Geography Reference
In-Depth Information
Table 13 . 2
Analysis of the statistical correlation between relational reputation and
centrality of a single actor inside the network
Dataset
Correlation
between 'RR'
and 'centrality'
p- value
(a = 0.10)
Network
density
(in terms of RR)
Network density
(in terms of
centrality)
Entire dataset
0.4116
<0.0001
/
/
Bormio
0.4229
0.0005
0.095
0.193
Cortina
0.5017
0.0047
0.062
0.143
Pila
0.3905
0.0971
0.163
0.322
6000000
R
R
4000000
2000000
0
6000000
Centrality
Figure 13 . 5
Correlation between relational reputation and centrality (QQ plot)
The sub-results with reference to the three contexts point to the same conclusion: rela-
tional reputation is closely correlated with the central position in the network. Table 13.2
shows also the degree of network density, calculated by using the adjacency matrix of
knowledge sharing. When comparing the correlation values with these densities, a low
network density seems to increase the positive association between the considered vari-
ables. Probably, a non-homogeneous distribution of the knowledge sharing l ows - that
is, the existence of many relevant sub-networks - amplii es the concentration of reputation
and, therefore, the i rm variety in terms of social assets. Maybe, when network density is
high, and, hence, there is frequent and intense knowledge sharing, the role of relational rep-
utation stays in the background. On the contrary, in a network characterized by a weaker
exchange of knowledge - low density - relational reputation plays an important role and
stands out in terms of polarization of knowledge l ows and develops signii cant lock-ins.
The second research hypothesis analyses the same problem in terms of competence-
based reputation and the i ndings support it as well. Structural indicators and the QQ
plot, shown in Table 13.3 and Figure 13.6, lead to similar conclusions, but with some
important dif erences.
First of all, the absolute levels of correlation between competence-based reputation
Search WWH ::




Custom Search