Mixed method social network analysis: Combining inductive concept development, content analysis, and secondary data for quantitative analysis


Title of the Article : Mixed method social network analysis: Combining inductive concept development, content analysis, and secondary data for quantitative analysis

Author Information : Trenton A Williams (Whitman School of Management, Syracuse University)
Dean A. Shepherd (Kelley School of Business, Indiana University)

Year of Publication : Organizational Research Methods (forthcoming)

Summary of Findings : We develop a new technique for gathering and analyzing data that allows for exploration of networks that are greater in scope, broader in access to diverse social actors, reduced bias, and increased capability for longitudinal designs.

Research Questions : 1. How can researchers draw upon and quantitize increasingly available (and novel) forms of data to explore the nature, influence, and role of social networks in explaining important business-related outcomes?

2. How can researchers balance the strengths and weaknesses of existing approaches to network data collection and analysis by using secondary data in text format?

What we know : Network analysis is useful in explaining how patterns of social relationships influence outcomes such as innovation, idea development, coalition building for new venture funding, trust in organizations, and many other business and entrepreneurship-related outcomes. To assess the influence of networks, researchers often rely on an interview process that can be time-consuming, limited in scope, and static. This process involves asking individuals about their connections and then following up with those connections to identify even more connections. This process is repeated until a network of relationships begins to emerge. While useful, this process is unable to identify large networks, which has led many to use secondary data (i.e., financial data, etc.). Secondary data can provide a broader access to networks but is limited in that it often fails to identify what is exchanged across a network.

Many researchers from a host of fields (sociology, psychology, management, economics, etc.) are interested in how social interactions shape and influence activities that impact our daily lives. This could include networks that foster trust within support groups, network relations that enable mentorship for new employees, networks that provide access to novel ideas for scientists or innovators, broad coalitions within communities that provide an effective voice for change, and so forth. Therefore, researchers (and consumers of research) are interested in data sources that accurately reflect social relationships in society.

Novel Findings : This paper’s primary contribution is the development of a novel research methodology that opens the door for future research contributions. Specifically, this paper outlines how researchers can combine qualitative and quantitative techniques to convert data such as annual reports, news articles, government inquiries, website information, and so forth into a format amenable to network analysis. What this provides is a means whereby researchers can explore the increasingly available text-based data made available by the internet and digital distribution of content. While there is a lot of content, a key challenge has been to find a way to systematically assess and analyze data. We demonstrate how by using key activities that traditionally are only employed in doing case-study research, researchers can develop valid and reliable measures for important network variables.

Novel Methodology : This paper involves introducing a new methodology--that is the focus of the paper. The new methodology for social network analysis involves several components. First, we articulate and outline a method used by a very minor number of scholars involving the development of organizational histories. In this process, researchers can gather information on organizations, companies, team members, etc. until reaching a level of saturation. This data can then be quantitized--that is, converted from text into a numerical value. This in and of itself provides a novel contribution. Second, we discuss how researchers can use techniques typically employed in qualitative analysis (e.g., data structuring) to develop new constructs. This is important for ensuring measures are valid and reliable. Finally, we provide recommendations for research decisions at various stages, giving scholars many different pathways to developing and analyzing network data. We anticipate that this can lead to identification of new network variables and provide a better understandanding of how networks shape things that matter to business and society.

Implications for Practice : This paper could ultimately produce a number of important implications for practitioners. Most professions require networking, having strong relationships, and strengthening ties that could aid with one’s success. Future studies building off the methodology of this study could provide clearer guidelines for how individuals, organizations, and communities might better cultivate, develop and deploy networks for positive outcomes. Furthermore, by opening a methodological pathway to new data sources, researchers might explore novel contexts where networks are urgently needed to address pressing community needs. For example, exploring how networks are used for locals to organize in response to a local disaster, or how companies rapidly respond to a crisis or challenge caused by a surprising event.

Implications on Research: Perhaps the greatest contribution of this paper is the potential it offers for future research. Specifically, we anticipate that it can benefit future research in entrepreneurship and management.

Entrepreneurship: Potential applications include the ability: to obtain data on early-stage transactions to explore new venture resource networks, which could avoid the “survivor bias” prevalent in organizational emergence studies; Identify network relationships between entrepreneurs and investment groups (e.g., crowd funds, angel investors, and venture capital funds), other new ventures, entrepreneurship education centers, etc., to explore a variety of relationships among and between different types of actors.; Possibly avoid “hubris” of entrepreneurs who may offer an exaggerated story of their network and influence. ; Possibly detail network changes over a firm’s growth cycles, including start up, early growth, “professionalization,” and even failure or firm sale.

Management: Provide greater clarity into board networks to explore network power, brokerage roles, and other factors influencing strategic decisions; Explore broader intra-organizational network structures across multiple levels to better understand network factors and their influence on individual or team performance; Identify connections between informal employee groups (within an organization) or community interest groups (outside an organization) that might influence employee actions toward an organization. Explore the role of formal organizational teams, information teams, and other organizational forms and the influence these structures have on a variety of variables (e.g., citizenship behavior, productivity, satisfaction, and performance); Assess the impacts of large-scale organizational network change initiatives among teams, divisions, and other organizational levels. Explore the relationship between these changes and variables of interest, such as organizational learning, employee effectiveness, turnover, and so forth.

Full Citations : Williams, TA & Shepherd, DA. "Mixed Method Social Network Analysis: Combining Inductive Concept Development, Content Analysis, and Secondary Data for Quantitative Analysis." Organizational Research Methods, Forthcoming.

Abstract : This paper outlines a mixed method approach to social network analysis, combining techniques of organizational history development, inductive data structuring, and content analysis to offer a novel approach for network data construction and analysis. This approach provides researchers with a number of benefits over traditional sociometric or other interpersonal methodologies, including the ability to investigate networks of greater scope, broader access to diverse social actors, reduced informant bias, and increased capability for longitudinal designs. After detailing this approach, we apply the method on a sample of 143 new ventures and suggest opportunities for general application in entrepreneurship, strategic management, and organizational behavior research.

Click here to access Full Paper

This study develops a new technique for gathering and analyzing data that allows for exploration of networks that are greater in scope.

Trenton Williams

Trenton Williams

Trent Williams was assistant professor of entrepreneurship at the Whitman School from 2016-17. His research focuses on entrepreneurial venture emergence, resourcefulness, decision-making and resilience. His work has appeared in the Journal of Management, Organizational Research Methods, Journal of Management Studies and The Academy of Management Learning and Education, among others. He is particularly interested in idea generation at early stages of venture creation.
Trenton Williams

Leave A Reply