RESULTS

Using moderated multiple regression results, the central hypothesis of a three way interaction received significant support in five of the six equations including all three of the equations involving two year sales mean as the performance growth measure. These results are reported in Tables 1 and 2. In all 12 equations reported across these two tables the values of all three conditional effect variables have been centered by subtracting the mean for the variable from the actual value of the variable for every observation. Therefore these variables have a mean of zero (0) and their standard deviations remain unchanged.

The results of the test of Hypothesis 1 can be found in Table 1. The test of the increase in R2 in moving from the simple regression model of Model 1A to the moderated regression model of Model 1B was significant. Therefore, the moderated model was considered to constitute a sufficiently strong contribution beyond the simple main effects model that the hypothesis could not be rejected before examining Model 1B. The partial test for the significance of the 3-way interaction between sales mean, the positiveness of NVT reaction to VC strategic assistance, and the level of global justice reaction of the NVT to interactions with the VC confirmed the expected positive and significant coefficient (b=43,994, tprob=.0065). Hypothesis 1 could not be rejected based on this analysis. This relationship is graphed in Figure 1 (Graphic unavailable). Under low global justice perception conditions, low strategic assistance is perceived by VCs as less risky over the entire relevant range of the sales mean variable. Under high global justice perception conditions, high strategic assistance is perceived as less risky across the entire range of above average sales performance conditions. Thus, a high global justice perception allows the VCs to freely use high levels of strategic assistance when performance effects appear likely to coincide with such assistance.

The results of the test of Hypothesis 2 can be found in Table 2. The test of the increase in R2 in moving from the simple regression model of Model 2A to the moderated regression model of Model 2B was significant. Therefore, the moderated model was considered to constitute a sufficiently strong contribution beyond the simple main effects model that the hypothesis could not be rejected before examining Model 2B. The partial test for the significance of the 3-way interaction between income performance level, the positiveness of NVT reaction to VC strategic assistance, and the level of global justice reaction of the NVT to interactions with the VC confirmed the expected positive coefficient (b=113,482, tprob=.0552) at a marginal significance level. Hypothesis 2 could not be rejected based on this analysis. This relationship is graphed in Figure 1. Under low global justice perception conditions, VCs see low strategic assistance assistance as less risky than high strategic assistance throughout the relevant range on mean income. Under high global justice perception conditions, VCs see high levels of strategic assistance as less risky under all above average income performance conditions. As was shown for sales performance but even more pronounced for income performance, a high global justice perception condition allows VCs to feel much less risk at becoming involved via strategic assistance whenever above average performance enhancements appear likely to coincide with such assistance.

The results of the test of Hypothesis 3 can be found in Table 1. The test of the increase in R2 in moving from the simple regression model of Model 3A to the moderated regression model of Model 3B was significant. Therefore, the moderated model was considered to constitute a sufficiently strong contribution beyond the simple main effects model that the hypothesis could not be rejected before examining Model 3B. The partial test for the significance of the 3-way interaction between sales performance mean, the positiveness of NVT reaction to VC operational assistance, and the level of global justice reaction of the NVT to interactions with the VC confirmed the expected positive and significant coefficient (b=84,053, tprob=.0010). Hypothesis 3 could not be rejected based on this analysis. This relationship is graphed in Figure 1 (Graphic unavailable). Under low global justice perception conditions, VCs tend to view higher levels of operational assistance as more risky at above average performance levels but less risky at below average performance levels. This suggests that VCs may require more extensive use of operational assistance in conditions where venture sales performance is below average and where NVT global justice perceptions are low in order to lower their own risk perceptions and to justify continuing to fund the venture. Under high global justice perceptions, a higher level of operational assistance lowers VCs’ risk perceptions across all above average sales performance levels.

The results of the test of Hypothesis 4 can be found in Table 2. The test of the increase in R2 in moving from the simple regression model of Model 4A to the moderated regression model of Model 4B was significant. Therefore, the moderated model was considered to constitute a sufficiently strong contribution beyond the simple main effects model that the hypothesis could not be rejected before examining Model 4B. The partial test for the significance of the 3-way interaction between income performance level, the positiveness of NVT reaction to VC operational assistance, and the level of global justice reaction of the NVT to interactions with the VC confirmed the expected positive coefficient (b=119,424, tprob=.0559) at a marginal significance level. Hypothesis 4 could not be rejected based on this analysis. This relationship is graphed in Figure 1. Under low global justice perception conditions, VCs see low operational assistance as less risky than high operational assistance throughout the relevant range on income mean. Under high global justice perception conditions, VCs see high levels of strategic assistance as less risky under all above average income performance conditions. As was shown for sales performance but even more pronounced for income performance, a high global justice perception condition allows VCs to feel much less risk at becoming involved via operational assistance whenever above average performance enhancements appear possible

The results of the test of Hypothesis 5 can be found in Table 1. The test of the increase in R2 in moving from the simple regression model of Model 5A to the moderated regression model of Model 5B was significant. Therefore, the moderated model was considered to constitute a sufficiently strong contribution beyond the simple main effects model that the hypothesis could not be rejected before examining Model 5B. The partial test for the significance of the 3-way interaction between sales performance mean, the positiveness of NVT reaction to VC conflict recognition assistance, and the level of mutuality-based justice reaction of the NVT to interactions with the VC confirmed the expected positive and significant coefficient (b=37,456, tprob=.00262). This relationship is graphed in Figure 1 (graphic unavailable). Under low mutuality-based justice perception conditions, VCs tend to have no pronounced risk preference between conflict recognition assistance levels. Under high mutuality-based justice perception conditions, a higher level of conflict recognition assistance lowers VCs’ risk perceptions across all above average sales performance levels.

The results of the test of Hypothesis 6 can be found in Table 2. The test of the increase in R2 in moving from the simple regression model of Model 6A to the moderated regression model of Model 6B was significant. Therefore, the moderated model was considered to constitute a sufficiently strong contribution beyond the simple main effects model that there was justification in examining Model 6B. Therefore, the moderated model was considered to constitute a sufficiently strong contribution beyond the simple main effects model that the hypothesis could not be rejected before examining Model 5B. The partial test for the significance of the 3-way interaction between income performance mean, the positiveness of NVT reaction to VC conflict recognition assistance, and the level of mutuality-based justice reaction of the NVT to interactions with the VC failed to confirmed the expected positive and significant coefficient (b=77,829, tprob=.2160). Hypothesis 6 was rejected based on this analysis. Although the relationship was too weak to be statistically significant, the parallel nature of the relationship can be viewed as graphed in Figure 1. Under low mutuality-based justice perception conditions, VCs appear to perceive low conflict recognition assistance as slightly less risky than high conflict recognition assistance levels, especially at above average performance levels. However, unlike the other five models, under high justice perception conditions, a higher level of assistance does not appreciably lower VCs’ risk perceptions except at extreme above average performance levels.

 

Top of Page
Previous Page Main Page Next Page
Return to 1996 Topical Index


1997 Babson College All Rights Reserved
Last Updated 4/5/97 by Cheryl Ann Lopez

To sign-up for the Center for Entrepreneurial Studies' publication lists,
please register with the
Entrepreneurship WebTeam.