Introduction to qualitative

social network analysis

By Simon Moss

Introduction

Overview of quantitative social network analysis

Many quantitative researchers conduct social network analysis—a set of tools and techniques to
investigate how information, resources, friendships, associations, and other attributes extend across
people, organizations, and other bodies. Typically, these researchers will generate a graph, similar to the
following example. In this graph

 each circle, usually called a node, will often represent a person or organization
 each line or arrow, often called an edge, represents the association between these nodes; for
example, a thick line could indicate the relationship is close

Page 1 of 8
After they construct this graph, researchers often calculate a set of metrics. For example, researchers
might

 calculate the betweenness centrality of every node or person; this metric indicates the extent to
which this person bridges or connects the various segments of the network.
 Determine which attributes correlate with this metric; for instance, the researcher might discover
that people who connect various segments of the network tend to be under 30.

Limitations of quantitative social network analysis

Quantitative social network analysis, although informative, does not answer vital questions about how
and why these networks develop. For example, quantitative social network analysis does not ascertain

 why the relationships between the individuals, organizations, or nodes are established
 how these relationships evolve over time
 the barriers that impede the formation or maintenance of these relationships
 the effects or consequences of these relationships

Overview of qualitative social network analysis

Qualitative social network analysis is designed to address these limitations. That is, qualitative social
network analysis explores why and how the relationships are established and maintained as well as
barriers or complications. Typically, the researcher will conduct interviews, administer surveys, observe
individuals, or analyse documents to resolve these questions. They might, for example

 first conduct a quantitative social network analysis or ask participants to describe their networks—
such as the colleagues to whom they converse most frequently
 then conduct interviews in which they ask various questions about these relationships, such as the
challenges they experience, depending on the research question

As this depiction implies, Heath et al (2009) and other researchers distinguish two main variants, as
defined in the following table. Perhaps the most common approach that researchers choose is to
complete a quantitative or conventional social network analysis first and then to apply the qualitative
social network analysis later.

Variant Qualitative vs quantitative Details

Qualitative social Qualitative only  The researcher asks questions to
network research characterise the social network of
participants
 The researcher might then ask
questions to explore why and how
the connections or relationships

Page 2 of 8
 were established and maintained
Qualitative social A quantitative phase usually  The researcher completes a
network analysis precedes a qualitative phase— quantitative social network analysis
although the two phases  The researcher might then ask
sometimes coincide questions to explore why and how
the connections or relationships
were established and maintained

Aim of this document

This document offers some insights into how researchers should conduct a qualitative social network
analysis. This document assumes you have skimmed, if not read, the document on quantitative social
network analysis. However, only a basic understanding of social network analysis is assumed.

When should you consider a qualitative social network analysis?

The following table outlines the circumstances in which a qualitative social network analysis may be
suitable.

When should I consider a qualitative social network

You have already conducted a quantitative social network analysis and would now like to
understand the causes, features, and consequences of these relationships in more depth
You would like to explore how and why information, resources, friendships, associations, and
other attributes spread across people, organizations, or other bodies.

In general, researchers who conduct qualitative social network do not embrace positivism, in which they
strive to uncover the true reality. Instead, these researchers tend to adopt

 interpretivism, in which they explore how individuals construct or understand their world rather than
attempt to unearth one true reality
 pragmaticism, in which they are willing to adopt any paradigm that facilitates their attempt to
address some problem.

Page 3 of 8
How to conduct qualitative social network analysis

Illustration

This section offers some insights into how to conduct a qualitative social network analysis. To introduce
these principles, consider an example in which a researcher wants to explore how research candidates
acquire knowledge and develop skills during their candidature.

Develop the networks

To conduct a qualitative social network analysis, researchers first need to construct a social network.
Most researchers will apply one of two approaches. First, they might conduct a quantitative social
network analysis. They might, for example

 ask PhD candidates at a university to indicate from which peers they acquired knowledge and
information about their course and research
 utilise the responses to generate a social network or sociogram, using the principles that appear in
this document to generate a graph that resembles the following display
 in this display, each circle represents one candidate, and each colour represents a distinct field

Alternatively, researchers can derive rough networks from interviews. To illustrate, some researchers
utilize an approach called egocentric sociograms or sometimes egonets (e.g., Dobbie et al., 2018).

Page 4 of 8
Egocentric sociograms implies a social network that revolves around each participant separately. To
generate these graphs, the researcher could

 present five concentric circles

 instruct participants to write their name in the centre or smallest circle
 invite participants to write names of other relevant individuals—such as colleagues in their PhD
program—on the circles; the distance from the centre should be inversely related to the importance
of each individual, as the following figure, derived from
https://acadeafic.org/2020/02/10/autophotography-and-egocentric-sociograms/, demonstrates
 reveal this approach is not precise—and participants can never feel certain about where to attach
each person

Several features will vary across studies. For example, depending on the research question, researchers
might ask participants to indicate each person who

 they regard as important

 they perceive as helpful
 they contacted in the last week
 they met in the last month, and so forth

Each colour might represent a distinct category, such as the field of research. These networks can both
guide subsequent conversation and, in some instances, may even constitute relevant data or displays.

Conduct interviews about the network

Page 5 of 8
Researchers will then, after constructing these networks, tend to conduct interviews with participants—
such as individuals or representatives of organizations. Typically, they might conduct between 40 and
60 interviews, but fewer interviews might be sufficient in particular circumstances. These interviews are
designed to clarify

 the changes or circumstances that promoted or hindered these relationships

 the features of these relationships
 the consequences of these relationships.

Usually, researchers will conduct semi-structured interviews. Typical questions might be

 when did you develop the relationships with peers from other fields?
 what events facilitated or hindered the development of these relationships
 how would you characterise these relationships?
 how have these relationships changed over time?
 what did you gain from these relationships?

Rather than interviews, other methods may be suitable. Occasionally, researchers might invite
participants to complete surveys instead, for example.

Analysis of the data

To analyse the data, most researchers apply techniques they would utilize to explore other interviews,
such as thematic analysis or content analysis. They would also consider the principles of exemplary
qualitative research, outlined in this document, such as audit trails, memos, debriefing, member
checking, and so forth. To illustrate, they might

 identify key themes, such as strategies people applied to develop relationships

 relate these themes to various metrics—such as discover that people who apply a particular strategy
tend to be the most central or important to social networks.
 consider which individuals, organizations, or nodes seem to be missing—such as reference to an
associate supervisor but not a principal supervisor

Examples

The principles of qualitative social network analysis are straightforward. However, to be adept in this
method, you should familiarize yourself with past examples and the insights these examples generated.

Page 6 of 8
Familiarity with these insights could help you clarify the research question, design excellent questions,
and interpret the data. The following table presents a few examples.

Example Insights from this study

This study explored how various organizations  The organizations that were most central
collaborated to promote a social movement to this movement tended to have provided
around food (Luxton & Sbicca, 2020) services in this region over many decades
 The organizations that were most central
to this movement tended to operate from
many sites
 The organizations that were most central
to this movement became important after
they collaborated to attract funding
This study examined the social network of  Gamblers often conceptualised their ex-
gamblers, exploring which individuals helped partners as central to their social network,
these gamblers overcome their addiction because they affect the lives of their
(Dobbie et al., 2018) children
This study examined how technology diffuses  Individuals often formed self-managed
across teams (Bertolotti & Tagliaventi, 2007) teams, and these informal teams facilitated
the transfer of knowledge

References

Bertolotti, F., & Tagliaventi, M. R. (2007). Discovering complex interdependencies in organizational

settings. Qualitative Research in Organizations and Management, 2(1), 43–61

Dobbie, F, Reith, G, McConville, S (2018) Utilising social network research in the qualitative
exploration of gamblers’ social relationships. Qualitative Research 18(2): 207–223.

Fischer, M. (2011). Social network analysis and qualitative comparative analysis: their mutual benefit
for the explanation of policy network structures. Methodological Innovations Online, 6(2), 27–51.

Heath, S., Fuller, A., & Johnston, B. (2009). Chasing shadows: Defining network boundaries in
qualitative social network analysis. Qualitative Research, 9(5), 645–661.

Hogan, B., Carrasco, J. A., & Wellman, B. (2007). Visualizing personal networks: Working with
participant-aided sociograms. Field Methods, 19(2), 116-144.

Krinsky, J., & Crossley, N. (2014). Social movements and social networks: Introduction. Social
Movement Studies, 13(1), 1–21.

Page 7 of 8
Luxton, I., & Sbicca, J. (2020). Mapping movements: a call for qualitative social network analysis.
Qualitative Research, 1468794120927678.

Page 8 of 8