|Income Adj. R2 Anova F sig||.00897 .2005||Empltrnd Adj. R2 Anova F sig||.01748 .1158||Satisf Adj. R2 Anova F sig||.05636 .0006||$Satisf Adj. R2 Anova F sig||.03334 .0125|
|Section 1||Coeff||Sig T||Coeff||Sig T||Coeff||Sig T||Coeff||Sig T|
|Income Adj. R2 Anova F sig||.00897 .2005||Empltrnd Adj. R2 Anova F sig||.00897 .2005||Satisf Adj. R2 Anova F sig||.00897 .2005||$Satisf Adj. R2 Anova F sig||.07002 .0001|
|Section 2||Coeff||Sig T||Coeff||Sig T||Coeff||Sig T||Coeff||Sig T|
More rigorous testing to assess the difference between pull-motivated and push-motivated entrepreneurs was conducted by MANOVA analysis, within SPSS. Correlations between the identified performance measures were first examined, confirming that each was assessing a unique aspect of performance and did not exhibit excessive correlations to the other measures (Table 5). While satisfaction with the standard of living provided is moderately correlated with both general satisfaction (.53) and income (.52), these are not deemed to be excessive.
Expanding success into four dimensions by adding measures of material and general satisfaction does, on the other hand, reveal statistically significant group differences (.007 at .93 power level). Once again, the underlying assumptions of independence of assumptions and homogeneity of variance were assessed; Boxs M test is not significant, and while the general satisfaction variable marginally violates the assumption of independence, the other indicators do not. This singular violation was not deemed as invalidating subsequent interpretation. Using the expanded multidimensional measure of venture success, pull-motivated entrepreneurs report higher average income, demonstrate a slightly higher tendency for employment growth, and are more satisfied, both in general and by the standard of living provided by their venture. While all four indicators demonstrate significance, it is notable that traditional performance measures of income and employment trend fail to demonstrate sufficient power to preclude a type II statistical error. The satisfaction indicators, on the other hand, exhibit acceptable discriminatory power (.9771 for general satisfaction, .8054 for standard of living satisfaction). (See table 6).
© 1997 Babson College All Rights Reserved
Last Updated 06/01/98