FIGURE 2-5
Tests for univariate equality of group means

 Predictor variables F Significance Father's occupation 10.2414 0.0016 ** Entrepreneurial experience 16.0815 0.0001 *** Family help experience 14.8418 0.0002 *** Neighborhood help experience 3.2030 0.0751 Overseas experience 0.7660 0.3825 Family tragedies 3.4416 0.0651 Illness, injury 2.8976 0.0903 Award experience 7.6705 0.0062 **

* p < 0.05
** p < 0.01
***p < 0.001

As a result of tests for univariate equality of group means, it was found that, four variables out of the eight - "father's occupation," "entrepreneurial experience," "family help experience," and "award experience" - produced different mean values between corporate employees aspiring to become entrepreneurs (Group 2.1) and those not aspiring to become entrepreneurs (Group 2.2).

Next, the stepwise method was used to identify the useful predictor variables. Then, we found the following three predictor variables in Figure 2-6 as the discriminant variables.

FIGURE 2-6
Pooled within-groups correlations
between discriminating variables and canonical discriminant functions

 Rank Variable Func 1 1 Family help experience 0.73906 2 Entrepreneurial experience 0.61011 3 Father's occupation 0.54742

As a result, the variable "family help experience" had the highest correlation with discriminant functions, indicating that a large percentage of corporate employees aspiring to become entrepreneurs have had the experience of helping the family. This was followed by "entrepreneurial experience" and "father's occupation."

Figure 2-7 shows the results of the classification results by the Mahalanobis' distance using the three aforementioned variables.

FIGURE 2-7
Classification results

 Actual group No. of cases Predicted group membership 2.1 2.2 Group 2.1:Corporate employees aspiring to become entrepreneurs 84* 5160.7%) 33 (39.3%) Group 2.2:Corporate employees not aspiring to become entrepreneurs 113* 25(22.1%) 88(77.9%)

Percent of "grouped" cases correctly classified: 70.56%

* After processing 216 cases, 19 cases were excluded from the analysis by at least one missing predictor variables or the variable that defines the groups.

Of 84 cases in Group 2.1 (Corporate employee aspiring to become an entrepreneur), 51 were predicted correctly to be members of Group 2.1 (60.7%), while 33 (39.3%) were assigned incorrectly to Group 2.2 (Corporate employee not aspiring to become an entrepreneur). Similarly, 88 out of 113 cases (77.9%) were identified correctly, and 25 (22.1%) were misclassified. The overall percentage of cases classified correctly was 70.56% (139 out of 216).

The classification results for corporate employees aspiring to become entrepreneurs (Group 2.1) and those not aspiring (70.56%) was higher than the classification results for entrepreneurs (Group 1) and corporate employees (64.93%). This is because some people had no choice but to become entrepreneurs due to changes in the environment and other factors. Thus, it is understandable that, from the perspective of forming an image of an entrepreneur, corporate employees aspiring to become entrepreneurs had higher classification results than people who are entrepreneurs.

Importance of Pseudo-Entrepreneurial Experiences

From the above two analyses, the predictor variables that contributed to discriminant functions can be made into matrices as shown in Figure 2-8.

FIGURE 2-8
Matrices that contribute to variables, based on two discriminant analyses

 Predictor variable Discriminant analysis of Group 1 and Group 2 Discriminant analysis of Group 2.1 and Group 2.2 Relationship with 3 stages in forming an image of entrepreneurs Father's occupation l n Role model of parents Entrepreneurial experience £ l Pseudo-entrepreneurial experience Family help experience £ Pseudo-entrepreneurial experience Family tragedies r Extraordinary experience Award experience n Extraordinary experience

lfirst place, £second place, nthird place, and rfourth place

Through two discriminant analyses, it was found that pseudo-entrepreneurial experiences contributes the most to discriminant functions followed by role model of parents and extraordinary personal experiences. As "fathers' occupation" is a given and not something people can choose of their own will, it is obvious that pseudo-entrepreneurial experiences is an important factor in forming one's image of an entrepreneur.

"Family help experience," which contributed the most in discriminating groups of corporate employees aspiring to become entrepreneurs from those not aspiring, produced degree analysis data that indicated that 63.0% of the corporate employees' fathers were corporate employees themselves, and a small percentage of corporate employees' fathers were entrepreneurs or self-employed. Therefore, it is easy to understand that "family help experience," rather than role models, plays an important role in forming one's image of an entrepreneur.

From these findings, we know that a large number of entrepreneurs and corporate employees aspiring to become entrepreneurs have had pseudo-entrepreneurial experiences, indicating the tremendous importance these experiences have in forming one's image of an entrepreneur.