THE LESSON LEARNT FROM Childers. 1986. Assessment of the psychometric properties of an opinion leadership scale

In this paper, the author revised the King and Summers' (1970) scale for opinion leadership. As many researchers have found, they argued that the opinion leaders were more likely to be product innovators and high in consumer creativity; were more likely to be risk preference and non-dogmatism; and were full of personal knowledge in certain products.
However, by using the KS scale (7-point response), each Alpha of the opinion leaders' personality was low (about 0.66). By using the same scale, the author could barely get high correlation between the result of KS scale and that which again got from KS scale. Thus, the validity of KS scale was doubted.
To improve this point, they utilized a 5-point scale and modified some KS scales (under the item-to-total correlation, they deleted the inappropriate items), in Table 2. In the end, the Alpha increased and the correlation between opinion leadership with creativity and ownership of the product was confirmed.
Also, to test the validity of the revised scale, different users of cable television (different in their degree of adoption of cable TV services) were analyzed. And, they found that, under Turkey HSD test of means, the average opinion leadership scores for premium subscribers and basic-only subscribers were significantly greater than the score for the group that had refused to subscribe to cable TV. Thus, the people who are familiar to TV services will found high in opinion leadership.

THE LESSON LEARNT FROM Chan & Misra. 1990. Characteristics of the opinion leader: A new dimension

In this paper, the authors analyzed the characteristics of opinion leaders (or givers). Historically, the opinion leaders were considered as important part in new-product adoption and diffusion, and venture business. In their personality, they tend to be less dogmatic, more innovative, more venturesome, likely to be confident in their appraisal of the product category, and more socially active. By analyzing 262 students and using Stepwise discriminant analysis, they found that contrary to previous findings, risk preference, dogmatism, and exposure to print media were not found to be significant discriminating characteristics of opinion leaders. But the product familiarity, personal involvement, and public individuation were significant in distinguishing opinion leaders from nonleaders (p. 57). The relationship between public individuation and opinion leadership is reasonable (p. 54).

THE LESSON LEARNT FROM Elfring & Hulsink, 2003. Networks in entrepreneurship: The case of high-technology firms

In this paper, the authors, using the theory of contingency, argued how strong and weak ties influenced the performance of high-tech start-ups in releasing incremental and radical innovations. They demonstrated that although strong and weak ties are both beneficial to venture businesses in terms of opportunity recognition, resource assembling, and legitimacy; their value is different when facing with different innovation types.
Thus, they proposed that a mix of strong and weal ties are crucial to the high-tech start-ups.
Towards discovery of opportunity, they demonstrated that networks could provide not only source of information bur help entrepreneurs evaluate the opportunities and potential markets. Here, the weak ties could provide more new and broad information whereas strong ties could be of importance of helping evaluating opportunities in greater detail.
Towards securing resources, strong ties provided privileged rights to help gather venture capital and work-forces. Also at the circumstance of legitimacy, the strong ties contributed in convincing potential market and existing industry to believe the innovative products.
In the end, the authors proposed that

1. For pursuing incremental innovations, weak ties are more likely to discover opportunities.
2. For pursuing radical innovations, the mix of strong and weak ties are more likely to discover opportunities.
3. Strong ties enable trusted feedback and exchange of tacit knowledge on the nature of the opportunity.
4. For pursuing incremental and radical innovations, strong ties are more likely to help secure resources.
5. Strong ties enable the exchange of tacit knowledge in the development of resources.
6. For pursuing incremental innovations, strong ties help ventures gain more legitimacy.
7. For pursuing radical innovations, the mix of strong and weak ties are more likely to gain legitimacy (cognitive ones and socio-political ones).
8. Weak ties are more important for radical innovations, as endorsement by outsiders is important in gaining legitimacy.

THE LESSON LEARNT FROM Freeman, L. 1978. Centrality in social networks: Conceptual clarification

First introduced by Bavelas in 1948, the idea of centrality in human community has been recognized. The centrality was related to group efficiency in problem solving, perception of leadership, and the personal satisfaction of participants. And, before arguing centrality, the graph-theory is necessary, where the words: "adjacent, degree of point, path, cycle, connected, distance, geodesics" have their unique meaning. 

In the terms of point centrality, "a point that falls on the communication paths between other points exhibits a potential for control of their communication." And, a central position is one that is not dependent upon others as intermediaries or relayers of messages. Short distances mean fewer message transmissions, shorter times, and lower costs. Especially, concern with communication activity will suggest a degree-based measure, interest in control of communication requires a measure-based upon betweenness; and concern with either independence or efficiency leads to the choice of a measure based upon closeness. 

The LESSON LEARNT FROM Borgatti, S. 1995. Centrality and AIDS

"Centrality measures are commonly described as indices of prestige, prominence, importance, and power - the four Ps." There are usually four measures, which were developed by Freeman (1979) and Bonacich (1972), being used in network analysis: degree, closeness, betweenness, and eigenvector centrality. "Degree centrality may be defined as the number of ties that a given node has." When all else is equal, we could describe degree centrality as measuring the risk (or opportunity) of receiving whatever is flowing through the network. Furthermore, in eigenvector centrality, it was demonstrated that "it wasn't just how many people a person knew that counted, but how many people the people that they knew knew." That's to say, an actor that is connected to many actors who are themselves well-connected is assigned a high score, but an actor who is connected only to near isolates is not. "Closeness centrality may be defined as the total graph-theoretic distance of a given node from all other nodes." "Betweenness centrality is defined as the number of times that a node needs a given node to reach another node." Thus, also according to Granovetter's (1973) definition, actors with many weak ties are more important than others because removing those actors would do the most damage to transmission possibilities throughout the network.