Information Technology Reference
In-Depth Information
namely, each position variables' goodness of fit for innovation performance are
different.
In the past, many scholars proved that network position had significant effect on
firm's performance or innovation performance (Zaheer & Bell, 2005; Ahuja, 2000;
Tsai, 2001). However, they often apply different network position variable to
explain innovation performance in different industries. Hence, one objective of this
research is to identify which network position variable is most suitable for
explaining the network of LCD industry which emphasizes collaboration of supply
chain in explaining innovation performance.
We use each adjusted R square
(degree of explanation toward innovation
performance) of regression analysis to present the distinction among different
position variables and networks in Figure 6. Consequently, it shows that the
explanation of coreness is the highest in each network. In other words, coreness
offers the best explanation for innovation performance in all four networks.
According to Borgatti and Everett (1999), in contrast to a more loosely connected
class of actors, a core formed by a group of densely connected actors makes up the
periphery of the network (refer to Figure 7). From the analysis results, it is likely
that there are two key firms which dominate over other LCD firms. Important
information and knowledge are mostly held and distributed by them, thus they have
better innovation performance than others. In other words, the important
information and knowledge often held by firms who are core in LCD industry.
Hence, it can be concluded that when key firms in a network hold critical
information, knowledge, or know-how, it is suitable to use coreness to explain
innovation performance.
Fig. 6
Differences among position variables in explaining innovation performance
On the contrary, structural holes offer least explanation in each network. In
addition, all network position variables get the best explanation toward innovation
performance in LCD-InsNet. The explanatory power of position variables gradually
