Justin Blaney Justin Blaney

The Strength of Strong Ties, A Critique

Justin Blaney

April, 17 2013

Click here to download as a PDF



The purpose of this paper is to provide a critique of Krackhardt’s scholarly work, The Strength of Strong Ties: The importance of philos in organizations (1992). This paper will analyze his methods, results, and discussion, and end with a conclusion on the quality of his work. Finally, a recommendation will be provided on the suitability of citing this work in other scholarly research literature.


Krackhardt, networking, strong ties, weak ties, granovetter, network connections, philos networks, trust networks, advice networks, friendship networks


Blaney, J. (2013). The strength of strong ties, a critique. Retrieved from www.justinblaney.com




Granovetter (1973) argued that weak ties are more beneficial than strong ties in some circumstances, but since this article was first published little consensus has formed around the definitional differences between a strong and weak tie. While Granovetter contrasted between ties in terms of measurable factors such as length and frequency of contact, scholars later discussed whether or not other words such as friend would be appropriate replacements for Granovetter’s ‘strong tie’. However, Fisher (1982) argued that friendship means many things to many people and isn’t a suitable scholarly term to utilize in the discussion of network relationships. By 1982, philos emerged as a term Fisher felt could provide a more accurate depiction of the type of relationship that Granovetter described (1982), yet the mere substitution of philos for strong tie does not completely explain the component parts of either term.

Krackhardt (1992) argued that too little attention has been paid by scholars to the individual characteristics of the ties themselves. He attempted to add to this discussion by analyzing how friendship or philos affects the relationships between a particular company’s employees. In particular, Krackhardt attempted to distinguish between two components of a strong tie by looking at two types of networks an individual has the potential to possess: the philos, or friendship network and the advice, or trust network. If successful, Krackhardt aimed to identify which of these two networks play a more powerful role in organizational change so that change agents can efficiently utilize the knowledge of these networks within the organization to create change in one direction or another (1992, p. 220).

To provide a critique of Krackhardt’s work (1992), this paper will analyze his methods, results, and discussion, and end with a conclusion on the quality of his work. Finally, a recommendation will be provided on the suitability of citing this work in other scholarly research literature.


Krackhardt began to describe his methods in his introductory section by contradicting his own sources with little support for doing so. After effectively arguing that philos is a superior term to friendship for the purposes of discussing network connections, he opted to combine the two terms into one category by saying it was logical to do so (1990, p. 220). He described the two types of networks he included in his analysis by clumping them into groups: philos/friendship and advice/trust (1990, p. 221), yet he cites sources who describe philos as being based more on trust. The support he provides for these distinctions works against his use of these groups, causing the results of his analysis to come needlessly into question. Krackhardt would have done better to clump the components of his groups differently or provide more robust support for the groups that he chose to use.

Though Krackhardt erred in his defense of the groups used to analyze relationships within the test company, he employed excellent technique in the methods used to investigate the network connections of the firm’s employees. Network data was collected from all 36 employees of the company four months before a unionization attempt was made by the employees (it is not clear whether this time gap was by design or whether it occurred as a result of serendipity). Krackhardt created a questionnaire to discover the nature of the networks by asking employees to identify both their own relationships and guess at the relationships of every other employee (1990, p. 222). This helped him understand three networking components of each individual: their advice network, their philos networks, and their ability to correctly recall the networks of others which Krackhardt used to measure the individuals centrality in the firm’s network (1992, p. 223).

After the unionization attempt failed, the author requested permission from senior management to interview three employees who were key to the union process. Later, Krackhardt returned to interview three additional employees to verify the reports of the first three (1990, p. 225). Unfortunately, he provided little support for the selection of these six informants. Furthermore, he only interviewed employees who the three founders of the company allowed him to interview. It stands to reason that the founders, who were all against the unionization of their company, may have purposefully or inadvertently directed Krackhardt to interview only employees who were sympathetic with their own points of view. This could cause the analysis of the case study to become colored by an anti-union sentiment which may or may have affected his conclusions. Another issue with the selection of informants is that only six where used. This number represents 17% of the company and, in light of the non-random process used to select these individuals, this sample has the potential to introduce individual bias that could have been diminished had a larger number of employees been interviewed. On the other hand, all six employees were granted anonymity, which likely allowed them greater latitude in the frankness of their responses.


Krackhardt found several major conclusions from the data and subsequent interviews that supported his theories that philos networks are more effective than advice networks in organizational change. The first is that union supporters had less friendship connections than antiunion employees (1992, p. 233). Second, Krackhardt found that employees who had high scores in the measurement for friendship network strength and low scores in advice network strength had more influence on the outcome of the unionization campaign than employees with an opposite mixture of network strength (1992, p. 234). Yet, the questionnaire used to accumulate this information was somewhat cumbersome to the respondent; because of this, it is possible the data was not as accurate as the author portrays. For example, each employee answered three sets of questions for each of the 35 other employees in the firm and each question had 36 possible answers (1992, p. 222). This results in thousands of possible outcomes and could be daunting and overly time consuming for many employees to complete in accurate detail.

It is an interesting side-note to highlight that employees can be high in one type of strong tie and low in another. This finding lends credibility to the purpose of Krackhardt’s work in that it proves how investigating tie strength in a general sense does not provide a complete picture of how network connections affect organizational change. Though this finding is supportive of the author’s arguments, it also highlights a potential problem with the results as presented. Krackhardt provides no explanation for which components of his findings were presented in this article (1992, p. 220). By keeping this aspect of his methods veiled, it is impossible to know whether there were contradictory findings that were simply ignored for the benefit of proving the hypothesis of the author. Perhaps Krackhardt felt that it was unnecessary to reveal every aspect of his methods because of his self-perceived status as an expert on the subject. If this is the case, based on the author’s resume and the number of times his articles have been cited by other scholars, he may be right. Yet, he would have removed this concern altogether by simply being more transparent.


Perhaps the most beneficial finding of this research is that philos networks are more beneficial in some circumstances, while advice networks are more beneficial in others (1992, p. 237). By failing to understand this, Krackhardt argues that the union failed, in part, because they were unable to leverage the correct network within the company to affect the change they desired. For example, the union ignored a highly connected individual who was a close friend of a top antiunion manager because they assumed the individual would side with his friend (1992, p. 236). Yet, since this individual was highly connected within the company’s philos network, he had a great deal of influence of many of the companies employees. Had the union attempted to persuade this employee to their side, they may have won many other employees as well. Given the results of this study, Krackhardt’s conclusion seems well validated.

The author does an excellent job of tying his research back to his original argument that not all strong ties are created equal. While Granovetter argued that weak ties were in fact superior to strong ties, Krackhardt has effectively shown that Granovetter may have only been correct in certain circumstances, especially those that relied heavily on the often less effective advice network strong ties. By brining the discussion back to the decades long debate about the nature and definition of strong and weak ties, Krackhardt helps the reader to see how these findings are generalizable in many circumstances beyond this single case study.


There are several components of Krackhardt’s work that diminish the potential applicability of the findings, yet the thoroughness of the author’s methods and his highly generalizable and useful conclusions indicate that scholars should be confident in citing this work as the foundation for future research.


Burns, R. B., & Burns, R. A. (2008). Business Research Methods & Statistics Using SPSS. London: Sage Publications Ltd.

Fisher, C. (1982). What do we mean by ‘friend’? An inductive study. Social Networks, (3), 287-306.

Girden, E. R, & Kabacoff, R. (2011). Evaluating research articles: From start to finish. 3rd ed. Los Angeles: SAGE.

Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology, 78(6), 1360–1380.

Krackhardt, D. (1992). The strength of strong ties: The importance of philos in organizations. In N. Nohria & R. Eccles (Eds.), Networks and Organizations: Structure, Form and Action (pp. 216–239). Boston, MA: Harvard Business School Press.

Krackhardt, D. (1990) Friendship patters and culture: The control of organizational diversity. American Anthropologist, 92(1), 142-154.

About the author

Justin Blaney is the author of #1 National Bestselling Novel, Evan Burl and the Falling. He currently is a doctoral candidate at the University of Maryland. Justin Blaney obtained a B.A. in Business Administration in 2004 at George Fox University and received an M.B.A. with a minor in Marketing in 2011 at Texas A & M University. He speaks and consults regularly on research topics such as network connections, behavior modification, and communication. Links to other works of scholarship authored by Justin Blaney can be found at http://www.justinblaney.com/category/publications/. He can be reached at justin@justinblaney.com.

Share Post :

More Posts