and to perceive the differences observed between the groups and which the factors that
are associated with the best performance.
We analyzed the differences between the clusters, to characterize each one of the groups
with respect to a set of attributes. The attributes considered were: process of innovation,
the mechanisms of knowledge, the networks and the system of governance.
4- Analysis of the involvement of companies to foment
innovation
For the territories to be more or less competitive and innovative, it is necessary a strong
involvement of all the local agents (public and private), in particular of the companies,
in innovation activities. Relative to the involvement of the companies in innovation
activities we grouping them in clusters analysis in accordance with a scale of
involvement in innovation activities find profiles or characteristics of each group proved
more effective. After we analyzed the different clusters to perceive the factors that are
associated to better performance and to demonstrate that it is possible to measure and to
evaluate the determinant conditions of involvement in innovation activities.
4.1 - Standard behaviours of companies involved in innovation
activities
Applying the k-means clusters analysis (of the SPSS- Statistical Package for Social
Sciences) to the group of variables previously defined for the companies results in 3
groups of companies, each one representing a standard behaviour of innovation with
respect to the attitude of each group. Table 2- summarizes the results of each group
relative to each of the variables previously presented. Note: that 0 corresponds to NO
(cluster is not involved in activities of innovation) and 1 corresponds YES (cluster is
involved in activities of innovation).