An Empirical Study on New Product Development Process by Nonparametric Combination (npc) Testing Methodology and Post-stratification

In many scientific disciplines and industrial fields researchers and practitioners are often faced with complex problems when dealing with comparisons between two or more groups using classical parametric methods although real problems rarely agree with the stringent assumptions required by such methods. The NPC methodology (Pesarin, 2001) offers an innovative but well tested approach that frees the researcher from stringent assumptions of parametric methods and allows a more flexible analysis both in terms of specification of multivariate hypotheses and in terms of the nature of the variables involved in the analysis. It does so by reducing the problem to a set of simpler sub-problems, each provided with its own suitable solution. Moreover, one of the most relevant features of NPC Test is that it does not need a modelling for dependence among variables. The case study was based on a sample of 85 firms, working in a B2B market in two specific manufacturing industries. Applying the NonParametric Combination (NPC) of dependent rankings methodology (Pesarin, 2000), the sample was divided into two groups on the basis of external performances, i.e. market, product and financial success criteria, for New Products launched over a three year period. The Best firms were those above the median score on a global success ranking, taking into account all three of the above mentioned success criteria. The other companies were called Rest firms. In the study we tested the hypothesis that Best firms have better internal performances, i.e. Time performances (Launch on Time and Time To Market Reduction) and better Quality Capability of New Product performance, than the Rest. We also wish to test if Best firms have greater internal environment support (e.g. NP Strategic Guide, Capabilities, etc.) and use some practices and drivers more intensively than Rest firms do. Results confirmed that Best firms do have greater internal environment support. Due to the multivariate nature of the problem, the testing procedure was properly broken down into a set of simpler sub-problems, each provided with its permutations solution; then, by nonparametric combination of them, we obtained an overall test concerned with the global null hypothesis. Moreover a confounding factor expresses by innovation level was tested as possible discriminating variable by some intermediate partial tests related to sample strata.