Table 1 shows the results of the logit regression analysis. Four models were used to develop the full analysis. Model 1 regresses the effect of industry on survival. Model 2 adds the eight types of managerial problems to the regression. Model 3 adds sales history to the analysis. And model 4 introduces the interaction between managerial problems and sales performance. Thus model 4 provides us with a full testing of the hypothesis exploring the impact of sales performance and managerial problems on survival of the new firm.
The results of Model 2 suggest that four of the managerial problem sets has a direct effect on the survival of new ventures after controlling for industry effects. New ventures that have solved managerial problems in developing good relations with unions and have solved problems in developing goals and preparing strategic plans have significantly greater odds of surviving (? = .866; p < .05 and ? = .533; p < .05 respectively). However, those firms reporting they were able to solve problems coping with government regulations had lower odds of survival (? = -.371; p < .05). Those who had start-up problems developing good relations with unions (? = -1.199; p < .05), selecting a lawyer or accountant (? = -.603; p < .05) when the business was established, and had not resolved that problem have significantly lower odds of surviving. Never having experienced managerial problems regarding the selection of an accountant or lawyer significantly increased the odds of survival (? = .606; p < .01). However, the results also suggest that firms reporting no problems in motivating and coordinating personnel decreased the odds of survival (? = -.479; p < .05). Finally, the results in Table 1 indicate that for firms who had start-up problems and were unable to solve the problems had significantly increased their chances of survival when the problem was coping with government relations (? = .424; p < .05) and finding qualified people (? = .541; p < .05). Thus, there is partial support for hypothesis 1.
Model 3 reports the results of adding sales performance history to the equation. The findings indicate that sales history significantly improves the fit of the model (?2 = 70.789;; p < .001). New business that get off to a slow start with low sales volume during the first year and continued low sales have significantly lower odds of surviving ( B = -.672; p < .05).
Model 4 reports the results for testing whether the relationship between problem solving effectiveness and new firm survival depends upon the pattern of sales history. Adding the interaction terms significantly improves the fit of the model (?2 = 166.974; p < .001) indicating an interaction effect. Examination of the interaction terms for sales performance by managerial problem found three sets to be significant. Sales performance history significantly moderates the relationship between problem solving relative to motivation and coordination of personnel and new firm survival (Wald = 16.026; p < .05). Businesses that had high sales in the first year followed by a low trend and who had solved problems motivating and coordinating personnel have significantly higher odds of surviving (? = 2.139; p < .05).
The second set of interaction terms for sales performance and problem solving found to be significant focused on developing goals and preparing strategic plans (Wald = 18.935; p < .01). Three interaction terms were of importance. Businesses with high sales in the first year and low sales in later years had lower odds of surviving (? = -2.706; p < .01). This suggests that initial sales volume does not buffer new ventures from continued problems in the development of goals and strategic plans. A second finding in this set identifies firms with low initial sales volume followed by high growth who had no problems with goals and strategic plans to increase their probability for survival (? = 1.998; p < .05). This was also true for firms with similar performance but were unable to solve their problem in developing goals and strategic plans (? = 1.399; p < .05)
The third set of interaction terms for sales performance and problem solving found to be significant concerned finding qualified executives (Wald = 14.591; p < .05). Firms with a high sales performance in the first year with continued high sales performance in later years while experiencing no problems in finding qualified executives were likely to fail (? =-4.178; p < .05). However, for firms with a high sales performance, but a lower rate of sales growth in subsequent years who reported no problems increased their odds of survival (? = 2.854; p < .01). However, probability of failure was evident for firms with a high sales performance at the start, but low sales performance afterwards who did solve their problems finding qualified executives (? = -3.189; p < .01).
A summary of the interaction effects is
displayed in Table 2.
© 1997 Babson College All Rights Reserved
Last Updated 03/30/98