Moreover, they were responsible for the reliability and validity of the various items and measures used (for further detail on item development, see Gartner et al., 2004).

In accordance with the procedures established by the ERC, we engaged SOM, a national polling firm, to select a representative sample of people engaged in the start-up process (nascent entrepreneurs) in Canada during the winter of 2000. From the initial 49,763 randomly selected telephone numbers, there were 29,855 usable numbers. Using a Computer Assisted Telephone Interviewing System (CATIS) interviews were completed with individuals in 21,116 households.

The initial screening interview produced a sample of 593 as some households had multiple nascent entrepreneurs. Out of these, 463 qualified for the longer telephone interview by answering affirmatively to all three screening questions (1) and 416 agreed to participate in the follow-up interview. In June 2000 detailed data collection began, using phone and mail survey instruments, covering a wide range of topics (which are detailed in Gartner et al., 2004).

At this point, the respondent's status as a nascent entrepreneur was double-checked. This involved determining whether the respondent was still in the gestation phase of the start-up process. Approximately one fifth (19%) of the respondents were dropped at this stage as they were too far advanced to be considered in the gestation phase of the entrepreneurial process. Among the remaining respondents, it was not possible to locate and contact about 24%. Another 17% of the remaining group would not or could not complete the phone interview.

A total of 151 verified and accessible nascent entrepreneurs completed the initial detailed telephone interview. This represents a 59% response rate of those that could be contacted and were eligible. Follow-up interviews were completed at 12-month intervals: 2001 (n=132), 2002 (n=119), 2003 (n=104), and 2004 (n= 91). The follow-up interviews enabled us to track activities and determine whether a firm was founded.

In this article we analyze attributional data from the detailed phone interviews along with data on the venture's start-up status. A venture that achieved operating status in any year of follow-up and maintained that status through to the 48-month follow-up was considered a success (as it had started and stayed in business). As noted above, our sample in 2004 consisted of 91 of the original 151 (60%). Among these 91, the distribution of startup status responses is: operating business (22 responses, 24%), active startup (6 responses, 7%), inactive startup (4 responses, 4%), abandoned start-up (59 responses, 65%). We excluded the active (still trying to start) and inactive (no current start-up efforts but venture not abandoned) start-ups from our analysis since the final outcome of these 10 startup efforts is yet to be determined. Consequently, our results are based on the 81 of the original 151 (54%) that had either succeeded in establishing and sustaining an operating business or abandoned their start-up efforts. Considering that the data for our analysis involved four phone interviews, each a year apart, this is a very good response rate. In comparison, Gartner and Shaver (2002) report a US response rate of 55% with only two follow-up interviews.

Initial Attributional Coding

To examine the relationship between attributions and success in starting a business, the present article draws on data collected from five open-ended questions included in the phone interview:

"Why are you starting this business?"

"Why do you expect your business to be successful?"

"What major problems have you had in starting this business?"

"What major problems do you expect in the future?"

"What is the most important reason why you gave up?"

All responses were coded by one of the principal researchers and an undergraduate research assistant according to the protocol developed and documented by Shaver et al. (2001). Prior to coding the sample's responses, training (which took about five hours) in the coding process was conducted using item sets provided by Shaver et al. (2001).

The coding procedure involved the following three steps. Acting independently, each coder first parsed every answer into the number of separate explanations it contained. Then for each separate explanation, the coder first decided whether the explanation identified a factor internal to the person or a factor in the external environment. Once the internal/external decision had been made, the coder then decided whether the explanation identified a stable characteristic (one that would not, or could not, change in the immediate short term) or a variable characteristic (one that would, or could, change in the immediate short term). For example, "Have more independence" is an internal-stable attribution for a reason to start, whereas "competition" is an external-variable attribution for a problem.

Reliability testing was based on the complete sample and was determined separately for each step in the process. Cohen's kappa was used to assess intercoder reliability since it accounts for agreement that could be expected to occur by chance. Despite the simplicity and widespread use of percent agreement, we did not use this measure as it is known to be a misleading and inappropriately liberal measure of intercoder agreement for nominal-level variables. (It overestimates true intercoder agreement.) Since the kappa index tends to be more conservative than other indices, a kappa coefficient of 0.61 is said to represent reasonably good overall agreement (Kvalseth, 1989; Wheelock et al., 2000; Landis and Koch, 1977).

Table 2 presents the intercoder reliability results using Cohen's kappa. With one exception (kappa of .517 for the stable/variable step of reasons for starting), these results show a reasonably strong degree of reliability. Indeed, the exception does fall within the range (between 0.41 and 0.60) that some scholars (Landis and Koch, 1977) contend provide a moderate level of agreement. The few disagreements in coding that did arise were resolved by discussions among the two coders.

We confined our analyses to the first-mentioned reason cited, as the literature has shown that the "top-of-mind" response is the one perceived to be the most important by a respondent (see, for example, Miranda, Konya and Havrila, 2005; Lesser and Thumuluri, 2000), and coded this element into the four attributional categories. Using SPSS, crosstabulations were conducted to compare categorical variables, with the significance of relationships determined using chi-square analysis.

Success in Starting a Business

One approach in operationalizing a concept is to break it down into different dimensions (Bryman, 1993). Returning to our general definition provided earlier, we identify three dimensions as important in meaningfully defining a "new firm": recency or newness, a form of organization, and sales.

Here we consider a business recently formed (or new) if it became operational within the previous 12 months, with newness being measured at the individual level according to self-reports of the status of the start-up and the reported date when operating status was achieved. During each follow-up phone interview, the respondents were asked: "How would you describe the current status of this start-up effort? Is it now an operating business, still in an active start-up phase, still a start-up but currently inactive, no longer being worked on by anyone, or something else?" A categorical variable was used to classify respondents according to whether their start-up was: operating, active, inactive, no longer worked on by anyone, or something else.

In adopting any of three main organizational forms--proprietorship, partnership, or incorportation--goods and services can be provided for sale. Therefore, self-reports of any of these forms would constitute an appropriate indicator of organizational form.

Sales of goods and/or services (Busenitz and Murphy, 1996) is one of the many criteria used in previous studies to identify a new operating firm. (For a critical assessment of these criteria see Carter, Gartner and Reynolds, 2004). While sales is the one criterion that would be an appropriate indicator of being operational for all firms, by itself, this indicator is problematic. For example, Diochon, Menzies and Gasse (2005b) found sales to be reported by those who have yet to consider themselves operational. However, if an individual perceives his or her start-up to be operational and reports sales of goods and services, arguably, this would be a valid indicator of firm founding. In sum, we contend that if a start-up exhibits all three dimensions outlined above, then it has made the transition from becoming to being a new business. As indicated earlier, we confined our analysis to those starting and staying in business and those reporting abandoning their efforts to start a business at the time of the 48-month follow-up.

Persistence after "Failure"

A categorical variable was used to measure entrepreneurial persistence after "failure" to establish a new business. Respondents who reported no longer working on their start-up were asked: "Would you say that you expect to be involved in another start-up?" Response categories include: "never again (n=16)," "under the right conditions (n=28)," "most certainly (n=13)."

Results

Attributions

Table 3 presents the coding results for the first mentioned responses to the attributional questions. As predicted, an internal stable attribution was the type of response most frequently provided in explaining positive situations--why they are starting a business (44.4%) and why they expect to be successful (35.4%). This evidence supports Hypothesis 1.

In terms of how the two negative situations were described--current problems and expected problems--both were most frequently ascribed internal variable attributions (50.0% and 61.6% respectively). Therefore, Hypothesis 2 was not supported. Although the next most frequently cited main reason for current and expected problems was external and variable, the evidence indicates that, overall, the sample does not appear to be very susceptible to the self-serving bias. These results contrast with those reported in Gartner and Shaver's (2002) investigation of nascent entrepreneurs in the United States (US). In their study the most frequently cited cause of problems was external and variable, with internal and variable the next most-used category. Interestingly, though, respondents from both studies attributed problems to variable (external and internal) reasons.

A 4x4 crosstabulation to test for the self-serving bias produced too many cells with expected values less than 5. Considering the previous results, we deemed it appropriate to focus on examining the locus of causality for the first mentioned reason for starting a business by the first mentioned problem. Table 4 shows that internal attributions were the most frequently offered first response to both questions, with no significant relationship between the attributions: [chi square] (df=1) =.425, [rho]=.515. In other words, there was no difference between respondents' locus of causality for starting a business and for the problems encountered. Generally, Canadian nascent entrepreneurs are not very susceptible to the self-serving bias. Instead, their attributions tend to demonstrate a distinctly internal pattern. Indeed, the evidence suggests there may be a cultural difference between Canadian and US entrepreneurs, whereby Canadians are more likely to take personal responsibility for setbacks or failures.

Overall, the attributions reflect the sample's predilection to look for internal causes for both positive and negative situations. For the two positive situations, the vast majority of the sample provided internal attributions for each (62.9% and 56.9% respectively). In describing the two negative situations a similar pattern was noted (61.3% and 72.6%). This same pattern was evident when the responses for "why start a business" were crosstabulated with "problems experienced." Respondents were almost twice as likely to attribute the cause of both to internal factors (see Table 4). Based on this evidence, it would appear that respondents' descriptions of positive and negative situations are, to some extent, determined by the cognitive structures of the respondents, as opposed to the data.

Attributions and Success in Firm Formation

A key objective of this article was to determine whether nascent entrepreneurs' attributions were related to success in forming a firm. Table 5 presents the results of analysis for testing whether nascent entrepreneurs who were successful in starting a business were less susceptible to the self-serving bias than nascent entrepreneurs who have given up on their efforts to start a business. Among those who successfully formed a firm too many cells had expected counts less than 5. Therefore, Hypothesis 3 could not be tested. However, it is interesting to note that the data presented in Table 5 indicates more of a susceptibility to the self-serving bias among those who were successful in creating a firm than among those who were unsuccessful. Indeed, among those who were successful in starting a business, 28.6% offered an internal locus of causality in describing why they started and an external locus of causality in describing problems. Among those who had abandoned their efforts to start a business, 20.3% were similarly categorized.

In light of the previous results, we explored whether success in forming a firm was related to the locus of causality for any of the first four attributional questions. Only one significant relationship was found--between success and the first reason given for starting a business (see Table 6). Nascent entrepreneurs who succeeded in starting a business were more likely to have an internal locus of causality than those who were unsuccessful in starting a firm. When describing their first reason for starting a business, about 40% of those reporting an internal locus of causality had succeeded in establishing a new business while only about seven percent of those with an external locus of causality succeeded in doing so. This suggests that personal reasons, rather than external market opportunities, are the primary drivers of decisions to create new ventures that eventually are successful. Tests of whether a stable internal pattern for "why start," and a variable internal pattern for problems, would be related to success in starting showed no significance.

Among those who did not persist in their efforts to start a business, Table 3 has shown that the most frequently provided response was an internal stable attribution when asked what the most important reason was for giving up [[chi square] (df=1) =3.872, [rho]=.049]. Consequently, Hypothesis 4--among nascent entrepreneurs who do not persist in their efforts to start a business internal stable reasons will be the least likely explanation of giving up--is not supported.

With 58% (n=33) providing internal reasons for giving up, clearly, there was no evidence of a "self-serving" bias. Considering this evidence in terms of the three requirements for task success presented by Heider, the intention to start a business was present, as was exertion in the direction of the intention. However, it would appear that personal ability was not perceived to exceed the difficulty of the task (starting a business).

In light of these results, it is somewhat surprising that the prospect of starting another venture was just as appealing to entrepreneurs who provided internal stable reasons for giving up as those providing other reasons for abandoning their efforts to start a business. Table 7 shows there were no statistically significant differences in expectations for involvement in another venture, based on the reasons provided for giving up. Therefore, Hypothesis 5 is not supported. Contrary to theory, 72.7% (n=16) of Canadian nascent entrepreneurs who attributed the cause of "failure" to internal stable factors expected to be involved in another start-up. What possible explanation would there be for this? Analysis of the various answers in this response category sheds light on the issue by revealing a number of themes. The most commonly cited reasons for giving up on the start-up include: taking a job offer/promotion at work (four individuals); health reasons (four individuals); sale of venture (three individuals); change in their personal circumstances (one had a baby and one moved to another town with their partner); and two said they found it too difficult. These findings suggest that the vast majority of respondents neither interpret giving up as a negative event nor as a "failure," thereby helping to explain why they expect to be involved in another start-up.

It is interesting to consider the results of this study in light of two noted tendencies among entrepreneurs: perceiving their own abilities, dedication, and efforts as crucial to success; and overestimating the odds for success (Baron, 1998). While the first tendency is evident, the second is less so. These Canadian nascent entrepreneurs were less likely to provide internal attributions for future success than they were for current and future problems (Table 3).

Discussion and Conclusion

This paper set out to explore whether individual differences in attributional style are associated with later success in starting (and staying) in business. Past research had shown that most people exhibit cognitive biases in their thinking and that entrepreneurs are more susceptible to the self-serving bias than others. In expecting to find a similar pattern among our sample of Canadian nascent entrepreneurs, we also expected less susceptibility to the self-serving bias among those who were successful in establishing a business.

Instead of exhibiting a self-serving bias, these nascent entrepreneurs were found to have an internal locus of causality when interpreting situations. While positive situations were perceived to be a result of their abilities (internal/stable), problems were perceived to be a result of their effort (internal/variable). In terms of the attributional framework, overcoming these problems would require greater effort.

With success in establishing a new business neither related to explanatory style nor to the utilization of a self-serving bias, these findings indicate that attribution theory is of little use in explaining why some people succeed in starting a business while others do not. Although the reasons Canadian nascent entrepreneurs gave for entering business were related to later success in establishing a new venture, their other attributions were neither related to success nor to entrepreneurial persistence.

There are several differences between our study and the research of others that might explain our unanticipated results. First, ours is the first real time study of nascent entrepreneurs' attributions that are examined in relation to "success" in both starting and staying in business. Second, our study is the first to explicitly define a "new business." Third, our study's measures emerge from the theoretical perspective taken. In using attribution theory as the theoretical lens, the focus is on the meaning nascent entrepreneurs give to their own activities/behaviour and start-up outcomes. As Gartner and Shane (1995) attest, there are very few articles in the literature that use measures of entrepreneurship that are theoretically derived. Finally, ours is the first study that uses a nationally representative sample to systematically study both those who succeeded in establishing a new venture and those who did not. Typically, only those who succeed are studied, as was the case when the attributions of "successful" entrepreneurs were compared to those of managers (rather than to those of people who did not "succeed" in establishing a business).

Fundamentally, these findings have implications for theory development. In particular they highlight the need to address the meaning and measurement of key terms. Within this research a sustainability dimension was incorporated into our definition of success in new firm formation. After four years of follow-up, 22 people reported starting and staying in business while 59 reported abandoning their efforts to start a business. Of these 59, 22% (n=13) had achieved operating status at one point in time and later gave up. Such findings underscore the importance of sustainability to the development of a better understanding of new firm formation.

For policymakers, the findings with respect to why ventures were abandoned point to the importance of ensuring that the outcomes of programs designed to encourage new firm formation are being validly measured and appropriately evaluated. Moreover, could the fact that Canadian nascent entrepreneurs are more inclined to take personal responsibility for their actions help explain why the advice and assistance of government agencies are generally known to be underutilized? This is an issue for future research.

For prospective entrepreneurs, there is cause for optimism concerning involvement in the start-up process. Indeed, although 59 of the 81 nascent entrepreneurs we studied in this article gave up on their start-up, most did not do so because they failed. The fact that the majority of these individuals expected to be involved in another start-up helps confirm that failure is being overstated. Furthermore, prospective entrepreneurs would benefit from considering their reason(s) for starting a business as these reasons appear to be related to starting and staying in business.

Indeed, the unique attributional style of Canadian nascent entrepreneurs merits further research, particularly on an international comparative basis. Not only would it be useful to more thoroughly investigate the role of cultural factors in the entrepreneurial process, it would also be useful to determine whether attributions change, subsequent to start-up.

No research is without limitations, and ours is no exception. Clearly, the sample size and its attrition rate constrain the nature and extent of analysis that could be conducted. For example, attributions could not be assessed according to demographic factors. Furthermore, in basing our conclusions about success on the fourth year follow-up data, the sample became too small to conduct separate analyses for those who started but did not stay in business as compared to those who never started a business and those who started and stayed in business. In future, research based on a larger sample is recommended.

Despite its weaknesses, this research does contribute to the literature. Many scholars have argued that attribution style might be useful in explaining the success of entrepreneurs. However, this is the first empirical study to investigate the issue among a representative sample of Canadian nascent entrepreneurs whose success in forming a new business was determined according to whether they started and stayed in business. Moreover, by assessing the reasons people gave for abandoning their efforts to start a business, valuable insight is provided about participants in the entrepreneurial process that we know very little about.

Acknowledgements

The authors are most grateful for the generous financial support of the Social Sciences and Humanities Research Council of Canada and Industry Canada.

Contact Information

For further information on this article, contact

Dr. Monica Diochon, Schwartz School of Business Administration and Information Systems, St. Francis Xavier University, Antigonish, Nova Scotia, Canada B2G 2W5 Tel: 902 867-5412 email: mdiochon@stfx.ca

Dr. Teresa V. Menzies, Faculty of Business, Brock University, St. Catharines, Ontario, Canada L2S 3A1 Tel: 905 688 5550 x 4118 email: tmenzies@brocku.ca

Dr. Yvon Gasse, Faculte des sciences de l'administration, Universite Laval, Sainte-Foy, PQ, Canada G1K 7P4 Tel: 418 656 7960 email: yvon.gasse@mng.ulaval.ca

References

Anderson, C.A. 1991. "How People Think About Causes: Examination of the Typical Phenomenal Organization of Attributions for Success and Failure," Social Cognition 9: 295-329.

Baron, R.A. 1998. "Cognitive Mechanisms in Entrepreneurship: Why and When Entrepreneurs Think Differently Than Other People," Journal of Business Venturing 12: 275-94.

Beaver, G. 2003. "Small Business: Success and Failure," Strategic Change 12, no. 3: 115-22.

Busenitz, L.W. and G.B. Murphy. 1990. "New Evidence in the Pursuit of Locating New Businesses," Journal of Business Venturing 11, no. 3: 221-31.

Cardon, M.S. and R.G. McGrath. 1999. "When the Going Gets Tough ... Toward a Psychology of Entrepreneurial Failure and Re-Motivation," http://www.babson.edu/entrep/fer/papers99/I/I_B/I_B%20Text.htm accessed on July 16, 2004.

Carland, J.W., J.C. Carland and J.W. Carland, III. 1995. "A Model of Entrepreneurship: The Process of Venture Creation," Retrived March 20, 1997, from http://www.sbaer.uca.edu/docs/proceedings/95sbi013.txt

Carter, N.M., W.B. Gartner and P.D. Reynolds. 2004. "Firm Founding." Pp 311-23 in W.B. Gartner, K.G. Shaver, N.M. Carter and P.D. Reynolds (eds.), Handbook of Entrepreneurial Dynamics: The Process of Business Creation. Thousand Oaks, CA: Sage.

Diochon, M., T. Menzies and Y. Gasse. 2005a. "From Becoming To Being: Measuring Firm Creation," paper presented at the Enterprise Development Growth and Expansion (EDGE) Conference, Singapore, July 11-13.

Diochon, M. T. Menzies and Y. Gasse. 2005b. "Exploring the Relationship Between Start-up Activities and New Venture Emergence: A Longitudinal Study of Canadian Nascent Entrepreneurs," International Journal of Management and Enterprise Development 2, nos 3/4: 408-26.

Forbes, D.P. 2005. "Are Some Entrepreneurs More Overconfident Than Others?," Journal of Business Venturing 20, no. 5: 623-40.

Ford, J.D. 1985. "The Effects of Causal Attributions on Decision Makers' Responses to Performance Downturns," Academy of Management Review 10: 770-86.

Gartner, W.B. and S.A. Shane. 1995. "Measuring Entrepreneurship Over Time," Journal of Business Venturing 10, no. 4: 283-301.

Gartner, W.B. and K.G. Shaver. 2002. "The Attributional Characteristics of Opportunities and Problems Described by Nascent Entrepreneurs in the PSED," paper presented at the Babson-Kauffman Entrepreneurship Research Conference, Boulder, CO.

Gartner, W.B., K.G. Shaver, N.M.Carter and P.D. Reynolds (eds.). 2004. Handbook of Entrepreneurial Dynamics: The Process of Business Creation. Thousand Oaks, CA: Sage.

Gray, C. 1998. Enterprise and Culture. London: Routledge.

Heider, F. 1958. The Psychology of Interpersonal Relations. New York: Wiley.

Kahneman, D. and D. Lovallo. 1993. "Timid Choices and Bold Forecasts: A Cognitive Perspective on Risk Taking," Management Science 39: 17-31.

Kvalseth, T.O. 1989. "Note on Cohen's Kappa," Psychological Reports 65: 223-26.

Kelley, H.H. and J.L. Michela. 1980. "Attribution Theory and Research," Annual Review of Psychology 31: 457-501.

Landis, J.R. and G.G. Koch. 1977. "The Measurement of Observer Agreement for Categorical Data," Biometrics 33: 159-74.

Lesser, J.A. and L.K Thumuluri. 2000. "Beyond Research Streams: Systematic Integration of Consumer Concepts. Twentieth Century Postscript," Management Research News 23, nos. 5/6: 55-72. Retrieved October 1, 2006, from ABI/INFORM Global database. (Document ID: 290263931).

Miranda, M.J., L. Konya and I. Havrila. 2005. "Shoppers' Satisfaction Levels Are Not the Only Key to Store Loyalty," Marketing Intelligence & Planning 23, nos. 2/3: 220-32.

Martocchio, M.J. and W.L. Gardner. 1987. "The Leader/Member Attribution Process," Academy of Management Review 12: 235-49.

Miller, A.G. and T. Lawson. 1989. "The Effect of an Informational Option on the Fundamental Attribution Error," Personality and Social Psychology Bulletin (June 1989) 194-204.

Nisbett, R.E., C. Caputa, P. Legant and J. Maracek. 1973. "Behavior as Seen by the Actor and as Seen by the Observer," Journal of Personality and Social Psychology 27: 154-64.

Palich, L.E., and D.R. Bagby. 1995. "Using Cognitive Theory to Explain Entrepreneurship Risk-Taking: Challenging Conventional Wisdom," Journal of Business Venturing 10, no. 6: 425-38.

Peterson, C. and P. Villanova. 1988. "An Expanded Attributional Style Questionnaire," Journal of Abnormal Psychology 97, no. 1: 87-89.

Pittman, T.S. 1993. "Control Motivation and Attitude Change." In G. Weary, F. Gleichner and K.L. Marsh (eds.), Control Motivation and Social Cognition. New York: Springer-Verlag.

Reynolds, P.D. 2000. "National Panel Study of U. S. Business Startups: Background and Methodology." Pp. 153-227 in J.A. Katz (ed.), Advances in Entrepreneurship, Firm Emergence and Growth. Stamford, CT: JAI Press Inc.

Rogoff, E.G., M-S. Lee and D-C Suh. 2004. "Who Done It? Attributions by Entrepreneurs and Experts of the Factors that Cause and Impede Small Business Success," Journal of Small Business Management 42, no. 4: 364-76.

Robbins, S.P., M. Coulter and R. Stuart-Kotze. 2003. Management: Canadian Seventh Edition. Toronto: Pearson Education Canada Inc.

Sapienza, H.J. 1991. "Comets and Duds: Characteristics Distinguishing High- and Low-Performing High-Potential Ventures," Frontiers of Entrepreneurial Research 6: 124-38.

Schermerhorn, J.R. 2005. Management, 8th ed. New York: John Wiley & Sons, Inc.

Shaver, K.G. 2004. "Attributions and Locus of Control." Pp. 205-32 in W.B. Gartner, K.G. Shaver, N.M.Carter and P.D. Reynolds (eds.), Handbook of Entrepreneurial Dynamics: The Process of Business Creation. Thousand Oaks, CA: Sage.

Shaver, K.G., W.B.Gartner, E. Crosby, K. Bakalarova, and E.J.Gatewood. 2001. "Attributions About Entrepreneurship: A Framework and Process for Analyzing Reasons for Starting a Business," Entrepreneurship Theory and Practice 26, no. 2: 5-32.

Shaver, K.G., W.B.Gartner, E.J.Gatewood and L.H. Vos. 1996. "Psychological Factors in Success at Getting into Business." Pp. 77-90 in P. Reynolds, S. Birley, J.E. Butler, W.D. Bygrave, P. Davidsson, W.B. Gartner and P.P. McDougall (eds.), Frontiers of Entrepreneurship Research 1996. Babson Park, MA: Babson College. http://www.babson.edu/entrep/fer/AWARDS/NFIB/html/nfib_paper.htm, accessed on June 21, 2005.

Shaver, K.G. and L.R. Scott. 1991. "Person, Process, Choice: The Psychology of New Venture Creation," Entrepreneurship Theory and Practice 16, no. 2: 23-45.

Stokes, D. and R. Blackburn. 2002. "Learning the Hard Way: The Lessons of Owner Managers Who Have Closed Their Businesses," Journal of Small Business and Enterprise Development 9, no. 1: 17-27.

Storey, D.J. 1994. Understanding the Small Business Sector. London: Routledge.

Thomson, A. and C. Gray. 1999. "The Determinants of Management Development in Small Business," Journal of Small Business and Enterprise Development 6, no. 2: 113-27.

Thornhill, S. and R. Amit. 2003. "Learning from Failure: Organizational Mortality and the Resource-Based View," Analytical Studies Branch Research Paper Series. Ottawa, ON: Statistics Canada.

Wheelock, A., W. Haney and D. Bebell. 2000. "What Can Student Drawings Tell Us About High-Stakes Testing in Massachusetts?" TCRecord.org. Available: http://www.tcrecord.org/Content.asp?ContentID=10634.

Weiner, B. 1985. "An Attributional Theory of Achievement Motivation and Emotion," Psychological Review 92: 548-73.

--. 1986. An Attributional Theory of Motivation and Emotion. New York: Springer-Verlag.

(1.) "Among the adults living your household, is there anyone who, alone or with others, is now trying to start a new venture?" "Will you be an owner, in part or in whole of this company or venture that you are trying to launch, alone or with others for your own business or that of your employer?" "During the last 12 months, have you done anything to help start this new business, such as looking for equipment or a location, organizing a start-up team, working on a business plan, saving money, or any other activity that would help launch a business?"

Monica Diochon, Schwartz School of Business Administration and Information Systems, St. Francis Xavier University, Antigonish, Nova Scotia

Teresa V. Menzies, Faculty of Business, Brock University, St. Catharines, Ontario

Yvon Gasse, Faculte des sciences de l'administration, Universite Laval, Sainte-Foy, Quebec

Table 1. Intentional Behaviour: Dimensions of Causal Attributions

                     Stability of the Cause

Locus of Causality   Stable            Variable
Internal             Ability           Effort
External             Task Difficulty   Luck

Table 2. Cohen's Kappa Results for Independent Ratings of the
Three-Step Coding Process of Respondents' Explanations

                                         Cohen's Kappa

                                 Parsing    Internal/   Stable/
Explanation of:                  Answers    External    Variable

Why business is being started      .808       .766        .517
Why business is expected to be     .792       .677        .691
  successful
Problems experienced               .942       .745        .623
Problems expected                  .908       .867        .701
Why gave up                        .791       .860        .862

Table 3. Coding Results for Attributions

                                          First      First
                              Average    Elements   Elements
Attributions        Sample   Number of   Internal   Internal
for:                (n=81)   Elements     Stable    Variable

Why start (n=81)      N        1.75         36         15
                      %                    44.4       18.5

Why expect            N        1.85         28         17
success (n=79)        %                    35.4       21.5

Current problems      N        1.39         9          40
(n=80)                %                    11.3       50.0

Expected problems     N        1.37         8          45
(n=73)                %                    11.0       61.6

Why gave up           N        1.41         22         11
(n=57)                %                    38.6       19.3

                     First      First
                    Elements   Elements
Attributions        External   External
for:                 Stable    Variable

Why start (n=81)       13         17
                      16.0       21.0

Why expect             21         13
success (n=79)        26/6       16.5

Current problems       6          25
(n=80)                7.5        31.3

Expected problems      6          14
(n=73)                8.2        19.2

Why gave up            10         14
(n=57)                17.5       24.6

Table 4. Locus of Causality Attributions for Reasons for Starting a
Business by the Reasons for Problems

                         Problems

Why Start   Internal     External     Total

             N       %    N       %    N       %

Internal    32    65.3   18    58.1   50    62.5
External    17    34.7   13    41.9   30    37.5
Total       49   100.0   31   100.0   80   100.0

[chi square] (df=1) =.425, [rho] =.515 n.s

Table 5. Locus of Causality Attributions for Reasons for Starting a
Business and Reasons for Problems by Firm Formation Status

                               Current Problems

Firm Formation Status                      Internal     External

                                           N       %    N       %

Successful (a)        Why      Internal   13    92.9    6    85.7
(operating)           Start?   External    1     7.1    1    14.3
                      Total               14   100.0    7   100.0

Unsuccessful (b)      Why      Internal   19    54.3   12    50.0
(efforts abandoned)   Start?   External   16    45.7   12    50.0
                      Total               35   100.0   24   100.0

                               Current Problems

Firm Formation Status                        Total

                                           N       %

Successful (a)        Why      Internal   19    90.5
(operating)           Start?   External    2     9.5
                      Total               21   100.0

Unsuccessful (b)      Why      Internal   31    52.5
(efforts abandoned)   Start?   External   28    47.5
                      Total               59   100.0

(a.) too many cells had expected counts less than 5.

(b.) [chi square] (df=1) =.105 [rho] =.746 n.s.

Table 6. Firm Formation Success by Locus of Causality for Starting a
Business

                                 Locus of Causality

                          Internal     External      Total

Formation Status          N       %    N       %   N       %

Successful (operating)   20    39.2    2     6.7  22    27.2
Unsuccessful             31    60.8   28    93.3  59    72.8
  (efforts abandoned)
Total                    51   100.0   30   100.0  81   100.0

[chi square] (df=1) =10.115, [rho] =.001

Table 7. Expectations for Future Start-up Involvement by
Reasons for Giving Up

                                   Attributions

                        Internal
                         Stable      All Other      Total

Expectations for        N       %    N       %    N       %
Future Involvement

Never again             6    27.3   10    28.6   16    28.1
Under the right        11    50.0   17    48.6   28    49.1
  conditions
Most certainly          5    22.7    8    22.8   13    22.8
Total                  22   100.0   35   100.0   57   100.0

[chi square] (df=2) =.014, [rho] =.993 n.s.