HYBRID CONJOINT-ANALYSIS - AN ESTIMATION PROBE IN NEW VENTURE DECISIONS

This article describes a portion of our ongoing research on the modelling of the venture capitalists' evaluation of new venture proposals. In particular, it presents the results of a pilot study compares the predictive validity of three models; self-explicated, traditional conjoint, and hybrid conjoint. These models were applied to venture capital decision making- The results indicate that these models can be used as a practical research method in venture capital decision-making. Secondly, the traditional conjoint and hybrid conjoint models out-perform the self-explicated model in terms of explanatory power but only marginally in terms of predictive power. Fourth, overall the predictive validity of the models examined is somewhat stranger for predictions based on the same data on which the models were tested - as expected - than for data not used by the models. However, the drop is not sharp. Fifth, the venture capitalists appeared to exhibit screening procedures that were conjunctive (non-compensatory), whereas in subsequent detailed analysis the procedure approximates a more compensatory one. Finally, it has been demonstrated that characteristics of the entrepreneur, the product and the market can replicate the venture capitalists' decision-making.