38
Network number 2, that of linkages dispersing knowledge to wineries within a valley, has
a density of linkages of 0.12. This shows that there is less interaction between firms
signifying that there is limited amount of knowledge flows. Both authors’ findings show
that in general firms with bigger knowledge bases were more involved with clusters and
that as Olavarria et al (2008) suggested there is indeed a link between firm size and
innovation. When carrying out the Pearson’s chi squared test a value of 33,728 was
produced for the relationship of size with innovation. This means that these two variables
are related with 99% reliability. Thus proving a larger firm means more chances of
innovation.
Olavarria et al (2008) and Giuliani (2006) differ in their findings on clusterization in the
Colchagua valley in two terms: the sub-groups formed; and the knowledge bases and
absorption capabilities of the key actors in these cliques. Giuliani finds that the sub-
groups can be separated into a core and periphery demography, with the firms with the
highest knowledge bases in the core with many linkages between each other. The
peripheral firms consist of those with weak knowledge bases who only have a small
density of linkages with few actors of the core and no links with other peripheral firms.
However, Olavarria et al (2008) find that there are in the network of knowledge flows
between wineries in Colchagua a number of small subgroups made up of mainly of three
of four connected nodes. Opposed to Giuliani (2006), the authors of Olavarria et al
(2008) find heterogeneity in the most connected firms in terms of actors most central in
the network. Of the five actors most central to this network in terms of grade of