SEEING WHAT'S NOT THERE: THE ENIGMA OF ENTREPRENEURSHIP JoAnn C. Carland James W. Carland Western Carolina University Wayne H. Stewart University of North Texas ABSTRACT The process is clear: entrepreneurs initiate business ventures. What is not clear is why they do so. The debate continues to rage about entrepreneurial behavior and this singular act of individual volition which is so vital to a nation's economic health and well being. The drives and personalities continue to be debated. Gartner (1988) asks, "Can one know the dancer from the dance?" Is it even important to try? Carland, Hoy and Carland (I 988) think it is essential because one cannot understand the dance without understanding the dancer. We think that the dance takes on the personality of the dancer. It is the dancer who interprets the dance and each artist makes the process his or her own. If we seek to understani the entrepreneurial process, we must have some insight into the entrepreneurial psyche. This is especially true if we wish to design educational and training programs for prospective and practicing entrepreneurs. In this work, we empirically examine 502 owner/managers of small businesses. We identify entrepreneurial vision, the ability to see what is not there, as their commonality. We empirically/ink that vision to the entrepreneurial psyche and use that to build insight into the entrepreneurial enigma, the process of entrepreneurship. INTRODUCTION The process is clear: entrepreneurs initiate business ventures. What is not clear is why they do so. The debate continues to rage about entrepreneurial behavior and this singular act of individual volition which is so vital to a nation's economic health and well being. The drives and personalities continue to be debated. Gartner (1988) asks, "Can one know the dancer from the dance?" Is it even important to try? Carland, Hoy and Carland ( 1988) think it is essential because one cannot understand the dance without understanding the dancer. We think that the dance takes on the personality of the dancer. Each artist makes the process his or her own. The outcome of the process of entrepreneurship is obvious for all to see: the creation of a new venture. That the venture may be the culmination of the dreams ofa life time seems unimportant in the face of the incontrovertible: the tangible outcome of the process. But we cannot lose sight of an another fact that is beyond debate: the trigger of the act is an individual. The key is that individual, the initiating force, the one who sees the opportunity, the challenge, and the one who takes that challenge. If we would understand the enigma of entrepreneurship, we must begin to circumscribe the behavior of the entrepreneur. Herron and Sapienza ( 1992) avow that the individual entrepreneur is the most salient unit of analysis in entrepreneurship research and theory. The value of the phenomenon of entrepreneurship is well established, both in terms of economic vitality and research and development. We may be interested in supporting and facilitating the process in the interests of national well being, economic development, or advancing the standard of living. We may be concerned about designing and administering entrepreneurial education and training programs to inculcate the skills and abilities required for success. In either event, but particularly so in the latter case, we must begin by understanding the initiator: the entrepreneur. How can one design a training program or a curriculum of entrepreneurship education if one does not understand the drives and characteristics which lead to the decision to initiate a venture, to concentrate on its growth, to take it public, to strive to dominate an industry? To teach the dance, one must teach the dancer. Here, then, is our effort at understanding the dancer, at unraveling the enigma of entrepreneurship. CHARACTERISTICS OF ENTREPRENEURS Much of the research in entrepreneurship has been founded upon the premise that entrepreneurs embody distinctive personality characteristics which can be identified (Cooper & Dunkel berg, 1987), and used to indicate a potential for entrepreneurship (Lachman, 1980). These approaches have been criticized because they tend to be difficult to operationalize (Gartner, 1988). Nevertheless, one must approach the explication of a gestalt by describing its contributory factors. ·The earliest identified entrepreneurial characteristic was risk taking. Cantillion (circa 1700) portrayed an entrepreneur as the individual who assumed the risk for the firm (Kilby, 1971 ), a perspective echoed by Mill (1848). Palmer(l 971) proffered that risk assessment and risk taking are the primary elements of entrepreneurship. Risk includes not only financial considerations, but also career opportunities and family relations (Liles, 1974). Yet, researchers are undecided about the role of the risk taking propensity of entrepreneurs (Brockhaus, 1987). Some studies have indicated no significant differences in risk taking propensities for entrepreneurs as compared to the general population (Brockhaus, 1980; Sexton & Bowman, 1983). Others have discovered a higher propensity for risk taking among entrepreneurs (Sexton & Bowman, 1986; Carland, Carland, Carland & Pearce, 1995), particularly when confronted with business risk (Ray, 1986), but moderated by business experience, age, education, and type of business (Schwer & Yucelt, 1984). Research has also shown that entrepreneurs evidence low uncertainty avoidance irrespective of culture (McGrath, MacMillan, & Scheinberg, 1992). Schumpeter (1934) posited that the burden of risk was inherent in ownership, and since entrepreneurs were not necessarily owners, the propensity for assuming risk should not be included as an entrepreneurial trait. Instead, according to Schumpeter, the central characteristic of entrepreneurship should be innovation. 2 . ,r- 11' ;.. : • ,. ·~-' • \ Schumpeter's view of entrepreneurial innovation was rooted in the classic theories of economists such as Say and Marshall (Hornaday, 1992). In the literature, innovation remains a frequently identified functional characteristic of entrepreneurs (e.g., McClelland, 1961; Hornaday & Aboud, 1971; Timmons, 1978; Brockhaus, 1982; Carland, Hoy, Boulton & Carland, 1984; Gartner, 1990). Timmons(l 978) suggested that creativity and innovation were conditions inherent in the role of entrepreneurship. Drucker (1985) actually defined entrepreneurshipas innovation in a business setting as the entrepreneur generates new capacity for wealth from limited resources. Olson (1985) included invention, an activity analogous to innovation, as a primary entrepreneurial activity. This contention was intensified by Carland, Hoy, Boulton and Carland (1984) who proposed that innovation was the critical factor in distinguishing entrepreneurs from managers and small business owners. Hornaday ( 1992) deftly illustrated that while innovation is a necessary element of entrepreneurship, alone it is insufficientto fully circumscribeentreprenetrial behavior because of the broad parameters of the function. Despite the often stated significance of creativity and innovation vis-a-vis entrepreneurs, relatively few studies have empirically investigated the proposed relationship. Perhaps the most ubiquitous entreprereurial characteristic is the need for achievement. This insight was initiated by the work of McClelland (1961 ). In a study of behavior in young men, McClelland (1961, 1965) concluded that a high need for achievement would influence the self selection of an 'entrepreneurial' position, defined as a salesman, company officer, management consultant, fund-raiser, or owner of a business. Thus, these studies did not actually link need for achievement with the founding or ownership of a business. Numerous subsequent studies have shown a positive relationship between achievement motivation and entrepreneurship (Hornaday & Bunker, 1970; Hornaday & Aboud, 1971; DeCarlo& Lyons, 1979; Lachman, 1980; Begley & Boyd, 1986). Other studies have shown that need for achievement is not the most important variable for predicting the likelihood of starting a business (Borland, 1974; Hull, Bosley, & Udell, 1980). Johnson (1990) suggested that because of the variabilityofthe samples, different operationalizations of the achievement motive, and convergent validity problems in instrumentation, more research is necessary to prove a definitive link between achievement motivation and entrepreneurship. The ability to identify and solve problems seems endemic to the entrepreneurial process. Jung ( 1971) posited that one's view of a problem is a function of how one perceives the world and assesses information. Jung labeled perception modes as sensation or intuition, and thought processes as either thinking or feeling. According to Jung, decision making that is based upon the thinking mode is methodical, while decision making based upon feeling is characterized by impulsiveness. Myers and Briggs (1962) extended the original work of Jung to develop further the orientation toward perception and judgment. A preference for one mode over another was considered to be an attitude. Four attitudes resulted: extroversion versus introversirn and perception versus judgment. These attitudes combined with four functions, sensation versus intuition and thinking versus feeling, produce sixteen permutations of preferences known as personality types or cognitive styles. These typologies are indicated by the Myers-BriggsType Indicator (MBTI). Keirsey and Bates (1984) used the MBTI types to identify four primary temperaments which represent the major cognitive distinctions among people. 3 Carland and Carland (1992), drawing largely upon the work of Jung (1971 ), Myers and Briggs ( 1962), and Keirsey and Bates ( 1984), analyzed the problem-solvingstyles of managers, entrepreneurs,and small business owners. Using the Keirsey and Bates temperaments of SP, SJ, NF, and NT, the authors explored the differences of innovation, risk and achievement by cognitive type and concluded that temperament did indeed go far toward explicating the phenomenon of entrepreneurship. THE ENTREPRENEURIAL GESTALT Hornaday's observation about the inability of innovation to circumscribe entrepreneurship(l992) is a result of the gestalt nature of the phenomenon. Even ifneed for achievement, preference for innovation and risk taking behavior are endemic to the entrepreneurial psyche, operationalizing the insight is problematic because each trait is normally distributed. To illustrate, Figures I, 2 and 3 display graphs of the distributions of scores for the group of502 small business owners which we will be examining in this study. The scores illustrated are those on established instruments which measure the need for achievement, the preference for innovation and the propensity for risk taking. Figure 1 Need for Achievement Need for Achievement Score Figure 3 Risk Taking Propensity Risk Taking Score Figure 2 Praforence for Innovation ~ .. nr:=====-i "m--~ ;.".....--- !!! .. LL Innovation Score The graphs may not visually resemble the bell shaped curve, but a statistical examination is required to determine whether the three distributions are normally distributed. For this purpose, we applied the Kolmogorov- Smimov test under the Lilliefors option (Wilkinson, 1990). This stat1st1c assumes no prior distribution but standardizes the variables and tests whether the standardized versions are normally distributed. The results, displayed in Exhibit I, show a high probability that the scores are normally distributed. 4 EXHIBIT I Kolmogorov-Smimov Test for Standard Normal Distribution Variable N Maximum Difference Lilliefors Probability Achievement Score 502 .181 .000 Innovation Score 502 .140 .000 Risk Taking Score 502 .085 .000 The conceptthat behavior traits of entrepreneurs are normally distributed is far reaching. It implies that prediction of individual behavior will be complicated by the relative strength of that individual'spersonalityorientatirn. This difficulty may well be the primary source of the confusion of results which seem to confound the advancement of the discipline. It is not surprising that different samples of entrepreneurs can be examined with totally different outcomes if one recognizes that the members of that sample may lie anywhere in a broad distribution of trait strength. This problem led us to focus this research on an empirical analysis of the interactions of the classic portrait of an entrepreneur. If entrepreneurship is a gestalt, we must study it as such in order to grasp the significance and practical applicability of our findings. We immediately face two basic questions in this quest. Are there any commonalities in entrepreneurial behavior? Is there some unifying perspective that has the potential to provide a structure for researching the phenomenon? The literature is rich in both process and trait work. Although the discipline seems at times disjointed, there is one perspective that seems to us to be common to all of the insights produced over the years: entrepreneurial vision. Whether we call it innovation and creativity, or the process of creating a venture, the commonality is that all ofus recognize that the entrepreneur had the ability to see what is not there. The vision is the key. It is the insight to identify an under-served market; the intuition to design new products, services or methods which can capture markets; the sixth sense that leads to an understanding of time, place, product and market. It is entrepreneurial vision that guides the act of volition which culminates in all of the phenomenon which we study: the creation of a venture; the guidance and nurture of a venture; and, the growth and development of a venture. The most promising approach to examining entrepreneurial vision, we felt, was the use of cognitive typologies. Recognizing that a major aspect of an individual's temperament is intuition, we decided to investigate whether typologies had the potential to form the structure for examining the entrepreneurial gestalt. To that end, we employed the Myers-Briggs Type Indicator (Myers & Briggs, 1962) to the subjects in this study and partitioned them into the four major temperaments which Keirsey and Bates ( 1984) espouse: NF, intuitive-feeling; NT, intuitive-thinking; SJ, sensingjudging; and, SP, sensing-perceptive. The immediate question is do the four temperaments display differences in the strength of the key traits, need for achievement, preference for innovation, or risk taking propensity, which are so well establisha:I in the literature? If they do, then we may have found a basis for understanding the interaction of the established personality traits in supporting entrepreneurial vision. 5 As a matter of fact, the temperaments do indeed, display differentstrengthson these traits. Pictured in Figures4, 5 and 6 are graphs illustrating that the two intuitive temperaments, the NF and NT cognitive groups of small busine&5 owners displayed higher scores on the need for achievement, preference for innovation and risk taking propensity than did the sensing based temperaments. Figure 5 Innovation Preference by Temperament e 0 11 ~,. f c .. :8 u .. " ~ 11 c" .E Temperament Figure 4 Need for Achievement by Temperament e 0 .. 11 ~ u c .. ~ u " ~ 12 " :c 10 ~ Temperament Figure 6 Risk Taking by Temperament Temperament This rather unscientific insight led us to choose cognitive typology as a structure for investigating the entrepreneurshipgestalt. That is, we intend to employ cognitivetemperamert as a foundation for statistical examination of the need for achievement, preference for innovation and risk taking propensity. We are now ready to begin our investigation of the enigma of entrepreneurship. Recognizing that advancing our understanding requires an empirical foundation, we designed a research methodology to pursue the heffalump (Hull, Bosley & Udell, 1980). RESEARCH METHODOLOGY The researchers in this study designed an instrument which contained established measures of the need for achievement, preference for innovation, risk taking propensity, and cognitive typology. Demographic and strategic questions rounded out the survey. Graduate business students selected the participants of the study on a convenience basis. The students solicited responses from employers, employers of their parents, acquaintances, or from individuals with whom they had some other form of contact. Participants in the study came from 30 states, however, most respondents lived in the Southeastern United States. Although the sample is convenience in nature, there are several benefits from this sampling technique. First, the sample was not anonymous, and the data set was controlled. 6 The questionnaires were examined upon submission, and incomplete questionnaires could be returned for completion. The lack of anonymity also ensures that the appropriate individual in the business actually completed the survey. Second, the rate of response was greater than that of the typical mail survey. Less than one in twenty individuals who were approached declined to participate in the study, suggesting that individuals participated in the survey who might not otherwise have responded. Therefore, while still existent, nonresponse bias is not as problematic as with the typical mail survey. Third, the technique supported the ability to generate a large sample size. The sample includes 502 active owner-managers of small businesses. The central limit theorem (Mason, 1982) suggests that, due to the sample size, the level of confidence of this sample approaches that ofa random sample. The businesses selected for study fit the Small Business Administration guidelines, i.e., a small business is independently owned and operated, and not dominant in its field. Moreover, the numberof employees and volume of sales of the firms complied with the Small Business Administration's guidelines for assistance. Consequently, every individual in the study was the principal owner and manager of a qualifying small business. Demographic information concerning the individual respondents and size and industrycharacteristicsoftheir firms is displayed in Exhibit 2. The Achievement Scale of the Personality Research Form (Jackson, 1974) was used to measure the need for achievement. Jackson (1974) reported that the test-retest reliability (N=l35) was .80, and odd-even reliability (N=l92) was reported to be .77. Jackson and Guthrie (1968), testing for validity, reported correlations with self ratings and peer ratings of .65 and .46, respectively. The authors concluded that the instrument contained convergent and discriminant validity. Risk taking propensity and preference for innovation were measured using the Risk Taking Scale and Innovation Scale of the Jackson Personality Inventory (Jackson, 1976). For risk taking propensity, Jackson (1976) tested the internal consistency reliability with two samples (N=82 and N=307), and reported values of .93 and .91 using Bentler's coefficient theta, and .81 and .84 using coefficient alpha. Testing for validity (N=70), Jackson (1976) reported correlations with the completion ofan adjective checklist, with selfrating and peer ratingof.75, .77, and .20 respectively. The internal consistency reliability of the Innovation Scale produced values of .94 and .93 using Bentler's coefficient theta, and .83 and .87 using coefficient alpha. Validity was checked using the completion of an adjective checklist, with self rating, and peer rating of .79, .73, and .37, respectively. Reliability for the instruments pertaining to risk taking propensity, preference for innovation, and need for achievement were analyzed in the current study using Cronbach's Alpha. The alphas were .76, .77, and .72, respectively. These scores suggest that the instruments accurately measure the characteristics, and that the individual items on the tests produce comparable patterns of responses over all cases. 7 EXHIBIT2 Demographic Data (N = S02) Type of Company Retail 19% Wholesale 3% Manufacturing 43% Construction 2% Service 29% Organization Proprietorship 79% Partnership 11% Corporation 10% Sales Less than $ IOOK 12% $100K to $SOOK 10% $SOOK to $IM 6% Over$1M 64% Employees Less than 10 24% IO to SO 11% SI to 100 6% IOI to 2SO S% Over 2SO 49% Age of the Owner Under 2S 4% 2S to 3S 33% 36 to 4S 36% 46 to SS 20% Over SS 6% Sex of the Owner Male 76% Female 24% Race of the Owner Majority 96% Minority 4% Education of the Owner Less than 12 years 3% 12 years 19% 13to!Syears 17% 16 years 46% Over 16 years 13% The 32-item forced-choiceshort form of the Myers-Briggs Type lndicator(MBTl)(Myers & Briggs, 1962) was included to measure the cognitive typology of the respondents. Mendelsohn (I 96S) reported that the TF (thinking versus feeling), SN (sensation versus intuition), and El (extroversion versus introversion) scales are independent, while the JP (judging versus perceiving) scale is consistently correlated with the SN scale. Internal 8 consistency reliabilities were reported in the range from .75 to .85, and a 14-month test-retest correlation of. 70 was reported. As advised by Myers and McCaulley ( 1985), the MBTI scores were converted to continuous distributions for each of the four pairs of characteristics. These scores are the basis for the determination of the four fundamental temperaments. Scores less than 100 imply a preference for the first letter in the scale, while scores greater than 100 indicate a preference for the second letter in the scale. This conversion also supports the use of the scores in parametric statistical operations. RE SUL TS OF THE ANALYSIS Using the MBTI scores to partition the data set resulted in identifying NFs, NTs, SJs, and SPs in proportions displayed in Exhibit 3. Interestingly, these proportions do not fit the national population distribution of38% SJs, 38% SPs, 12% NFs and 12% NTs (Keirsey & Bates, 1984 ). Small business owners may not be a mirror of the national population. NTs are represented in this sample at more than twice the expected level. SJs and NFs are slightly higher herethan in the national population, while SPs are dramatically under represented. This may have to do with the relative satisfaction of various temperaments under normal conditions of employment. Temperament SJs SPs NFs NTs EXHIBIT 3 Distribution of MBTI Scores Number Percent 222 44% 56 11% 77 15% 147 29% Normal Percent 38% 38% 12% 12% The first step in the statistical analysis was the examination of the scores which each of the four temperament groups produced on each of the three personality profile instruments. That is, we examined the scores for need for achievement, preference for innovation, and risk taking propensity, for each of the groups, NFs, NTs, SJs, and SPs. The first step was to examine each of the distributions for normality. As was the case with the overall scores, the individual distributions were normal, with one exception. Exhibit 4 shows the results of the Kolmogorov-Smimov test under the Lilliefors option (Wilkinson, 1990) for each of the four temperaments. As the table shows, all three instruments produced normally distributed data for each of the four temperament groups except for risk taking among SPs. 9 EXHIBIT4 Kolmogorov-Smirnov Test Variable N Maximum Difference Lilliefors Probability NFs Achievement Score 77 .186 .000 Innovation Score 77 .175 .000 Risk Taking Score 77 .116 .012 NTs Achievement Score 147 .186 .000 Innovation Score 147 .188 .000 Risk Taking Score 147 .079 .024 SJs Achievement Score 222 .177 .000 Innovation Score 222 .106 .000 Risk Taking Score 222 .119 .000 SPs Achievement Score 56 .163 .001 Innovation Score 56 .118 .049 Risk Taking Score 56 .109 .097 The second step in the statistical analysis was a search for differences in the scores which each of the four temperament groups produced on each of the three personality profile instruments. Descriptive statistics are displayed in Exhibit 5. The mean scores in the table are more revealing in light of the range of scores which each of the instruments produces. The need for achievement instrument produces a range of scores from 0 to 16. The preference for innovation scale produces a range of scores from 0 to 20. The risk taking propensity instrument also produces a range of scores from 0 to 20. The table shows that the two intuitive groups, NFs and NTs, displayed higher mean scores on all three instruments than did the sensing groups. The descriptive statistics do not indicate whether differences in scores are statistically significant, consequently, the next phase of the investigation involved Analysis of Variance (ANOV A) on the scores of each of the instruments for each of the temperaments. The results of those statistics are displayed in Exhibit 6. The table shows that the scores on each of the three instruments are significantly different for the various cognitive temperaments. In fact, the level of significance is quite high. 10 EXHIBIT 5 Descriptive Statistics N MEAN SCORE VARIANCE STD DEVIATION NFs Achievement 77 13.260 6.353 2.520 Innovation 77 15.013 17.671 4.204 Risk Taking 77 11.779 26.806 5.177 NTs Achievement 147 16.605 4.542 2.131 Innovation 147 16.517 9.430 3.071 Risk Taking 147 12.980 19.705 4.439 SJs Achievement 222 12.707 7.131 2.670 Innovation 222 12.599 21.906 4.680 Risk Taking 222 7.874 22.645 4.759 SPs Achievement 56 12.607 8.788 2.965 Innovation 56 12.214 18.171 4.263 Risk Taking 56 8.482 20.181 4.492 EXHIBIT 6 Analysis of Variance Achievement N=502 Squared Multiple R: 0.026 Source Sum of Squares DF Mean Square F Ratio p Temperament 85.789 3 28.596 4.443 .004 Error 3205.246 498 6.436 Innovation N=502 Squared Multiple R: 0.159 Source Sum of Squares DF Mean Square F Ratio p Temperament 1621.830 3 540.610 31.450 .000 Error 8560.443 498 17.190 Risk Taking N=502 Squared Multiple R: 0.196 11 Source Sum of Squares DF Mean Square F Ratio p Temperament 2680.718 3 893.573 40.349 .000 Error 11028.636 498 22.146 The last finding led us to examine whether the actual cognitive scores could shed any more light on the entrepreneurship function. Accordingly, we turned to converted MBTI scores for El, SN, TF and JP. These scores, as reported above, had been converted into continuousdistributionsto support statistical examination (Myers & Mccaulley, 1985). The first step in the examination was a correlation statistic. Exhibit 7 displays a Pearson correlatirn matrix and the statistical significance for each cell. EXHIBIT 7 Pearson Correlation Matrix Ach Inn Risk EI SN TF JP Achievement 1.000 Innovation 0.375 1.000 Risk Taking 0.241 0.414 1.000 EI Score -0.146 -0.249 -0.294 1.000 SN Score 0.176 0.485 0.510 -0.336 1.000 TF Score -0.095 -0.092 -0.084 -0.063 O.I56 1.000 JP Score 0.070 0.164 0.352 -0.113 0.506 0.228 I 1.000 Bartlett Chi-Square Statistic: 697.28I DF=21 Probability= .000 MATRIX OF PROBABILITIES Ach Inn Risk EI SN TF JP Achievement 0.000 Innovation 0.000 0.000 Risk Taking 0.000 0.000 0.000 EI Score 0.001 0.000 0.000 0.000 SN Score 0.000 0.000 0.000 0.000 0.000 TF Score 0.033 0.040 0.060 O.I58 0.000 0.000 JP Score 0.I 18 0.000 0.000 0.012 0.000 o.ooo I 0.000 12 As the table shows, scores on the need for achievement were significantly correlated with the El, SN, and TF scales of the MBTI, as well as the scores for preference for innovation and risk taking propensity. In fact, the preference for innovation was significantly correlated with all of the other measures. Risk taking propensity scores were also significantly correlated to everything except the TF scale. From the perspectiveofthe temperamert scores, the SN scale performed the best in terms of significance for all three of the personality trait instruments. These findings led us to speculate about the ability of the cognitive temperament to drive the personality traits. The theoretical underpinning of our empirical analysis involved an understanding of the value ofintuition in explaining entrepreneurial vision. The SN scale of the MBTI explicitly measures the degree of a respondent'sreliance upon intuition in his or her cognitive processes. Accordingly, we conducted regression analyses employing the three personality traits as dependent variables and the SN score as independent variable. The results are displayed in Exhibit 8. EXHIBIT 8 Regression Analyses Dependent Variable: ACHIEVEMENT Squared Multiple R: 0.031 Variable Coefficient Std Error Std CoefTolerance T p Constant 11.040 0.516 0.000 21.414 .000 SN 0.021 0.005 0.176 .IOOE+OI 3.989 .000 Analysis of Variance Source Sum of Squares OF Mean Sq F Ratio p Regression IOI.477 I 101.477 15.909 .000 Residual 3176.465 498 6.378 Dependent Variable: INNOVATION Squared Multiple R: 0.235 Variable Coefficient Std Error Std CoefTolerance T p Constant 4.332 0.805 0.000 5.383 .000 SN 0.102 0.008 0.485 .IOOE+Ol 12.378 .000 Analysis of Variance Source Sum of Squares OF Mean Sq F Ratio p Regression 2381.691 I 2381.691 153.222 .000 Residual 7740.957 498 15.544 Dependent Variable: RISK TAKING Squared Multiple R: 0.260 Variable Coefficient Std Error Std CoefTolerance T p Constant -1.848 0.918 0.000 -2.013 .045 SN 0.124 0.009 0.510 .IOOE+ol 13.241 .000 Analysis of Variance 13 Source Sum of Squares OF Mean Sq F Ratio p Regression 3544.617 I 3544.617 175.324 .000 Residual 10068.333 498 20.218 As the table shows, the SN scale was a significant predictor of the need for achievement, preference for innovation and risk taking propensity. The scale only explained 3% of the variance in the need for achievement, but it explained 23% of the variance in preference for innovation scores and 26% of the variance in risk taking propensity scores. The regression coefficients were positive except for risk taking: the greaterthe level of intuition, the stronger the need for achievement and preference for innovation. However, stronger sensing preferenoe is associated with higher risk taking propensity. This last finding is at odds with the earlier descriptive statistics which suggest that intuitives have higher mean scores on risk taking than do sensing groups. There is one final set of statistics which come to mind. If the SN score is so successful at predicting scores on the need for achievement, preference for innovation, and risk taking propensity scales, how do the other cognitive meters perform? To grapple with that question, we conducted a stepwise regression matching the four MBTI scales to the three personality trait measures. The results are displayed in Exhibit 9. Stepwise regression is not an analytical technique which can produce conclusions because it ignores correlations among independent variables as they are entered into the regression equation. Nevertheless, stepwise regression can show whether additional study is justified. If the change in percentage of variance explained by a model is significant as the result of adding independent variables, that suggests that the additional variables merit further investigation. An insignificant change in explained variance suggests that additional variables are relatively unimportant to a model. The table in Exhibit 9 shows that, in comparison to the other scales, the SN scale is the dominant determinant of all three personality traits. The SN scale accounted for 26% of the variance in risk taking in step I, while the remaining three steps of the procedure only increased the R Square to 33%. The story is the same with innovation. The SN scale explainoo 24% of the variance in the innovation score, while three more regression steps could only raise the explanatory power to 28%. With regard to need for achievement, the SN scale only accounted for 3% of the variance in the instrument, however, two more regression runs could only raise the R Square to 6%. Clearly, the SN scale is the most important cognitive factor in understanding the personality traits. We cannot lose sight of the fact that the percentage of variance explained by the SN cognitive scale is small. Explaining 26% of the variance means that 74% of the variance is unexplained. In other words, there are other factors, not included in this analysis which are clearly important in determining the strength of the independent variables in these models. This is an important consideration and one which requires additional research to understand. 14 EXHIBIT 9 STEPWISE REGRESSION ANALYSES Achievement Step# I R Square: 0.031 Variable Coefficient Std Error Std CoefTolerance F p SN 0.021 0.005 0.176 .IE+.01 15.909 .000 Step# 2 R Square: .046 SN 0.023 0.005 0.196 0.97554 19.462 .000 TF -1.016 -0.006 -0.126 0.97554 8.066 .005 Step# 3 R Square: .055 SN 0.019 0.006 0.162 0.86885 12.029 .001 TF ·0.016 -0.006 -0.127 0.97541 8.264 .004 El -0.014 -0.006 -0.100 0.88708 4.612 .032 Innovation Step# I R Square: .235 SN 0.102 0.008 0.485 .IE+OI 153.22 .000 Step# 2 R Square: .264 SN 0.!08 0.008 0.512 0.97554 172.69 .000 TF ·0.039 -0.009 -0.172 0.97554 19.48 .000 Step# 3 R Square: .273 SN O.IOI 0.009 0.479 0.86885 135.84 .000 TF -0.039 -0.009 ·0.173 0.97541 19.92 .000 El -0.024 -0.0IO -0.099 0.88708 5.98 .015 Step# 4 R Square: .276 SN 0.108 0.010 0.514 0.66375 119.96 .000 TF -0.037 .0.009 -0.162 0.94508 17.03 .000 JP -0.016 -0.011 ·0.070 0.71674 2.39 .122 El -0.023 -0.0IO ·0.095 0.88221 5.41 .020 Risk Talcing Step# I R Square: .260 SN 0.124 0.009 0.5!0 .IE+.01 175.32 .000 Step# 2 R Square: .288 SN 0.131 0.009 0.537 0.97554 196.01 .000 TF -0.044 -0.0IO -0.168 0.97554 19.20 .000 Step# 3 R Square: .307 SN 0.11 I 0.011 0.457 0. 74172 I I0.97 .000 TF ·0.051 -0.0IO -0.193 0.94564 25.26 .000 JP 0.044 0.012 0.165 0.72069 14.01 .000 Step# 4 R Square: .328 SN 0.098 0.011 0.400 0.66375 78.24 .000 TF ·0.052 -0.0IO ·0.197 0.94508 26.93 .000 JP 0.048 0.012 0.177 0.71674 16.59 .000 El ·0.042 -0.011 ·0.153 0.88221 15.12 .000 15 CONCLUSIONS Have we been able to explicate the entreprereurship enigma? Not completely, but ifthe results of this study are confirmed by future research, we have made progress in .describing the dancer and understanding the dance. The results of this study suggest that entrepreneurs are not homogenous. They may well be characterized by need for achievement, preference for innovation and risk taking propensity, but some of them are more highly driven than others. In fact, any given group of entrepreneurs is likely to contain such a distribution of individuals which makes drawing conclusions about their personality traits difficult. This problem may well be the basis for those who espouse abandoning trait research because it provides no insight into the entrepreneurial process. Despite the problems which normally distributed trait strengths produce for researchers, there is much of value to be gained from understanding the entrepreneurial psyche. The process ofany and all entrepreneurial action is the result of an individual'sdecision to take that action. That decision is rooted in personality and cognition. We must gain knowledge of that personality in order to support, train and educate the entrepreneur. The results of this study suggest that the core insight which can support an understanding of the entrepreneur is intuition. In every statistical test, the cognitive function of intuition served to form a basis for understanding the behavior patterns. Those entrepreneurs with stronger intuition translate that vision into innovative action. They are supported in the drive to activate the vision by high need for achievement. They are less dismayed in the face of risk and may well see their actions as less risky. They see what is not there and see it extremely well. Those entrepreneurs with less intuition in their cognitive typologies will prefer a more concrete approach to the entrepreneurial process. They may be less creative than their intuitive brothers and sisters, but they will be more practical in their approaches to business. It is well established that sensing typologies make better managers but intuitive types make better changeagents(Keirsey& Bates, 1984). Both can be highly driven by need for achievement, but sensing types are more cognizant of risk and generally less risk taking. They see what is not there less well and are more apt to focus on the concrete and the here and now. The foregoing descriptions are simply of individuals at the two poles of the intuition continuum. In reality, most entrepreneurs will fall somewhere between those poles. What they see best will be what they focus upon and will form the basis for their individual approaches to the process of entrepreneurship. The process will be directed by the entrepreneurial vision. The depth and breadth of that vision varies along a normal distribution. Consequently, in any given group of entrepreneurs we are likely to find a cross section of people with various intuitive strengths and levels of drives. This does not mean that we do not understand the dancer. In fact, it means that we understand the dancer quite well. Like the original artist from whom we borrow this imagery (Yeats, 1956) each dancer will interpret the dance differently. To teach them we must test their insight. Those with stronger intuitive leanings will benefit more from educational programs grounded in reality and based on cold, hard facts. Those with stronger sensing orientation will benefit more from educational programs designed to foster and support paradigm bridging actions. The former need structure 16 ....... for their vision and insight, the latter need vision and insight to which they can apply their structure. Both can dance, but both can benefit from wise programs which mitigate their weaknesses and concentrate on how to make the best decisions for their future success. The entrepreneurial gestalt is truly that: a whole which is much greater than the sum of its parts and an outcome which transcends its inputs. We may never master the enigma, but we must recognize that to be ignorant of it is to forever limit our insight. Absent an understandingofthe entrepreneurial psyche we are left with attempting to interpret outcomes and processes through our own cognitive typologies. Those of us with more sensing orientation are likely to view the entrepreneurial process more mechanistically and to focus on the planning and strategic management processes. Those of us with more intuitive leanings are more likely to interpret entrepreneurial behavior as the result of leaps of logic which sometimes defy description. Like the blind men describing the elephant, each of us will be right about some things, but none of us will see the whole. If intuition is the glue which supports entrepreneurial vision, as this research suggests, then it is even more difficult than describing the elephant. We are describing the heffalump, a creature no one has ever seen (Hull, Bosley & Udell, 1980), and we must approach the creature carefully because each of them will be different and each of them will be unique and each of us will only be able to see one small aspect of the whole. It is this uniqueness that makes entrepreneurs the same and makes them so fascinating. That is the true enigma of entrepreneurship, the Gordian knot of our discipline. If we would understand these dancers, if we would interprettheir dances, if we would explore how they see what is not there, then we must look for the commonalities in our work rather than the differences. Each of us contributes another piece to the puzzle. As yet, we cannot know the pattern, we cannot see the portrait, but working with and building on each other, we will solve the enigma. It is our collected minds and works which will serve as Alexander's sword. 17 REFERENCES Begley, T. & Boyd, D. ( 1986). A comparison of entrepreneurs and managers of successful firms. Journal of Management, 13(1), 99-108. Birley, S. (1987). New venture and employment growth. Journal of Business Venturing, 2(2), 155-165. Boden, R.J. ( 1987). The role of small business in efficient resource allocation. In The State of Small Business. Washington, DC: U.S. Government Printing Office, 105-132. Borland, C. ( 1974 ). Locus of control, need for achievement and entrepreneurship. Unpublished doctoral dissertation, University of Texas at Austin. Brockhaus, R.H. (1980). Risk taking propensity of entrepreneurs. Academy of Management Journal, 23(3), 509-520. Brockhaus, R.H. (1982). The psychology of the entrepreneur. In C. Kent, D. Sexton & K. Vesper (Eds.), Encyclopedia of Entrepreneurship. Englewood Cliffs: Prentice-Hall, 39-57. Brockhaus, R.H. (I 987). Entrepreneurial folklore. Journal of Small Business Management, 25(3), 1-6. Carland, J.W., Hoy, F., Boulton, W.R. & Carland, J.A. ( 1984). Differentiating entrepreneurs from small business owners: A conceptualization. The Academy of Management Review, 9(2), 354-359. Carland, J.W., Hoy, F. & Carland, J.A. (1988). Who is an entrepreneur? Is a question worth asking. American Journal of Small Business, 12(4), 33-39. Carland, J. W. & Carland, J.A. ( 19~2). Managers, small business owners, and entrepreneurs: The cognitive dimension. Journal of Business and Entrepreneurship, 4(2), 55-62. Carland,J.W., III, Carland, J.W., Carland, J.A. & Pearce, J.W. (1995). Risk taking propensity among entrepreneurs, small business owners, and managers. Journal of Business and Entrepreneurship, 7( I), 15-23. Cooper, A. & Dunkelberg, W. (1987). Entrepreneurial research: Old questions, new answers and methodological issues. American Journal of Small Business, 11(3), 11-23. De Carlo, J. & Lyons, P. ( 1979). A comparison of selected personal characteristics of minority and non-minority female entrepreneurs. Journal of Small Business Management, 17(4), 22-29. Drucker, P. (1985). Innovation and entrepreneurship. New York, NY: Harper & Row. Gartner, W. (1988). Who is an entrepreneur? ls the wrong question. American Journal of Small Business, 12(4), 11-32. Gartner, W. (1990). What are we talking about when we talk about entrepreneurship?. Journal of Business Venturing, 5(1 ), 15-28. Glebes, M.C. ( 1987). The small business contribution to the service sector. In The State of Small Business. Washington, DC: U.S. Government Printing Office, 271-302. Herron,C. & Sapienza, H.J. (1992). The entrepreneur and the initiation of new venture launch activities. Entrepreneurship: Theory & Practice, 17(1 ), 49-55. Hornaday, J. ( 1992). Thinking about entrepreneurship: A fuzzy set approach. Journal of Small Business Management, 30(4), 12-23. Hornaday, J. & Aboud, J. (1971). Characteristics of successful entrepreneurs. Personnel Psychology, 24, 141-153. Hornaday,J. & Bunker, C. (1970). The nature of the entrepreneur. Personnel Psychology, 23, 47-54. 18 ~···-,~-~-----c .. c-,c-,---c-c----------------- Hull, 0., Bosley, J. & Udell, G. (1980). Reviewing the hefTalump: Identifying potential entrepreneurs by personality characteristics. Journal of Small Business Management. 18(1), 11-18. Jackson, D.N. (1974). Persona/ityresearchform manual. Goshen, NY: Research Psychologists Press. Jackson, D.N. (1976). Personality inventory manual. Goshen, NY: Research Psychologists Press. Jackson, D.N. & Guthrie, S.M. ( 1968) A multitrait-multimethod evaluation of the personality research form. Proceedings of the American Psychology Association, 177-178. Johnson, B. (1990). Toward a multidimensional model of entrepreneurship: The case of achievement motivation and the entrepreneur. Entrepreneurship: Theory and Practice, 14(3), 39-54. Jung, C. (1971 ). Psychological types. (H.G. Baynes, Trans., revised by R.F.C. Hull). Princeton, NJ: Princeton University Press (Original work published in 1921). Keirsey, D. & Bates, M. (1984). Please understand me. Del Mar, CA: Prometheus Nemesis Books. Kilby, P. ( 1971 ). Entrepreneurship and economic development. New York, NY: Free Press. Lachman, R. (1980). Toward measurement of entrepreneurial tendencies. Management International Review, 20(2), I 08-116. Liles, P.R. (1974). New business ventures and the entrepreneur. Homewood, IL: Richard D. Irwin. McClelland, D. (1961 ). The achieving society. Princeton, NY: Van Nostrand. McGrath, R.G., MacMillan, I.C., & Scheinberg, S. (1992). Elitists, risk-takers, and rugged individualists? An exploratory analysis of cultural differences between entrepreneurs and non-entrepreneurs. Journal of Business Venturing, 7(2), 115-135. Mendelsohn, G.A. (1965). The myers-briggs type indicator. In Buras (Ed.), The sixth mental measurements yearbook. Highland Park, NJ: The Gryphon Press. Mill, J.S. (1848). Principles of political economy. London, England: John W. Parker. Myers, J.B. & Briggs, K.C. (1962). The myers-briggs type indicator. Princeton, NJ: Educational Testing Service. Myers, J.B. & Mccaulley, M.H. (1985). Manual: A guide to the development and use of the myers-briggs type indicator. Palo Alto, CA: Consulting Psychologists Press. Olson, P. (1985). Entrepreneurship: Opportunistic decision makers. Journal of Small Business Management, 11 (2), 25-31. Palmer, M. (1971). The application of psychological testing to entrepreneurial potential. California Management Review, 13(3), 32-39. Ray, D.M. (1986). Perceptions of risk and new enterprise formation in singapore: An exploratory study. In R. Ronstadt, J. Hornaday, R. Peterson, and K. Vesper (Eds.), Frontiers of Entrepreneurship Research. Wellesley, MA: Babson College, 119-145. Reynolds, P.O. (1987). New firms: Societal contribution versus survival potential. Journal of Business Venturing, 2(3), 231-246. Schumpeter, J. (1934). The theory of economic development. Translated by R. Opie. Cambridge, MA: Harvard University Press. Schwer, K. & Yucelt, U. (1984). A study of risk-taking propensities among small business entrepreneurs and managers: An empirical evaluation. American Journal of Small Business, 8(3), 31-40. 19 Sexton, D.L. & Bowman, N .B. ( 1983). Comparativeentrepreneurshipcharacteristicsof students. In J. Hornaday, J. Timmons, and K. Vesper (Eds.), Frontiers of Entrepreneurship Research. Wellesley, MA: Babson College, 465-478. Sexton, D.L. & Bowman, N.B. (1986). Validation of a personality index: Comparative psychological characteristics analysis of female entrepreneurs, managers, entrepreneurshipstudents and business students. In R. Ronstadt, J. Hornaday, R. Peterson, and K. Vesper(Eds.),FrontiersofEntrepreneurship Research. Wellesley, MA: Babson College, 40-57. Timmons, J. (1978). Characteristics and role demands of entrepreneurship, American Journal of Small Business, 3( I), 5-17. Whiston, W.B. (1987). The State of Small Business. In The State of Small Business. Washington, DC: U.S. Government Printing Office, 1-64. Wilkenson, L. (1990). SYSTAT: Thesystemforstatistics. Evanston, IL: SYSTAT, Inc. Yeats, W .B ( 1956). Among school children. In The collected poems of WB. Yeats. New York: Macmillan. 20 l !