The role of attributional explanatory style in the perceived outcomes of entrepreneurial venture failure.

Askim-Lovseth, Mary K. ; Feinberg, Richard A.

Introduction

Failure of a business must be viewed as a chapter in the life of an entrepreneur. That chapter and the experiences throughout the business venture may impact many aspects of the entrepreneur's well-being. Some note that failure can be a crossroads that can lead to future success (Parks, 1977; Sellers, 1995). Others find that it can be a time of transition for the entrepreneur and a stressful and emotional event (Patzelt and Shepherd, 2011; Ronstadt, 1985; Shepherd, 2003; Shepherd, Wiklund and Haynie, 2009), and that experiences which preclude a venture start-up can influence the entrepreneur's attitude towards failure (Politis and Gabrielsson, 2007; Ucbasaran et al., 2010). How entrepreneurs can cope with failure is gaining greater attention in the literature (Minniti and Bygrave, 2001; Patzelt and Shepherd, 2011; Shepherd, 2003, 2004).

The element of failure can be present in the minds of entrepreneurs because of the risk that is undertaken. Brockhaus (1980) defined three components of entrepreneurial risk: the risk-taking propensity of an entrepreneur, the perception of failure for the venture, and the perceived consequences resulting from failure. The first component, the propensity to take risks, has held the greatest interest for researchers. The second component, the perception of failure for the venture, has taken the direction of causes of venture failure via ex post facto analysis. That is understandable because entrepreneurs believe they will be successful and their higher motivation to achieve is not congruent with focusing on failure. The third component, the perceived consequences of failure, has centered on career decisions.

That there is a level of risk involved in starting a business has not been debated and there have been numerous studies (Ahmed, 1985; Hornaday and Aboud, 1971; Masters and Meier, 1988) that have focused on risk-taking propensity. Developing strategies to minimize or avoid the risks more appropriately characterizes the behavior of an entrepreneur (Mitton, 1989; Palmer, 1971).

Though risk is evident in creating the venture, little has been written about what is at personal risk for the entrepreneur. Assuredly, the entrepreneur does take a personal risk. This risk and not succeeding with the venture are likely to have an impact on the entrepreneur who has failed. This impact may be related to how the ex-entrepreneur explains the causes of failure to himself. This study proposes to answer the question, is there a relationship between how the entrepreneur explains the failed venture and the perceived outcomes related to that failure?

There is a well-developed and understood theoretical and empirical literature in social psychology called attribution theory to help understand the causal interpretation of events. While the logical underpinnings of connecting this social psychological work to the current work is conceptually logical (attributions affect decisions, therefore, attributions should affect business decisions), the connection in the entrepreneurial research literature has not been made. The theoretical technique used in this study of taking a well-understood area of theory and research in one area and applying it to another to increase understanding is called substantive modeling (Popper, 1963).

In this research, the better understood stream of research called attribution theory will be used to understand the relationship of failure to interpreting the outcomes of failure and the future decisions of the entrepreneur. Explanatory style is a means by which these attributions may be determined. In much the same way an individual's attributional explanatory style influences consequent attributions and decisions, an entrepreneur's explanatory style should have a role in the perceived outcomes of a failed entrepreneurial venture, which in turn affects consequent decisions. Since failure is likely for many start-up businesses, then knowing the entrepreneur's attributional style prior to venture startup can imply how the entrepreneur may react if the venture fails. It may even influence the likelihood of another venture startup. There has been a void of research in this area of entrepreneurial study.

We begin by presenting the conceptual framework for this study, including hypotheses development. This is followed by a discussion of the sample, variables, study design, and data analysis. The paper concludes with addressing the limitations of the study and implications for professional practice or applied settings.

Conceptual Framework

Attribution literature had generally focused on two main streams--the basic processes of attributions and the consequences of attributions (Harvey and Weary, 1984). The relationships among causal ascription, affect, and action were noted in some of the earliest works of attribution theory (Heider, 1958; Kelley, 1967) and continued on (Abramson, Seligman and Teasdale, 1978; Chandler, Seibel and Spies, 1990; Mikulincer, 1986; Peterson and Seligman, 1984; Peterson, Maier and Seligman, 1993; Weiner, 1986, Weiner et al., 1971; Weiner et al., 1972).

Since attributions can affect the thoughts and behaviors relating to consequences of an event and can also be used to predict outcomes (Seligman and Schulman, 1986), explanatory style should be a particularly relevant attributional issue for understanding the relationships between the causal interpretation of the events that led to the failure of a business and what the outcomes resulting from that failure may be for the entrepreneur.

Attribution can be defined as the linking of conditions and causes to an event in order to give it meaning; it is the process by which people can interpret an event and make causal explanations (Heider, 1958; Kelley, 1967, 1973). Numerous causes may be used to explain an event and the work in attribution theory has helped develop clear understanding and rules in the relationship between attributions and behavior. These rules may help in understanding and predicting the relationship between success/failure in an entrepreneurial activity and its outcomes for the entrepreneur.

The research in this area has uncovered three fundamental attributions. These fundamental attributions have been found to have significant, enduring, and widespread effects on behavior and decision-making (Darley and Cooper, 1998). First, success or failure can be attributed to either internal ("I did it") or external (luck, fate, powerful other) factors. Second, the nature of the environment in which events occur varies or is stable. What an individual does following success or failure depends in large part on the attributions concerning the stability or instability of the environment in which that success or failure occurred. The third attribution is the degree to which an individual believes that an event is global or specific.

That numerous causes, not single variables, clearly explain behavior warrants a classification system that would provide a deeper understanding to attributing causality of an event. This realization has led to the development of a multidimensional attributional style (Abramson, Seligman and Teasdale, 1978; Crittenden, 1989; Curren, Folkes and Steckel, 1992; Fletcher et al., 1986; Kelley, 1967, 1973; Peterson, Maier and Seligman, 1993; Weiner, 1979; 1986; White, 1991; Williams, Lees-Haley and Brown, 1993). Abramson, Seligman and Teasdale (1978) created a multidimensional scale to measure the three dimensions of causal explanation: internal/external, stable/unstable, and global/specific.

In the context of a failure, an internal locus of causality would lead an individual to explain the experienced failure as a personal characteristic of that individual. An external attribution leads to the cause of the failure being environmental or situational factors. In this latter scenario, the outcome would be no different if someone else were placed in that same situation. The stability dimension refers to the causes being the same over time; unstable causes are transient. Concerning globality, global causes generalize across many situations whereas the contrary is situation specific.

If the individual has an internal, stable, and global explanatory style concerning the causal beliefs of the unsuccessful event, nothing can be done to improve the outcomes of ensuing events. Each of these styles, according to Abramson, Seligman and Teasdale (1978), have measurable and demonstrable impacts. A depressed state is most likely to result when a failure is attributed to internal, stable, and global factors because the individual expects future outcomes to be similar to the past. There will be a repetitive pattern of outcomes no matter what the individual does. If the individual makes an external, unstable, and specific attribution, Abramson and his colleagues (1978: 70) assert that the "... external attribution for failure raises self-esteem, an unstable one cuts the deficits short, and a specific one makes the deficits less general." The failure does not generalize to other situations for the individual.

The dimensions of internality, stability, and globality are rooted in the reformulated theory of learned helplessness (Abramson, Seligman and Teasdale, 1978; Peterson, Maier and Seligman, 1993), which hypothesized how individuals respond to uncontrollable events. It basically asserts that a state of helplessness results for an individual when exposed to unsolvable and uncontrollable problems. Yet, Abramson, Seligman and Teasdale (1978: 54) noted failure and uncontrollability are not always related.

"In current attribution theories...success and failure refer to outcomes. Success refers to obtaining a desired outcome and failure to not obtaining a desired outcome. According to this framework, then, the term failure does not embrace all cases of uncontrollability. Thus, from a strict attributional point of view, failure and uncontrollability are not synonymous: Failure is a subset of uncontrollability involving bad outcomes."

In this study, the assumption cannot be made that all events leading to the failure of a business venture will be viewed as uncontrollable by the entrepreneur. The controllability dimension of causality presented by Weiner (1979, 1986) has relevance to understanding the causal belief systems of entrepreneurs.

Weiner (1979, 1986) posited that individuals in achievement situations make causal inferences for their success or failure outcomes. This cognitive processing includes perceptions about the events leading to the outcome which influence subsequent behavior and expectations. Weiner tendered three dimensions in his theory of motivation based upon attributions of causality for success and failure: locus of causality, stability, and controllability.

Combining the dimensions of the previously noted researchers produces the four dimensions of this study: internality (internal/external), stability (stable/unstable), globality (global/specific), and controllability (controllable/uncontrollable). The application of these dimensions to causal explanations of failure in an entrepreneurial situation can be illustrated by the following general examples.

An individual ascribing failure of a business venture to lack of general ability or competence would be using internal, stable, global, and uncontrollable attributions. The inherent lack of ability would be seen by the entrepreneur as a trait that would manifest itself in other undertakings. As a result, such attributions would lead to lower probability of undertaking a new venture following failure. On the other hand, if individuals attributed a failed venture to outside forces which impaired their competence (government regulations led to a business failure), the individuals might be likely to start a new venture if that external attribution was perceived as unstable, specific, and controllable.

Acknowledging there is a personal risk factor involved with business ventures, Liles (1974) identified four areas being influenced by the entrepreneurial start-up: (1) financial well-being, (2) career opportunities, (3) family relations, and (4) psychic well-being. Ronstadt (1981, 1982, 1984, 1985) and Brockhaus (1985) have done initial work on entrepreneurial careers and the time of exit from the career; the likelihood of starting another venture after failure; and the effects of failure on financial, personal, and family factors. Since the early research of Ronstadt and Brockhaus, others have investigated these risk areas--personal expectations of financial success (Cooper et al., 1992; Naffziger, Hornsby and Kuratko, 1994; Wicker and Conn, 1990); career opportunities (Shepherd, Wiklund and Haynie, 2009); family relations (Dyer, 1992; Harrell, 1994); and psychic well-being or self-esteem (Bunker and Webb, 1992; Harrell, 1994; Kernis, Brockner and Frankel, 1989; O'Connor and Wolfe, 1987; Rhodewalt et al., 1991).

In understanding the impact of failure on Liles' (1974) four areas of risk undertaken by the entrepreneur, certain explanatory styles should be more pervasive dependent upon the controllability of the situation. When the event that had the greatest effect on each of the risk areas is perceived as controllable and the risk (e.g., career opportunities) was negatively affected (e.g., would not start another entrepreneurial venture), explanatory style would more likely be internal, global, and stable. The entrepreneur would take more personal blame, have thoughts that it would occur again in the next undertaking, and would have lower expectancies for future success. If the event was uncontrollable and the entrepreneur's career was less negatively affected or there was the possibility of starting another venture because of the lessons learned, explanatory style would more likely be external, specific, and unstable. If the entrepreneur felt helpless because of the uncontrollability factor, yet that helplessness was only related to the particular event, it should not deter the entrepreneur from the possibility of starting another venture sometime in the future. Following this logic, the entrepreneur's financial well-being, family relations, and psychic well-being would also have fewer negative effects if the explanatory style was external, unstable, specific, and uncontrollable.

Based on this conceptual framework, the following major hypotheses were postulated concerning the overall explanatory style (though not noted, hypotheses were proposed for each dimension of explanatory style for each outcome):

H1: There is a positive relationship between external, unstable, specific, uncontrollable attributions and the likelihood to have improved financial well-being or have financial well-being unaffected after a venture failure.

H2: There is a positive relationship between external, unstable, specific, uncontrollable attributions and the likelihood of continuing with an entrepreneurial career after a venture failure.

H3: There is a positive relationship between external, unstable, specific, uncontrollable attributions and the likelihood to have improved family relations or have family relations unaffected after a venture failure.

H4: There is a positive relationship between external, unstable, specific, uncontrollable attributions and the likelihood to have improved psychic well-being or have psychic wellbeing unaffected after a venture failure.

Combining the dimensions in any other manner is predicted to not result in more positive outcomes. The most desirable style is that of external, unstable, specific, and uncontrollable for bad events (failure), coupled with internal, stable, global, and controllable for good events (success). With the occurrence of a business failure, the entrepreneur would fare better if the perception were less personal responsibility assessed, less likelihood of its reoccurrence, less pervasiveness over other situations, and less control over the event. This study proposes that an individual's explanatory style plays an influential role in the perceived outcomes of a failed entrepreneurial venture.

Methodology

This research is within the tradition of theory testing of relationships; it extends the attributional explanatory theory and generalizations to the area of entrepreneurial research to determine the relationship between the causal (independent) variables and the outcomes (dependent variables) to be predicted (Malhotra, 1993). In this study, the relationship between attributional explanatory style and the perceived outcomes of failure on decisions concerning financial, career, family, and self-esteem were assessed using a within-subjects correlational design. The study involved three measurement instruments. The first measured entrepreneurial attitudes of the sample by using the Entrepreneurial Attitude Orientation (EAO) scale developed by Robinson et al. (1991). The second assessed cross-situational tendencies to explain events in a particular way (trait approach) using the Attributional Style Questionnaire (ASQ) developed by Peterson et al. (1982). The third measured causal explanations for specific events that have occurred (situational approach). Three business scenarios related to venture success, failure, and exit with subsequent statements concerning inancial well-being, career opportunities, family relations, and psychic well-being (self-esteem) outcomes were developed. Figure 1 illustrates the experiment phases and determination of the final sample subset.

[FIGURE 1 OMITTED]

Sample

The past problems in finding "real and actual" ex-entrepreneurs for research (Ronstadt, 1985; Wicker and Conn, 1990) because they have a tendency to disappear after the venture failure, necessitated supporting the efficacy of using college students in this type of study. Though many students start business ventures while attending college and/or participate in student organizations that promote the development and success of young entrepreneurs, two other important benchmarks were considered: (1) students are appropriate subjects for tests of theory in social science, and (2) students have been used in entrepreneurial research in the past.

Students are commonly used in theoretically driven research. It is argued that samples need not be representative if one is testing a prediction rather than making one (Berkowitz and Donnerstein, 1982; Calder, Phillips, and Tybout, 1981, 1983; McGrath and Brinberg, 1983; Mook, 1983). External validity would be an issue if the results of this study were to be generalized to the population of interest, ex-entrepreneurs. In order for the theoretical explanation to be generalizable, the theory must survive attempts at falsification and attempts must be made to extend or falsify theory to business context. Since the intent was to test the application of combining theory to entrepreneurial situations in a controlled experimental setting, Calder, Phillips, and Tybout (1981) assert that the best sample for this type of study should be homogeneous on dimensions that would most likely influence the variables of interest. The use of college students, homogeneous on a dimension of entrepreneurial attitude (i.e., innovation), would satisfy this criterion.

In the past, research has used college students as subject populations successfully. Three areas of entrepreneurial research noted include instrument development (Robinson et al., 1991), determining entrepreneurial and non-entrepreneurial characteristics (Koh, 1996; Sexton and Bowman, 1983a, 1983b, 1984, 1986), and factors affecting entrepreneurial career aspirations (Brenner, Pringle and Greenhaus, 1991; Crant, 1996; Krueger, 1993; Scott and Twomey, 1988).

In this study, students were an appropriate sample for the theoretically driven research since prior research supports the notion that college students can be used to develop and uncover relations between variables. Therefore, the subjects were students in 15 upper-level business classes from two major Midwestern universities (N = 336). The sampling method was purposive nonprobability sampling. A median split of the mean scores of the innovation dimension of the EAO scale produced the sample subset. The mean score range of the dimension was 14.06 to 26.00, from a possible 1.00 to 26.00. The subset included those subjects who had mean scores > = 20.33.

Entrepreneurial Attitude Orientation (EAO) scale. The purpose of the EAO scale was to determine from the sample those respondents who had an entrepreneurial attitude versus those that could be classified as having a non-entrepreneurial attitude. The respondents with an entrepreneurial attitude were the sample subset used in this research.

The EAO scale is comprised of four attitude subscales--achievement, personal control, innovation, and self-esteem. The 75 items on the Likert-type scale are scored on a "strongly agree" to "strongly disagree" 10-point scale. Each item also represents an affective, cognitive, or behavioral reaction. The four subscales were found to have a 77 percent accuracy in predicting between entrepreneurial and non-entrepreneurial groups (Robinson et al., 1991).

There are two alternatives for scoring the EAO. Mean scores can be combined on each of the four dimensions for each individual to constitute a total mean individual score or the mean score on the innovation dimension can be used as the primary indicator. The latter alternative was chosen for this study because the innovation dimension had the highest reliability factor in the instrument's testing; Cronbach coefficient alpha of .90 in comparison to .70 to .84 for the other three dimensions (Robinson et al., 1991).

Variables

The independent variables were the situations (entrepreneurial success, failure, and exit); the explanatory style dimensions of internality, stability, globality, and controllability; and the overall attributional explanatory style. Each of the four explanatory style dimensions had two levels: internality--internal and external; stability--stable and unstable; globality--global and specific; and controllability--controllable and uncontrollable. Attributional explanatory style was measured by using an adapted version of the ASQ (Peterson et al., 1982). The dependent variables reflected choices made concerning financial well-being, family relations, psychic well-being (self-esteem), and subsequent entrepreneurial activity (career opportunities).

Attributional Style Questionnaire (ASQ). The ASQ has been the primary instrument for measuring explanatory style. It consists of 12 hypothetical situations describing equal numbers of good and bad outcomes. The situations are equally representative of events that are achievement related and those that are of an interpersonal/affiliative nature. This composition built "cross-situational generality into the measure" (Peterson et al., 1982: 290).

With the original ASQ (which does not include the controllability dimension because the instrument was based on perceived uncontrollability by the respondents), respondents are asked to identify one major cause of each outcome and each cause is rated on a seven-point scale for degree of internality, stability, and globality. The dimensions are based on a continuum, not a dichotomy. The seven-point scales are anchored so that lower scores are received for the external, unstable, and specific attributions and higher scores are received for the internal, stable, and global attributions (Peterson et al., 1982; Tennen and Herzberger, 1985). Peterson et al. (1982) recommended not to delineate the scales into smaller subscales, but rather to use the composites for each dimension and the overall composites for good and bad events.

The adapted version of the ASQ used in this study involved adding the fourth dimension of controllability to the scale for each of the 12 hypothetical situations of good and bad events. Following the pattern of the scaling used with the other three dimensions, the seven-point scale was anchored so that a lower score reflected the uncontrollable attribution and a higher score correlated with the controllable attribution (this resulted after accommodating for reverse scoring). A mean was calculated for this dimension across all the events, separately for positive and negative events. Composite scores and the total score also reflected the inclusion of the controllability dimension.

Business venture scenarios. The third measurement instrument of the study was developed to operationalize the dependent variables--financial well-being, career opportunities, family relations, and self-esteem--and included three scenarios involving the success, failure, and exit of an entrepreneurial venture. These scenarios were constructed with the intent of letting the attributional style influence the perceived outcomes, rather than the outcomes being indicative in the scenario. This reflects the premise of attributional explanatory styles (Alloy et al., 1984).

The scenarios that were developed incorporated several important factors. Since outcomes related to the four areas of risk were the focus of this research, "ambiguous" information was included concerning two of these areas; specifically, those which had an interpersonal component rather than personal, for example, financial well-being and family relations. Financial support of the venture came from both personal and outside sources, but no degree of support was indicated. Concerning family relations, only that the family had an element of involvement was expressed. Personal inferences had to be made on the other two factors--self-esteem and continuing to pursue entrepreneurial activities (career opportunities). Self-esteem is a personal factor, so no reference was made in the scenario about self-confidence, self-worth, or any descriptors related to self-esteem. The decision to continue in the entrepreneurship arena is an outcome that is likely linked to multiple factors within the business venture experience. Reflections on how the experience affected financial well-being, family relations, and self-esteem should influence the next step in the individual's life.

The length of time in a business also influences the reasons for exiting a business. Ronstadt (1985) found that financial factors were more prevalent for early exits, and later exits were more influenced by personal and family factors. Being ambiguous with longevity in the business ("for a period of time") was important so as not to influence the outcomes in any particular direction.

In the failure scenario, "lack of success" rather than the word "failure" was used. The words "lack of success" hopefully allowed the inference that the venture did not live up to the individual's expectations. This point ties in with research noted earlier related to exit and threshold expectations of success (Cooper et al., 1992).

With the intended vagueness of the scenarios, the intent was for attributional explanatory style to influence the perceived outcomes. Each of the scenarios was identical with the exception of the situation--success (S), failure (F), or exit (E). Based on this framework, the following scenario represented the venture experience of failure (success and exit outcomes are not part of this study):

 You have owned your own business for a period of time and have
 decided to exit the entrepreneurial venture because of lack of
 success. This is your first venture experience and your family has
 had an element of involvement in the venture with you. Your venture
 was funded by internal (personal) and external financial resources.

Following each scenario, the format was consistent and followed that of the ASQ. The one major cause of each outcome was identified and each cause again was rated on a seven-point scale for degree of internality, stability, globality, and controllability.

The final aspect of the business venture scenarios instrument included 16 statements relating to the four dependent variables. Subjects responded to a 10-point scale anchored by strongly disagree (1) and strongly agree (10); there was a forced choice at midscale (slightly disagree [4] and slightly agree [5]). The 16 statements were culled from the numerous articles written by entrepreneurs who had failed and the descriptive research articles on ex-entrepreneurs that have been previously cited in this research. From the content of these sources, comments were extracted that related to each of the four areas--financial wellbeing, career opportunities, family relations, and self-esteem. A group of business students categorized these statements into the four respective categories. There was 100 percent agreement among the students about the statement classifications.

The content validity of the scale is reflected in that these statements were made by entrepreneurs who had failed and these were their personal comments about their experiences. Item-total correlation using Cronbach's alpha was used to test the internal consistency of each of the risk areas to assure that each scale item was tapping a single unidimensional construct. This analysis was done on the full sample (N = 336) and on the sample subset (n = 168). Consistent with both analyses, two statements were eliminated in order to improve the coefficient alpha. A statement was also deleted from the career opportunities scale because alpha was improved by .06 in the sample subset. Table 1 shows the items for the four scales and their Cronbach alphas.

Experiment

The instruments were divided into two packets for experimental purity. The first packet comprised the EAO and the ASQ instruments. The second packet included the business venture scenarios and subject profile information. Within each packet, the order of the instruments or the order of the scenarios was changed to test for order effects.

Subjects were provided the instrument packets in a classroom setting. All subjects were first given the EAO-ASQ packet; once returned, they were provided the second packet, which contained the business venture scenarios. All were asked to put some mark of identification (such as a favorite movie or a random symbol) on the top of their respective packets in order to match the two packets. With the return of each of the first and second packets, a check was done for incomplete data in order to reduce the number of missing values. A match was made of the two packets according to the identifier the subject used. A total of 364 matched packets were returned, with 28 being removed for incomplete data; final sample size was 336.

Assumptions Protocol and Statistical Analysis

Before analyzing the data, the tests for normality and homoscedasticity were completed on the full sample (N = 336) and the sample subset (n = 168). Testing the assumption of normality was done using several different means. The stem-and-leaf plot, the expected normal probability plot (or Q-Q plot), and the Kolmogorov-Smirnov test with Lilliefors significance correction were used to test for normality. The assumption of normality was upheld. Levene's Test of Equality of Error Variances was used to test the assumption of homoscedasticity. Levene's test was not significant for the criterion variable or for the independent variables (the attributional explanatory style dimensions) for either the full sample or the sample subset. Levene's test was significant for only one dependent variable, financial well-being, for both samples ([F.sub.2,333] = 4.637, p < .01; [F.sub.2,165] = 4.061,p < .05, respectively), concluding that the assumption of homoscedasticity was adequately present.

In order to test for order effects, the counterbalancing method was used (Malhotra, 1993). The instruments in the first packet were reversed for half of the sample (EAO-ASQ and ASQ-EAO). The order of the second packet placed each scenario in a respective first, second, and third position (success-failure-exit, failure-exit-success, and exit-success-failure). This resulted in six different orders. Order effects were initially tested on the EAO and ASQ instruments from the first packet and then the business scenarios (success-failure-exit) from the second packet using one-way ANOVA. Bonferroni's test was done as a post hoc multiple comparisons test to evaluate the differences between all possible pairs of group means. To better determine where there might be the presence of order effects, a tiered approach was taken. Analysis was first done with the six groups (based on the possible combinations of the EAO, ASQ, and business venture scenarios) for the EAO and ASQ instruments. For the full sample, the composite negative score was statistically significant ([F.sub.5,330] = 2.45, p < .05); for the sample subset, there were no significant differences between the mean scores of the EAO criterion variable (innovation) or the six composite scores from the ASQ of the six groups. Based on these results, the six groups were collapsed into three for further order effects analysis on the business venture scenarios. The three groups consisted of success-failure-exit, SFE (n = 112, full sample; n = 55, sample subset); failure-exit-success, FES (n = 121, full sample; n = 67, sample subset); and exit-success-failure, ESF (n = 103, full sample; n = 46, sample subset).

One-way ANOVA (see Table 2) was used to determine if the mean scores on the dependent variables (financial well-being, career opportunities, family relations, and self-esteem) were different based on the order the subjects were given of the business venture scenarios, with failure the scenario of interest. For the full sample, significant differences in the mean scores were found for all four of the dependent variables. Significant differences were found for three of the variables with the sample subset--financial well-being, career opportunities, and family relations. Bonferroni's multiple comparison test was subsequently done. Order effects were present to a greater degree in the full sample than in the sample subset. In the full sample, the order of SFE-FES significantly affected the mean scores of financial well-being (MD = -1.074, p < .0001), career opportunities (MD = -1.014, p < .0001), and self-esteem (MD = -.0513, p = < .05). When the success scenario was first rather than the failure scenario (either in the SFE-FES or FES-SFE comparison), the mean scores of these dependent variables concerning failure were significantly lower. When the failure scenario preceded the success scenario (either in the FES-ESF or ESF-FES comparison), mean scores for financial well-being (MD = 0.947, p < .0001), career opportunities (MD = 1.005, p < .0001), family relations (MD = 0.745, p < .01), and self-esteem (MD = 0.574, p < .01) were significantly higher. There were no effects present when comparing the SFE-ESF order, when failure always followed success.

An order effects pattern was less consistent with the sample subset and was only present in the financial well-being and family relations variables. When the success scenario preceded the failure scenario (either in the SFE-FES or FES-SFE comparison), mean scores were significantly lower for financial well-being (MD = -0.947, p < .01). Mean scores were significantly higher for financial well-being (MD = 1.102, p < .01) when the subjects responded to the failure scenario before the success scenario (either in the FESESF or ESF-FES comparison). Scores for family relations were significantly higher when failure preceded success (MD = 0.951, p < .05) in the FES-ESF or ESF-FES comparison. In one instance, when success always preceded failure (in the SFE-ESF or ESF-SFE comparison), family relations scores from the failure scenario were significantly higher (MD = 0.886, p < .05).

Since significant differences in scores were found when subjects responded to the success scenario prior to the failure scenario (lower scores), it was evident that their responses varied as a function of the success scenario. As a result, the FES group (failure, exit, success) was used for hypotheses testing to ensure there was no sequential confounding of task and order presentation.

The hypotheses were tested using the Pearson correlation coefficient which measured the degree and direction of the linear relation between each attributional explanatory style dimension and the composite style with each of the four areas of risk--financial wellbeing, career opportunities, family relations, and self-esteem. This was an appropriate statistical test because the independent variables (the subjects' explanatory style) cannot be controlled or manipulated so causation cannot be determined (Gravetter and Wallnau, 1992). Since a directional prediction was made in the hypotheses, a one-tailed test was done (which increases the statistical power).

Results

Sample Characteristics

Descriptive statistics concerning sample characteristics were run on both the full sample and the sample subset to determine if the sample profile remained consistent on all characteristics (excluding the EAO innovation criterion variable used for the determination of the sample subset).

With both the samples, characteristics of the subjects remained consistent across the groups, indicating that the primary distinction between the two groups was the median split of the entrepreneurial orientation dimension of innovation. The sample of interest, those who were above the median split on the innovation dimension, were generally 21 to 30 years of age and never married, with no greater proportion represented by either gender. Over 40% of their parents owned a business, with agriculturally based being the most dominant type. Very few of the subjects were business owners themselves, nine from the full sample and six from the sample subset, with services the most common type of business owned. Proportionately (compared with the other sample characteristics), there was a greater percentage (67%) of business ownership that carried over into the sample subset. Of those subjects owning a business, only one had previously owned a business and that individual had owned two other businesses prior to the present one. These businesses were generally service-related. The sample subset indicated a greater likelihood of owning a business in the future, 73% compared with 63%. This may be due to the fact that these individuals scored in the upper half of one of the dimensions addressing entrepreneurial orientation (the EAO scale).

Attributional Explanatory Style Dimensions

Results of a one-way ANOVA indicated no significant differences with the groups (SFE. FES. ESF) for any of the explanatory style dimensions, indicating any differences that would appear in the entrepreneurial risk areas across the groups would be a result of the business venture scenario order. The ASQ was completed prior to the business venture scenarios, so the attribution scores are not a result of which groups the subjects were in.

Hypotheses Testing

The four dimensions, along with the composite negative style and the overall explanatory style, will be addressed collectively for each of the risk areas for the FES group. Since the seven-point ASQ scale was anchored by low scores indicating external, unstable, specific, and uncontrollable dimensions, it was hypothesized that the relationships with the risk areas would be more positive with this attributional style; therefore, a lower dimension score would indicate this relationship. Results are found in Table 3.

Financial well-being. Three of the four attributional explanatory style dimensions were found to be significant. The correlations between external attribution, specific attribution, and uncontrollable attribution supported the hypotheses that these attributions were positively related to the perceived outcomes of financial well-being resulting from a venture failure. Meaning, subjects who had a lower score on these dimensions of the ASQ for bad events (reflecting external, specific, and uncontrollable attributions) responded more positively to statements that judged financial well-being. There was no significant correlation to support the hypothesis there was a positive relationship between unstable attributions and the effect of the venture failure on financial well-being.

There were two ways to assess a general explanatory style, those being a composite negative score and an overall explanatory style score. The composite negative relates only to bad events (that is, failure), while the overall explanatory style is a more generalized attribution style and reflects how an individual overall attributes causality to events. Both were used to test the hypothesis relating to overall style. The composite negative score was found to be significant, indicating a more external, unstable, specific, and uncontrollable attribution style concerning failure and more positive perceived outcomes related to financial well-being. There was a significant correlation of overall explanatory style with financial well-being; meaning across all situations (good and bad/success and failure), the dominant style is to have internal, stable, global, and controllable attributions when good things happen and external, unstable, specific, and uncontrollable attributions when bad things happen. The significant correlation indicates outcomes for financial well-being were perceived to be better for a failed entrepreneurial venture when the attributions were collectively external, unstable, specific, and uncontrollable, supporting the hypothesis.

Career opportunities. There were no significant correlations between any of the four attributional explanatory style dimensions and the perceived outcomes concerning career opportunities.

There was also no significant correlation between composite negative style and career opportunities. There was a significant correlation for overall explanatory style indicating the responses to statements concerning outcomes for career opportunities were more positive for a failed entrepreneurial venture when attributions were collectively, external, unstable, specific, and uncontrollable. Since this relationship was weak, it must be accepted with caution, as it is not being supported by the composite negative relationship.

Family relations. Two of the attributional explanatory style dimensions were found to be significant. The correlations between unstable attribution and specific attribution supported the hypotheses that these attributions were positively related to the perceived outcomes of family relations resulting from a venture failure. No significant correlations were found between external attribution and uncontrollable attribution regarding family relations.

The composite negative was close to the level of significance and in the hypothesized direction. There was a significant positive correlation of overall explanatory style with family relations, indicating the principal style is to have external, unstable, specific, and uncontrollable attributions when bad situations occur and, as a result of this, perceive the outcomes (family relations) related to the bad situation (entrepreneurial failure) to be more favorable.

Self-esteem. The correlations between external attribution, specific attribution and uncontrollable attribution and self-esteem scores were found to be statistically significant. This supported the hypotheses that these attribution dimensions were more positively related to more positive responses for assessing self-esteem with a failed venture. There was no significant correlation between unstable attribution and the effect of the venture failure on self-esteem.

A significant correlation was also found for the composite negative score and for overall explanatory style with self-esteem. This indicates that self-esteem was perceived to be better (higher self-esteem scores) for a failed entrepreneurial venture when the attributions were collectively external, unstable, specific, and uncontrollable than if attributions were the contrary, supporting the hypothesis.

Discussion and Limitations

Though there was not consistent support across the board for all areas of risk with all attributional dimensions, the strongest relationships were with the outcomes of financial well-being and self-esteem, with all dimensions being significant excluding stability. Previous research generally has not singled out any one attribution (with the noted exception of internality), so inferences about the stability dimension alone must be made with caution. But a plausible explanation may lie with the career opportunities data. Since no relationship was found for any of the individual dimensions with career opportunities it is possible that if the individual were not continuing his or her career in entrepreneurship, the unstable attribution would not be dominant. The stability dimension is based on whether an event is likely to occur in a similar situation in the future (i.e., another entrepreneurial venture). If the individual does not continue the entrepreneurial career path after venture failure, then the stability dimension's correlation with outcomes is questionable. Contrary, if the individual were to continue with another venture after failure, an unstable attribution should have been the prevailing style because it would not be desired for the cause of failure to "follow" the individual into the next venture and possibly preempt the failure of that business, too.

Peterson, Maier and Seligman (1993) noted that it may be difficult to abstract singular causal relationships concerning attribution because there could be many causes that confound with one another when trying to attach meaning to events. It is also noted that the internal-external dimension and the stable-unstable dimension have an orthogonal relationship (Weiner, 1986; Weiner et al., 1971) and that the globality dimension has an orthogonal relationship with the internality and stability dimensions (Abramson, Seligman and Teasdale, 1978). The inconsistency between external and unstable attributions and the perceived outcomes related to financial well-being, family relations, and self-esteem (the financial well-being hypothesis supported with external attribution and not supported with unstable attribution) aligns with the position of Weiner and his colleagues. Meaning, there is perfect nonassociation between the variables. For example, with the cause of the failure being temporary lack of effort, the attribution would be internal and unstable. If the cause was that this type of business has a higher failure rate than other types of businesses, then the attribution would be categorized as external and stable. Without tracking and analyzing the specific causes, it cannot be said that this was the case in this study. Along with this, the inconsistency concerning supporting or not supporting the hypotheses across the dimensions with financial well-being, family relations, and self-esteem (no hypotheses were supported for career opportunities) makes the point credible that overall explanatory style should serve as a better indicator for how attributions influenced perceived outcomes related to venture failure.

Because of the possible difficulty in abstracting singular causal relationships concerning attribution, as noted by Peterson, Maier and Seligman (1993), overall explanatory style may be more influential for how individuals interpret events and behaviors and make causal explanations for answering why things happen. The preferred dominant styles of internal, stable, global, and controllable attributions for good events (success) and external, unstable, specific, and uncontrollable attributions for bad events (failure) would be evidenced in higher scores for overall explanatory style. This favored style when attributing causality to bad events would make the individual less prone to chronic depression and helplessness, which would influence the individual's perception of his or her personal and professional life.

Across all areas of risk--financial well-being, career opportunities, family relations, and self-esteem as identified by Liles (1974)--an external, unstable, specific, and uncontrollable attributional explanatory style resulted in more positive perceived outcomes related to these risk areas. These findings support the premise that individuals may have an overall explanatory style when attributing causes of failure (Abramson, Seligman and Teasdale, 1978; Alloy et al., 1984; Burns and Seligman, 1989; Nurmi, 1992; Peterson and Seligman, 1984), and based on this attribution of causality, judgments and responses are affected. When this style was present, the individual's financial outlook was better, there was a greater likelihood to continue along the entrepreneurial career path, family relations were enhanced, and the individual felt good about his or her performance and what learning opportunities were presented even though the venture met with a lack of success.

Early prior research (Brockhaus, 1985; Meyer, Zacharakis and De Castro, 1993; Ronstadt, 1981, 1982, 1984, 1985, 1986; Wicker and Conn, 1990), though descriptive and narrative in nature, presents a diversity of findings concerning how failure has affected the ex-entrepreneur. Later research (Patzelt and Shepherd, 2011; Politis and Gabrielsson, 2007; Shepherd, 2003, 2004; Shepherd, Wiklund and Haynie, 2009; Ucbasaran et al., 2010) is paying greater attention to the emotion, stress, and coping mechanisms present when failure occurs. For some, their financial status, relations with family members, and self-esteem were impaired as a result of the failure, and they were ending their entrepreneurial career path. For others, the contrary was true. This study draws a link between why the attitudes of these individuals may vary based on their attributional explanatory style.

A limitation of the study to note is regarding a possible influence of the findings concerning career opportunities. The Cronbach alpha of .58 for the statements that reflected career opportunities was lower than for the other risk areas-financial well-being, family relations, and self-esteem (these were in the .64 to .75 range for the sample subset). Improvement of the alpha coefficient may have found significant findings in the relationship of each of the four attributional explanatory style dimensions with career opportunities (yet, the overall explanatory style was found to be significant).

The study was designed to extend or falsify attribution theory and explanatory style to the context of entrepreneurial venture failure. Intentionally, the sampling method used was purposive nonprobability sampling, an appropriate method for testing theory. The primary limitation of this method is the lack of generalizability to the population of interest, ex-entrepreneurs. Ultimately, further extension of the relationships found depends on testing the relationships with ex-entrepreneurs as the sample.

Implications for Professional Practice or Applied Settings

Businesses are born and businesses die. As the title of Brockhaus' (1985) work on ex-entrepreneurs is so aptly put, "Is There Life After Death?" Many entrepreneurs who have left one venture, whether by choice or not, proceed to undertake another. They do not attribute failure to themselves as much as they do to situational factors. They persevere, become resilient, and take their lessons learned into the next venture.

Learning about the next chapter in the ex-entrepreneurs' lives has had problems because of the difficulty in locating them after the venture failure. This should not deter the continuance of this area of research. Rather this "... supports the need for continued efforts to better research the experiences of ex-entrepreneurs after they cease doing business. In this way, future ex-entrepreneurs can better assess the consequences of failure" (Brockhaus, 1985: 476).

Locating ex-entrepreneurs is one solution to the problem (and as noted, is a problem in itself). But this study implies there is another alternative. If causal explanations affect the way an individual responds to situations, such as failure, then knowing the entrepreneur's attributional style prior to venture startup can imply how the entrepreneur may react if the venture fails.

There is value in understanding how entrepreneurs or ex-entrepreneurs attribute the causes of a failed venture. First, it would enable new entrepreneurs to go in with their eyes open with no false expectations. A business can start with much aforethought and planning or hastily with the generally misguided thoughts that money can be made quickly and the entrepreneur can have an easier schedule than with current employment. With both of these instances, often the entrepreneur does not assess all that is at risk besides personal time and money. Entrepreneurs, in creating the venture, indeed assume a level of risk, but often do not perceive to be taking such great risks because they believe so intensely in their ideas and that they will succeed. Yet, the fear of failure and its resultant consequences are ever present in the mindsets of entrepreneurs (Harrell, 1994). Running a business affects the family, self-esteem, and the individual's career path. Knowing one's attributional explanatory style will provide insight into what the responses may be after venture failure concerning the impacts on other aspects of the entrepreneur's life.

Second, when a venture fails the entrepreneur may decide to proceed with another venture start-up. Erroneous attributions can lead to actions that fail to correct the problem(s) in another entrepreneurial situation. In fact, those actions could even intensify the problem. Considering the limited resources with most venture start-ups (whether that be financial, time with family, or the expertise gained from managing a previous business), inaction or taking the wrong action could lead to failure for the first venture and continued failure for subsequent ventures. Knowing how explanatory style affects determining causation for events could improve decision-making and thereby new venture survival.

Third, as educators, as consultants, and as policy advisors, helping entrepreneurs understand that the decisions made and actions taken during the business venture can have far reaching effects beyond the scope of the business. Rather than always taking a retrospective look, attributional explanatory style may provide a look into how outcomes may be perceived if a venture does not succeed; it can imply how the entrepreneur may react if the venture fails.

References

Abramson, L.Y., M.E.P. Seligman, and J.D. Teasdale. 1978. "Learned Helplessness in Humans: Critique and Reformulation." Journal of Abnormal Psychology 87: 49-74.

Ahmed, S.U. 1985. "nAch, Risk-taking Propensity, Locus of Control and Entrepreneurship." Personality and Individual Differences 6: 781-782.

Alloy, L.B., C. Peterson, L.Y. Abramson, and M.E.P. Seligman. 1984. "Attributional Style and the Generality of Learned Helplessness." Journal of Personality and Social Psychology 3: 681-687.

Berkowitz, L., and E. Donnerstein. 1982. "External Validity is more than Skin Deep." American Psychologist 37(3): 245-257.

Brenner, O.C., C.D. Pringle, and J.H. Greenhaus. 1991. "Perceived Fulfillment of Organizational Employment Versus Entrepreneurship: Work Values and Career Intentions of Business College Graduates." Journal of Small Business Management 29(3): 62-74.

Brockhaus, R.H., Sr. 1980. "Risk Taking Propensity of Entrepreneurs." Academy of Management Journal 23: 509-520.

Brockhaus, R.H., Sr. 1985. "Is There Life After Death? The Impact of Unsuccessful Entrepreneurial Endeavors on the Life of the Entrepreneurs." In J.A. Hornaday, E.B. Shils, J.A. Timmons, and K.H. Vesper (eds.), Frontiers of Entrepreneurship Research (pp. 468-481).

Bunker, K.A., and A.D. Webb. 1992. Learning How to Learn from Experience: Impact of Stress and Coping (Rep. No. 154). Center for Creative Leadership: Greensboro, NC.

Burns, M.O., and M.E.P. Seligman. 1989. "Explanatory Style Across the Life Span: Evidence for Stability Over 52 Years." Journal of Personality and Social Psychology 56: 471-477.

Calder, B.J., L.W. Phillips, and A.M. Tybout. 1981. "Designing Research for Application." Journal of Consumer Research 8(2): 197-207.

Calder, B J., L.W. Phillips, and A.M. Tybout. 1983. "Beyond External Validity." Journal of Consumer Research 10(1): 112-114.

Chandler, T.A., C.A. Seibel, and C.J. Spies. 1990. "Attributions and Affects Related to Actual vs Perceived Success in an Actual Achievement Context." Psychological Reports 66: 971-976.

Cooper, A.C., T. Folta, J. Gimeno-Gascon, and C.Y. Woo. 1992. "Entrepreneurs' Exit Decisions: The Role of Threshold Expectations." In J.L. Wall and L.R. Jauch (eds.), Best Papers Proceedings, Academy of Management (pp. 75-79).

Crant, J.M. 1996. "The Proactive Personality Scale as a Predictor of Entrepreneurial Intentions." Journal of Small Business Management 34(3): 42-49.

Crittenden, K.S. 1989. "Causal Attribution in Sociocultural Context: Toward a Self-presentational Theory of Attribution Processes." The Sociological Quarterly 30(1): 1-14.

Curren, M.T., V.S. Folkes, and J.H. Steckel. 1992. "Explanations for Successful and Unsuccessful Marketing Decisions: The Decision Maker's Perspective." Journal of Marketing 56(2): 18-31.

Darley, J., and J. Cooper. 1998. Attribution and Social Interaction: The Legacy of Edward E. Jones. The American Psychological Association: Washington, DC.

Dyer, W.G., Jr. 1992. The Entrepreneurial Experience: Confronting Career Dilemmas of the Start-up Executive. Jossey-Bass Publishers: San Francisco.

Fletcher, G.J.O., P. Danilovics, G. Fernandez, D. Peterson, and G.D. Reeder. 1986. "Attributional Complexity: An Individual Differences Measure." Journal of Personality and Social Psychology 51: 875-884.

Gravetter, F.J., and L.B. Wallnau. 1992. Statistics for the Behavioral Sciences (3rd ed.). West Publishing Company: St. Paul, MN.

Harrell, W. 1994. For Entrepreneurs Only: Success Strategies for Anyone Starting or Growing a Business. Career Press: Hawthorne, NJ.

Harvey, J.H., and G. Weary. 1984. "Current Issues in Attribution Theory and Research." Annual Review of Psychology 35: 427-459.

Heider, F. 1958. The Psychology of Interpersonal Relations. Lawrence Erlbaum Associates: Hillsdale, NJ.

Hornaday, J.A., and J. Aboud. 1971. "Characteristics of Successful Entrepreneurs." Personnel Psychology 24(2): 141-153.

Kelley, H.H. 1967. "Attribution Theory in Social Psychology." In D. Levine (ed.), Nebraska Symposium on Motivation, Volume 15 (pp. 192-240). University of Nebraska Press: Lincoln.

Kelley H.H. 1973. "The Processes of Causal Attribution." American Psychologist 28: 107-128.

Kernis, M.H., J. Brockner, and B.S. Frankel. 1989. "Self-esteem and Reactions to Failure: The Mediating Role of Overgeneralization." Journal of Personality and Social Psychology 57: 707-714.

Koh, H.C. 1996. "Testing Hypotheses of Entrepreneurial Characteristics: A Study of Hong Kong MBA Students." Journal of Managerial Psychology 11(3): 12-25.

Krueger, N.F. 1993. "'Growing up entrepreneurial?' Developmental Consequences of Early Exposure to Entrepreneurship." In D.P. Moore (ed.), Best Papers Proceedings, Academy of Management (pp. 80-84).

Liles, P.R. 1974. New Business Ventures and the Entrepreneur. Richard D. Irwin, Inc.: Homewood, IL.

Malhotra, N.K. 1993. Marketing Research: An Applied Orientation. Prentice Hall: Englewood Cliffs, NJ.

Masters, R., and R. Meier. 1988. "Sex Differences and Risk-taking Propensity of Entrepreneurs." Journal of Small Business Management 26(1): 31-35.

McGrath, J.E., and D. Brinberg. 1983. "External Validity and Research Process: A Comment on the Calder/Lynch Dialogue." Journal of Consumer Research 10(1): 115-124.

Meyer, G.D., A.L. Zacharakis, and J. De Castro. 1993. "A Postmortem of New Venture Failure: An Attribution Theory Perspective." In N.C. Churchill, S. Birley, J. Doutriaux, E.J. Gatewood, F.S. Hoy, and W.E. Wetzel, Jr. (eds.), Frontiers of Entrepreneurship Research (pp. 256-269).

Mikulincer, M. 1986. "Attributional Processes in the Learned Helplessness Paradigm: Behavioral Effects of Global Attributions." Journal of Personality and Social Psychology 51: 1248-1256.

Minniti, M., and W. Bygrave. 2001. "A Dynamic Model of Entrepreneurial Learning," Entrepreneurship Theory and Practice 25: 5-16.

Mitton, D.G. 1989. "The Compleat Entrepreneur." Entrepreneurship Theory and Practice 13(3): 9-19.

Mook, D.G. 1983. "In Defense of External Invalidity." American Psychologist 38: 379-387.

Naffziger, D.W., J.S. Hornsby, and D.F. Kuratko. 1994. "A Proposed Research Model of Entrepreneurial Motivation." Entrepreneurship Theory and Practice 18(3): 29-42.

Nurmi, J. 1992. "Cross-cultural Differences in Self-serving Bias: Responses to the Attributional Style Questionnaire by American and Finnish Students." The Journal of Social Psychology 1: 69-76.

O'Connor, D.J., and D.M. Wolfe. 1987. "On Managing Midlife Transitions in Career and Family." Human Relations 40: 799-816.

Palmer, M. 1971. "The Application of Psychological Testing to Entrepreneurial Potential." California Management Review 13(3): 32-38.

Parks, G.M. 1977. "How to Climb a Growth Curve: Eleven Hurdles for the Entrepreneur-manager, Part II." Journal of Small Business Management 15(2): 41-45.

Patzelt, H., and D.A. Shepherd. 2011. "Negative Emotions of an Entrepreneurial Career: Self-employment and Regulatory Coping Behaviors." Journal of Business Venturing 26: 226-238.

Peterson, C., S.F. Maier, and M.E.P. Seligman. 1993. Learned Helplessness: A Theory for the Age of Personal Control. Oxford University Press: New York.

Peterson, C., and M.E.P. Seligman. 1984. "Causal Explanations as a Risk Factor for Depression: Theory and Evidence." Psychological Review 91: 347-374.

Peterson, C., A. Semmel, C. von Baeyer, L.Y. Abramson, G.I. Metalsky, and M.E.P. Seligman. 1982. "The At tributional Style Questionnaire." Cognitive Therapy and Research 6: 287-300. Politis, D., and J. Gabrielsson. 2007. "Entrepreneurs' Attitudes Towards Failure--An Experiential Learning Approach." Frontiers of Entrepreneurship Research 27(6). Available online: http://papers.ssrn.com/sol3/papers. cfm?abstract_id=1064982.

Popper, K. 1963. Conjectures and Refutations: The Growth of Scientific Knowledge. Harper Torchbooks: New York.

Rhodewalt, F., C. Morf, S. Hazlett, and M. Fairfield. 1991. "Self-handicapping: The Role of Discounting and Augmentation in the Preservation of Self-esteem." Journal of Personality and Social Psychology 61: 122-131.

Robinson, P.B., D.V. Stimpson, J.C. Huefner, and H K. Hunt. 1991. "An Attitude Approach to the Prediction of Entrepreneurship." Entrepreneurship Theory and Practice 15(4): 13-31.

Ronstadt, R. 1981. "Entrepreneurial Careers and Research on Entrepreneurs." In K.H. Vesper (ed.), Frontiers of Entrepreneurship Research (pp. 591-600).

Ronstadt, R. 1982. "Does Entrepreneurial Career Path Really Matter?" In K.H. Vesper (ed.), Frontiers of Entrepreneurship Research (pp. 540-567).

Ronstadt, R. 1984. "Ex-entrepreneurs and the Decision to Start an Entrepreneurial Career." In J.A. Hornaday, F.A. Tardley, Jr., J.A. Timmons, and K.H. Vesper (eds.), Frontiers of Entrepreneurship Research (pp. 437-460).

Ronstadt, R. 1985. "Every Entrepreneur's Nightmare: The Decision to Become an Ex-entrepreneur and Work for Someone Else." In J.A. Hornaday, E.B. Shils, J.A. Timmons, and K.H. Vesper (eds.), Frontiers of Entrepreneurship Research (pp. 409-434).

Ronstadt, R. 1986. "Exit, Stage Left: Why Entrepreneurs End Their Entrepreneurial Careers Before Retirement." Journal of Business Venturing 1: 323-338.

Scott, M.G., and D.F. Twomey. 1988. "The Long-term Supply of Entrepreneurs: Students' Career Aspirations in Relation to Entrepreneurship." Journal of Small Business Management 26(4): 5-13.

Seligman, M.E.P., and P. Schulman. 1986. "Explanatory Style as a Predictor of Productivity and Quitting Among Life Insurance Sales Agents." Journal of Personality and Social Psychology 50: 832-838.

Sellers, P. 1995. "So You Fail. Now Bounce Back!" Fortune 131(8): 48-66.

Sexton, D.L., and N.B. Bowman. 1983a. "Comparative Entrepreneurship Characteristics of Students: Preliminary Results." In J.A. Hornaday, J.A. Timmons, and K.H. Vesper (eds.), Frontiers of Entrepreneurship Research (pp. 213-232).

Sexton, D.L., and N.B. Bowman. 1983b. "Determining Entrepreneurial Potential of Students." In F.H. Chung (ed.), Proceedings, Academy of Management (pp. 408-412).

Sexton, D.L., and N.B. Bowman. 1984. "Personality Inventory for Potential Entrepreneurs: Evaluation of a Modified JPI/PRF-E Test Instrument." In J.A. Hornaday, F.A. Tardley, Jr., J.A. Timmons, and K.H. Vesper (eds.), Frontiers of Entrepreneurship Research (pp. 513-528).

Sexton, D.L., and N.B. Bowman. 1986. "Validation of a Personality Index: Comparative Psychological Characteristics Analysis of Female Entrepreneurs, Managers, Entrepreneurship Students and Business Students." In R. Ronstadt, J.A. Hornaday, R. Peterson, and K.H. Vesper (eds.), Frontiers of Entrepreneurship Research (pp. 40-51).

Shepherd, D.A. 2003. "Learning from Business Failure: Propositions of Grief Recovery for the Self-employed." Academy of Management Review 28(2): 318-328.

Shepherd, D.A. 2004. " Educating Entrepreneurship Students About Emotion and Learning from Failure." Academy of Management Learning and Education 3(3): 274-287.

Shepherd, D.A., J. Wiklund, and J.M. Haynie. 2009. "Moving Forward: Balancing the Financial and Emotional Costs of Business Failure." Journal of Business Venturing 24: 134-148. Tennen, H., and S. Herzberger. 1985. "Attributional Style Questionnaire." In D.J. Keyser and R.C. Sweetland (eds.), Test Critiques, Vol. 4. (pp. 20-30). Test Corporation of America: Kansas City.

Ucbasaran, D., P. Westhead, M. Wright, and M. Flores. 2010. "The Nature of Entrepreneurial Experience, Business Failure and Comparative Optimism." Journal of Business Venturing 25: 541-555.

Weiner, B. 1979. "A Theory of Motivation for Some Classroom Experiences." Journal of Educational Psychology 71: 3-25.

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

Weiner, B., I.H. Frieze, A. Kukla, L. Reed, S. Rest, and R.M. Rosenbaum. 1971. Perceiving the Causes of Success and Failure. Silver Burdett/ General Learning Press: Morristown, NJ.

Weiner, B., H. Heckhausen, W. Meyer, and R.E. Cook. 1972. "Causal Ascriptions and Achievement Behavior: A Conceptual Analysis of Locus of Control." Journal of Personality and Social Psychology 21: 239-248.

White, P.A. 1991. "Ambiguity in the Internal/External Distinction in Causal Attribution." Journal of Experimental Social Psychology 27: 259-270.

Wicker, A.W., and L. Conn. 1990. "The Missing Persons of Entrepreneurship Research: Owners of Discontinued Businesses." In L.R. Jauch and J.L. Wall (eds.), Best Papers Proceedings, Academy of Management (pp. 74-78).

Williams, C.W., P.R. Lees-Haley, and R.S. Brown. 1993. "Human Response to Traumatic Events: An Integration of Counterfactual Thinking, Hindsight Bias, and Attribution Theory." Psychological Reports 72: 483-494.

Mary K. Askim-Lovseth, Department of Marketing, University of North Dakota

Richard A. Feinberg, Consumer Sciences and Retailing, Purdue University

Contact

For further information about this article, contact:

Mary K. Askim-Lovseth, Department of Marketing, University on North Dakota, Grand Forks, ND, 58202-8366

E-mail: maskim@business.und.edu

Tel: 701-777-2930 (office); 701-777-2225 (fax)

Table 1. Scale Items and Cronbach's Alpha for the Four Scales

 Cronbach's
 Alpha

Scale Items Total Total
 Sample (a) Subset (b)

Financial Well-being .63 .64

 I have been able to at least
 recover my original investment
 with the venture.
 I have been able to develop
 financial contacts
 as a result of this venture that
 will help me in the future.
 My financial future does
 not look very positive. (c)

Career Opportunities .59 .58

 I intend to continue my
 entrepreneurial career.
 I am ready to get into another
 line of work. (c, d)
 I found my place with this
 business.

Family Relations .75 .75

 The amount of emotional
 support from my family has
 been limited. (c, d)
 The communication within
 my family has improved
 as a result of this
 undertaking.
 My family has developed a
 closer relationship because
 of this business.
 My family felt terribly
 restricted being tied down
 to the business. (c)
 This endeavor has put a strain
 on family relations. (c)

Self-esteem .75 .72

 I feel that I have worked
 hard in the business.
 This has been a positive
 learning experience for me.
 I still feel confident about
 my abilities.
 I am concerned about what
 others think about me.(c, d)
 I feel I have much to be proud of.

(a) N = 336. (b) n = 168. (c) Reverse-scored items. (d) Deleted from
scale.

Table 2. Analysis of Variance Summaries for
the Perceived Outcomes of Venture Failure
Concerning Order Effects for Full Sample and
Sample Subset

Source SS df

Financial Well-being

Between Groups 80.292 2
Within Groups 783.014 333

Career Opportunities

Between Groups 78.904 2
Within Groups 1170.221 333

Family Relations

Between Groups 31.618 2
Within Groups 883.328 333

Self-Esteem

Between Groups 22.964 2
Within Groups 659.243 333

Note. N = 336, full sample.

Financial Well-being

Between Groups 42.302 2
Within Groups 482.470 165

Career Opportunities

Between Groups 28.827 2
Within Groups 678.649 165

Family Relations

Between Groups 8.514 2
Within Groups 524.039 165

Self-Esteem

Between Groups 10.220 2
Within Groups 335.844 165

Source MS F

Financial Well-being

Between Groups 40.146 17 073 ****
Within Groups 2.351

Career Opportunities

Between Groups 39.452 11 23 ****
Within Groups 3.514

Family Relations

Between Groups 15.809 5.96 **
Within Groups 2.653

Self-Esteem
Between Groups 11.482 5.80 **
Within Groups 1.980

Note. N = 336, full sample.

Financial Well-being

Between Groups 21.151 7.233 ***
Within Groups 2.924

Career Opportunities

Between Groups 14.414 3.504 *
Within Groups 4.113

Family Relations

Between Groups 14.257 4.489 *
Within Groups 3.176

Self-Esteem

Between Groups 5.110 2.511
Within Groups 2.035

Note. n = 168, sample subset

* p < .05. ** p < .01. *** p < .001.
**** p < .0001.

Table 3. Descriptive Statistics and Correlation Coefficients of the
Attributional Explanatory Style Dimensions with the Risk Factors

 Dimensions (a)

Risk Factors M SD I S G

Financial Well-being 7.189 1.370 -.219 * -.103 .223 *
Career Opportunities 6.351 1.754 .040 .003 -.131
Family Relations 6.261 1.776 -.073 -.250 * -.209 *
Self-esteem 8.056 1.320 -.217 * -.036 -.209 **

 Dimensions (a)

Risk Factors C CoNeg Overall

Financial Well-being -.248 * -.276 * .283 *
Career Opportunities -.094 -.065 .252 *
Family Relations -.025 -.200 .304 **
Self-esteem -.306 ** -.296 ** .386 ***

Note. n = 67. [sup.a.ASQ] dimensions of internality (I), stability
(S), globality (G), controllability (C), composite negative (CoNeg),
and overall explanatory style (Overall). Lower scores for CoNeg
indicate an external, un-stable, specific, and uncontrollable
explanatory style for bad outcomes. Higher scores for overall
explanatory styles indicate internal, stable, global, and controllable
attributions for good situations and external, unstable, specific, and
uncontrollable attributions for bad situations.

* p < .05, one-tailed. ** p < .01, one-tailed. *** p <.001,
one-tailed.