-
Title
-
How Networks Form: Homophily, Opportunity, and Balance
-
Author
-
Lewis, Kevin
-
Research Area
-
The Individual and Society
-
Topic
-
Social Networks
-
Abstract
-
Owing to rapid advances in available data and methods, social network analysis has recently been propelled into a new era: Instead of documenting patterns in static network structures, we are increasingly able to pinpoint the principles governing the evolution of these structures as well as how they emerged in the first place. In this essay, I trace the contours of this new trend. First, I describe foundational research on three mechanisms of network generation that have received particular attention in the literature: homophily, opportunity constraints, and structural balance. Next, I outline cutting‐edge research that has built on this foundation. In just the past several years, scholars have broadened prior approaches into larger, encompassing analytic frameworks; disentangled the various underlying processes that give rise to observed patterns in network structures; and distinguished between attribute‐driven network change and network‐driven attribute change—all largely thanks to advances in modeling tools that have overcome prior obstacles and enabled theoretical progress. Finally, I discuss three directions for future research. While recent scholarship has revolutionized our understanding of network dynamics, our grasp of how tie‐generating mechanisms operate and interact remains comparatively shallow; counterintuitive divisions exist between major sites of relational research and there remains much room for comparative work; and for all the promise of computational social science, there is risk that this movement will return us to the descriptive techniques of prior days but on a much larger scale.
-
Related Essays
-
Global Economic Networks (Sociology), Nina Bandelj et al.
-
Problems Attract Problems: A Network Perspective on Mental Disorders (Psychology), Angálique Cramer and Denny Borsboom
-
Migrant Networks (Sociology), Filiz Garip and Asad L. Asad
-
Interdependence, Development, and Interstate Conflict (Political Science), Erik Gartzke
-
The Development of Social Trust (Psychology), Vikram K. Jaswal and Marissa B. Drell
-
Herd Behavior (Psychology), Tatsuya Kameda and Reid Hastie
-
Emerging Trends in Social Network Analysis of Terrorism and Counterterrorism (Sociology), David Knoke
-
Network Research Experiments (Methods), Allen L. Linton and Betsy Sinclair
-
Culture, Diffusion, and Networks in Social Animals (Anthropology), Janet Mann and Lisa Singh
-
Gender and Women's Influence in Public Settings (Political Science), Tali Mendelberg et al.
-
The Role of School-Related Peers and Social Networks in Human Development (Psychology), Chandra Muller
-
Social Relationships and Health in Older Adulthood (Psychology), Theodore F. Robles and Josephine A. Menkin
-
How Do Labor Market Networks Work? (Sociology), Brian Rubineau and Roberto M. Fernandez
-
Creativity in Teams (Psychology), Leigh L. Thompson and Elizabeth Ruth Wilson
-
Identifier
-
etrds0164
-
extracted text
-
How Networks Form: Homophily,
Opportunity, and Balance
KEVIN LEWIS
Abstract
Owing to rapid advances in available data and methods, social network analysis
has recently been propelled into a new era: Instead of documenting patterns in static
network structures, we are increasingly able to pinpoint the principles governing
the evolution of these structures as well as how they emerged in the first place.
In this essay, I trace the contours of this new trend. First, I describe foundational
research on three mechanisms of network generation that have received particular
attention in the literature: homophily, opportunity constraints, and structural
balance. Next, I outline cutting-edge research that has built on this foundation.
In just the past several years, scholars have broadened prior approaches into
larger, encompassing analytic frameworks; disentangled the various underlying
processes that give rise to observed patterns in network structures; and distinguished between attribute-driven network change and network-driven attribute
change—all largely thanks to advances in modeling tools that have overcome prior
obstacles and enabled theoretical progress. Finally, I discuss three directions for
future research. While recent scholarship has revolutionized our understanding
of network dynamics, our grasp of how tie-generating mechanisms operate and
interact remains comparatively shallow; counterintuitive divisions exist between
major sites of relational research and there remains much room for comparative
work; and for all the promise of computational social science, there is risk that this
movement will return us to the descriptive techniques of prior days but on a much
larger scale.
INTRODUCTION
The study of social networks has a rich interdisciplinary history (Freeman, 2004). In contrast to standard quantitative social science—which has
been described as a “sociological meatgrinder” because its tools typically
require that all respondents in a sample are independent of one another,
thus stripping individuals from their social contexts (Barton, 1968)—it is
precisely these interdependencies among respondents that are the focus of
network research. Traditional network studies have examined patterns of
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.
1
2
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
relationships across a staggering array of settings: from interactions among
monks (Sampson, 1968) to friendships among fraternity housemates (Newcomb, 1961), business ties among Renaissance Florentine families (Padgett &
Ansell, 1993) to social event coattendance among southern socialites (Davis,
Gardner, & Gardner, 1941). And while the network approach to social
science has been gaining momentum since the “renaissance” of the 1970s
(Freeman, 2004), this movement has progressed particularly rapidly—and
taken a particular shape—in just the past decade: Instead of documenting
the characteristics of static network structures, advances in available data
and methods have enabled contemporary network analysts to study how
these structures evolve as well as how they emerged in the first place.
This new frontier of research has returned network analysts to questions
at the heart of the sociological enterprise: What are the interpersonal affinities and antagonisms that characterize a given social structure (Laumann &
Senter, 1976)? In other words, if we recognize that patterns of human relationships reflect both the deliberate choices of individuals as well as the constraints in which these decisions are made, the “choice” part of this equation
provides direct insight into the structure of intergroup boundaries: who is
willing and unwilling to affiliate with whom, and therefore the strength of
various social cleavages. Unfortunately, disentangling the role of choice from
the role of constraint—the classic sociological distinction between agency
and structure—has not always been so easy.
FOUNDATIONAL RESEARCH
Contemporary research on the genesis and evolution of social networks
tends to draw on three primary traditions of research. Each tradition
corresponds to a distinct mechanism whereby social ties are created and
maintained. Interestingly, the three traditions developed somewhat independently of one another, and it has only been relatively recently that
these strands have been integrated into a more comprehensive analytic
framework.
HOMOPHILY
Arguably the most well-known principle in the networks literature is that
of homophily: the notion that “birds of a feather flock together” or “similarity breeds attraction.” Homophily has been studied across a striking array
of settings, relationships, and attributes. This research was deftly reviewed
by McPherson, Smith-Lovin, and Cook (2001)—a paper that continues to be
one of the most-cited works in sociology today. The motivating principle
behind this research is simple: Insofar as similar individuals seek out one
How Networks Form: Homophily, Opportunity, and Balance
3
another and dissimilar individuals avoid one another, the ensuing patterns
of relationships have powerful implications for the information we receive,
the attitudes we form, and the interactions we experience (McPherson et al.,
2001). Importantly, as mentioned, the strength of this tendency itself can be
used as a barometer of social relations: Rather than asking individuals about
intergroup prejudices using various forms of social distance measures (e.g.,
Bogardus, 1947), we can instead infer these prejudices from observed patterns of social ties (e.g., Kalmijn, 1991).
One noteworthy feature of this body of research is the surprising degree
of ambiguity and inconsistency regarding what the label of “homophily”
signifies in the first place. Literally, the term homophily is derived from the
Greek words homoios (equal, similar) and philia (friendship, love, affection;
see Skopek, Schulz, & Blossfeld, 2011, note 1). In other words, homophily
refers to a preference—the enhanced degree of psychological attraction
between two similar birds—rather than a pattern—the actual observed
tendency for similar birds to flock together more frequently than dissimilar
birds. However, the term homophily is often confusingly applied to both.
In addition, regardless of whether the focus is preference or pattern, an
important question is how exactly to measure this phenomenon; and in
particular, what is the baseline degree of attraction/affiliation that should
be used as a basis for comparison. In other words, what level of homophily
should we expect merely as a consequence of chance? This discussion
leads naturally to the next mechanism of tie formation—a mechanism that
recognizes there is at least one important reason to expect similar birds to
flock together that has nothing to do with their preference to do so.
OPPORTUNITY
Social networks are hardly structured by individual preferences alone.
Rather, these preferences operate, most basically, within the constraints of
who is available to affiliate with in the first place. Let us imagine we were
to acquire a list of who is dating whom at a certain high school—and we
discovered that, lo and behold, every single romantic couple on this list
is composed of two students who both have brown eyes. One possible
explanation for this finding is that brown-eyed students display a strong
degree of preference for one another (brown-eyed student homophily).
Another possibility, however, is that the entire school is simply full of
brown-eyed people—in which case the pattern we observe tells us nothing
about students’ preferences and everything about the limited pool of dating
partners from which each student had to choose.
In order to accurately assess the role of homophily in governing the selection of social ties, therefore, we must first account for the opportunity structure
4
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
from which these ties were selected; and as a baseline expectation without
any role of homophily, we would expect that patterns of interpersonal affiliation would naturally reflect the patterns of characteristics in a population.
This basic, intuitive insight—one that, surprisingly, remains ignored in many
studies of social networks today—was developed and tested in two foundational texts by Blau (Blau 1977; Blau & Schwartz, 1984). Importantly, Feld
(1981, 1982)’s concept of social foci provides an additional, more precise theoretical tool for examining the impact of opportunity structures on social
networks. Drawing on Homans’ (1950) social elements of activity, interaction,
and sentiments, Feld defines a social focus as “a social, psychological, legal,
or physical entity around which joint activities are organized”—as a consequence of which, individuals whose activities are organized around the same
focus will be much more likely to form a social tie (Feld, 1981). In other words,
“opportunity structures” of potential contacts are not distributed randomly
in physical space; rather, they are concentrated at areas of joint activity.
BALANCE
As noted by Feld (1981), foci posit a “sociological” explanation for why
proximity breeds affiliation: Two individuals are much more likely to
become connected if they share a social context in common. A complement
to this approach is the “psychological” explanation provided by balance
theory—where two people who share a friend in common are more likely to
become friends themselves.
Balance theory was originally formulated by Heider (1946), formalized
by Cartwright and Harary (1956), and further developed by Davis (1963).
Its central insight is that certain configurations of relationships (namely,
“unbalanced” configurations) produce psychological strain for all parties
involved and tend to be avoided. It is very uncomfortable, for instance, to be
friends with two people who intensely dislike each other—and so the only
options for alleviating this strain (and restoring “balance” to the triad) are for
the focal individual to abandon one or the other of the friendships (i.e., to take
a side) or to try to get the two friends to make amends and start liking each
other. The upshot of this process is that “unclosed triads” in social networks
(configurations where A is friends with B and B is friends with C but A and
C are not themselves friends) are extremely rare (Davis, 1970), because there
are psychological pressures that induce A and C to become friends (thereby
“closing” the triad). It is important to note that the scope of structural balance
predictions is not limited to configurations of three people, but extends
also to more or fewer persons (as well as to configurations with nonhuman
objects about which two people may have attitudes). Another regularity in
social networks, for instance, is the tendency for the psychological strain
How Networks Form: Homophily, Opportunity, and Balance
5
produced by asymmetric ties (situations where A is friends with B but B is
not friends with A) to be avoided: If B does not eventually reciprocate the
tie, A is likely to withdraw it (e.g. Hallinan, 1978).
CUTTING-EDGE RESEARCH
While research on each of the above mechanisms—homophily, opportunity,
and structural balance—has progressed for nearly a century, it is only relatively recently that scholars have developed integrated theoretical frameworks involving all three. This development is due, in large part, to recent
advances in available methods for analyzing cross-sectional and longitudinal
social network data.
THEORETICAL DEVELOPMENTS
Theoretical advances in recent years have developed along three fronts: first,
expanding the above-mentioned three mechanisms into a broader, encompassing framework; second, differentiating between observed patterns and
underlying processes and disentangling the unique role of various processes
in generating these patterns; and third, expanding focus to include fixed as
well as endogenously changing individual attributes.
Broader Analytic Framework. In an important, recent contribution, Rivera,
Soderstrom, and Uzzi (2010) provide a thorough review of the various
possible mechanisms that lead to the formation and persistence of social network ties. They classify these mechanisms under three headings: assortative,
proximity, and relational mechanisms. This framework is a natural extension
of the three processes described earlier—homophily is but one assortative
mechanism, social foci are but one proximity mechanism, and structural
balance is but one relational mechanism—but there are other mechanisms of
each type, as well. For instance, in addition to seeking similarity in a partner,
there are various reasons—often in collaborative settings (e.g., Moody,
2004)—for preferring dissimilarity in interaction partners, or heterophily.
Physical proximity influences not just the creation of ties through meeting
opportunities but also the likelihood of ties being maintained over long distances (e.g., Martin & Yeung, 2006). And “structural” mechanisms that have
nothing to do with individual attributes include not only reciprocity and
triadic closure but also dynamics of repetition (the tendency for interactions
to be repeated) and degree (the tendency for individuals with many ties
to accumulate new ties at a faster rate than individuals with fewer ties, or
“preferential attachment”).
6
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
Disentangling Underlying Processes. There is a pervasive tendency in
research on social networks—and on homophily in particular—to assume
that observed patterns in network structures directly parallel the underlying processes that generated them. In other words, if certain types of
individuals (e.g., from the same racial background) tend to be connected
in a social network, this must be because individuals from the same racial
background prefer to connect with one another (i.e., racial homophily).
In much the same way that similarity among network affiliates could be
produced by either homophily or opportunity constraints, recent research
has begun to acknowledge that the relationship between underlying
processes and observed outcomes is still more complex. This work is not
without its forbears: Granovetter (1973), for instance, noted that triadic
closure could equally plausibly result from structural balance, homophily,
or opportunity mechanisms; Feld (1982) noted that social foci bring together
disproportionately homogeneous sets of people, leading to the appearance
of homophily; and Rivera et al. (2010) note that work on each mechanism
tends to progress in relative isolation, and we know relatively little about
the relative strength of various network-generating factors (see also Kalmijn,
1998).
Building on these leads, a handful of recent papers have not only mapped a
number of theoretical relationships between different underlying processes
and the same observed pattern; they have statistically disentangled and
quantified the relative contribution of each process, as well. Goodreau, Kitts,
and Morris (2009), for instance, consider the relationship between three
underlying processes (sociality, selective mixing, and triadic closure) and
three observed outcomes (a network’s degree distribution, mixing pattern,
and level of transitivity). Kossinets and Watts (2009) investigate the origins
of “homophily” in a large university community, and show that even a
modest degree of homophilous preference can be compounded over many
“generations” of interaction due to triadic and focal closure. Meanwhile,
Wimmer and Lewis (2010) distinguish between racial “homophily”—the
preference for a racially similar partner—and “homogeneity”—the pattern
of racial similarity in social networks—and show that the importance of
racial homophily is distorted unless alternative mechanisms of tie formation
(ethnic homophily, sociality, and structural balance) are also taken into
account.
Network and Behavioral Coevolution. A final source of theoretical progress in
fact has to do with another mechanism that can generate observed similarity
between partners in a social network—but only with respect to characteristics
that, unlike racial background, potentially change over time. In other words,
How Networks Form: Homophily, Opportunity, and Balance
7
just as network “homogeneity” in certain characteristics can be produced
by homophily, opportunity, and structural balance, so too can homogeneity be generated by peer influence—the tendency to adopt the characteristics and behaviors of our peers. In other words, are friends similar to one
another because similar people become friends or because friends become
more similar over time? While the puzzle of social selection and peer influence has been a focus of research for decades (e.g., Kandel, 1978), a number of
methodological issues (Lyons, 2011; Noel & Nyhan, 2011; Shalizi & Thomas,
2011) have plagued traditional and contemporary work on this topic; and it
is only relatively recently that methodological advances (e.g., Steglich, Snijders, & Pearson, 2010) have enabled a clearer answer to this longstanding
theoretical puzzle. Perhaps unsurprisingly, results vary tremendously based
on the context, relationship, and attribute at hand: While Mercken, Snijders,
Steglich, Vartiainen, and de Vries (2010) find that selection as well as influence processes play an important role in adolescent smoking behavior and
de Klepper, Sleebos, van de Bunt, and Agneessens (2010) show that students
adjust their own military discipline to that of their friends, Lewis, Gonzalez,
and Kaufman (2012) find that peer influence among friends on Facebook is
virtually nonexistent.
METHODOLOGICAL ADVANCES
The above-mentioned progress was largely enabled by unprecedented
developments in available methods for analyzing social network data.
While network analytic tools have rapidly advanced on a number of fronts
(see review in Snijders, 2011), of particular note here are two methods that
are capable of statistically disentangling the various underlying mechanisms
that give rise to observed patterns: one, exponential random graph modeling, which uses cross-sectional network data to understand how observed
social networks were generated (Robins, Pattison, Kalish, & Lusher, 2007);
and the other, stochastic actor-based modeling, which uses longitudinal
network data to understand how observed social networks evolve over time
(Snijders, van de Bunt, & Steglich, 2010). While these are certainly not the
only techniques available for examining the complex underlying processes
governing social network emergence and evolution, they do provide several
distinct advantages compared to prior approaches (e.g., Steglich et al., 2010);
feature recently developed model terms and estimation techniques that help
overcome the limitations of prior specifications (e.g., Robins, Snijders, Wang,
Handcock, & Pattison, 2007); and are increasingly featured in mainstream
sociological publications (e.g., Schaefer, Kornienko, & Fox, 2011; Srivastava
& Banaji, 2011; Wimmer & Lewis, 2010).
8
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
KEY ISSUES FOR FUTURE RESEARCH
As a consequence of the advances described, recent scholarship has built on
the foundation created by decades of prior work to advance social network
analysis to a new level. Rather than describing patterns in social networks
and inferring from these patterns the underlying processes that generated
them, the relationship between process and pattern is itself increasingly
scrutinized—leading to knowledge that is genuinely explanatory rather
than descriptive. Still, there is room for progress in three directions.
DEPTH
First, our grasp of the processes governing network formation and evolution
remains relatively shallow—a necessary consequence of tackling comparatively “superficial” issues first (e.g., what is the relative importance of
various mechanisms) before plunging deeper (e.g., what are the interactions
between these mechanisms and how does their importance vary over
time). For instance, does the strength of triadic closure vary across racial
groups? What types of social foci are particularly conducive to friendship
development and maintenance? And how does the importance of each
mechanism vary over time as networks emerge and evolve (cf. Rivera et al.,
2010)? Greater depth can also be achieved by returning to the basics of
how network analysts approach their data and the assumptions implicit
in much of our theory and methods. How is the process of tie formation
different from that of tie maintenance—not to mention tie dissolution?
How do these dynamics vary when we are dealing with valued, rather
than dichotomous, social relationships? Transcending such distinctions
altogether, future research would benefit from greater integration of qualitative and quantitative approaches focused on the meaning of ties, nodes,
and groups (e.g., Fuhse, 2009; Pachucki & Breiger, 2010; Yeung, 2005), and
even by returning to longstanding, taken-for-granted mechanisms such as
homophily and acknowledging how little we actually know about how they
work in practice (see DiMaggio & Garip, 2012, p. 111).
BREADTH
Second, while our understanding of network dynamics has quickly
expanded to new settings and relationships, curious gaps—and the potential for still further improvement—remain. For instance, social network
analysis is fundamentally concerned with the study of interpersonal relationships, of which there is arguably no more important example than
romantic ties. However, the literature on romantic relationships (which
has largely focused on marriage due to the importance of the marriage
How Networks Form: Homophily, Opportunity, and Balance
9
relationship and the availability of accurate, nationally representative data)
has developed largely independently from the broader literature on social
networks. This division is understandable due to the unique nature of
“marriage” as a network tie (individuals can have, at most, only one “alter”
at any given time, making for a rather uninteresting social network) as
well as the unique nature of marriage data (while complete social network
data provide information on ties as well as non-ties, marriage records, for
instance, include only the former, precluding the application of many network analytic techniques). On the other hand, this division has also stymied
progress because each line of research has been slow to take advantage
of the theoretical insights developed by the other. More generally, there is
great room for comparative work across networks with multiple types of
relationships in multiple types of settings, as well as the development of
additional analytic techniques that can handle multiple types of ties in the
same setting, that is, network multiplexity (McPherson et al., 2001).
SCALE
Finally, there is increasing enthusiasm for a movement commonly referred
to as “computational social science”—in which network analysis will
almost certainly play a central role. Owing to the ubiquity of electronic
communication—and especially the explosion of social network sites such as
Facebook—digital traces of human interaction are available to researchers in
unprecedented quantities. While traditional network datasets (such as those
cited in the introduction) consisted of networks on the order of tens of nodes,
we are now seeing available network datasets on the order of hundreds,
thousands, and sometimes millions of nodes—leading some to term this
the moment of “big data” in social science. The promise and obstacles such
a movement entails have been discussed (and foreshadowed) elsewhere
(e.g., Lazer et al., 2009; Rogers, 1987). What has received less attention is the
tension this situation has created for forward-thinking network analysts. We
have reached a stage of rapid methodological development in which we are
able to do more today with the same data than we have ever done in the
past. However, this methodological development has not yet caught up with
the scale of available data—leading to situations in which available methods
are applied beyond their intentions (and worse, beyond their assumptions);
or, more commonly, they are not applied at all. Ironically, then, for all of its
promise, the movement of computational social science risks pushing us
back to the very type of descriptive analyses that characterized foundational
research on social networks—because we simply do not yet have the tools
available to do anything else.
10
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
Perhaps the most promising area for future development, then, is striking a
productive compromise between the two: acknowledging what we can and
cannot do with the massive quantities of data that are increasingly available
and ensuring that explanatory breadth and depth are not sacrificed because
our reach exceeds our grasp. The three mechanisms of homophily, opportunity, and balance will surely continue to be central to our understanding of
how networks form and evolve. But particularly at a time when an increasing
proportion of human interaction occurs digitally and technology is used as
much to maintain social ties (Ellison, Steinfield, & Lampe, 2007) as to create
them (Rosenfeld & Thomas, 2012), the next great advance in social network
thinking and research remains to be seen.
REFERENCES
Barton, A. H. (1968). Bringing society back in: Survey research and macromethodology. American Behavioral Scientist, 12, 1–9.
Blau, P. M. (1977). Inequality and heterogeneity: A primitive theory of social structure. New
York, NY: Free Press.
Blau, P. M., & Schwartz, J. E. (1984). Crosscutting social circles: Testing a macrostructural
theory of intergroup relations. Orlando, FL: Academic Press.
Bogardus, E. S. (1947). Measurement of personal-group relations. Sociometry, 10,
306–311.
Cartwright, D., & Harary, F. (1956). Structural balance: A generalization of Heider’s
theory. Psychological Review, 63, 277–293.
Davis, A., Gardner, B. B., & Gardner, M. R. (1941). Deep south. Chicago, IL: University
of Chicago press.
Davis, J. A. (1963). Structural balance, mechanical solidarity, and interpersonal relations. American Journal of Sociology, 68, 444–462.
Davis, J. A. (1970). Clustering and hierarchy in interpersonal relations: Testing two
graph theoretical models on 742 sociomatrices. American Sociological Review, 35,
843–851.
de Klepper, M., Sleebos, E., van de Bunt, G., & Agneessens, F. (2010). Similarity in
friendship networks: Selection or influence? The effect of constraining contexts
and non-visible individual attributes. Social Networks, 32, 82–90.
DiMaggio, P., & Garip, F. (2012). Network effects and social inequality. Annual Review
of Sociology, 38, 93–118.
Ellison, N. B., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook “friends:”
Social capital and college students’ use of online social network sites. Journal of
Computer-Mediated Communication, 12, 1143–1168.
Feld, S. L. (1981). The focused organization of social ties. American Journal of Sociology,
86, 1015–1035.
Feld, S. L. (1982). Social structural determinants of similarity among associates. American Sociological Review, 47, 797–801.
How Networks Form: Homophily, Opportunity, and Balance
11
Freeman, L. C. (2004). The development of social network analysis: A study in the sociology
of science. Vancouver, BC: Empirical Press.
Fuhse, J. A. (2009). The meaning structure of social networks. Sociological Theory, 27,
51–73.
Goodreau, S. M., Kitts, J. A., & Morris, M. (2009). Birds of a feather, or friend of a
friend? Using exponential random graph models to investigate adolescent social
networks. Demography, 46, 103–125.
Granovetter, M. (1973). The strength of weak ties. American Journal of Sociology, 78,
1360–1380.
Hallinan, M. T. (1978). The process of friendship formation. Social Networks, 1,
193–210.
Heider, F. (1946). Attitudes and cognitive organization. Journal of Psychology, 21,
107–112.
Homans, G. C. (1950). The human group. New York, NY: Harcourt, Brace & World.
Kalmijn, M. (1991). Status homogamy in the United States. American Journal of
Sociology, 97, 496–523.
Kalmijn, M. (1998). Intermarriage and homogamy: Causes, patterns, trends. Annual
Review of Sociology, 24, 395–421.
Kandel, D. B. (1978). Homophily, selection, and socialization in adolescent friendships. American Journal of Sociology, 84, 427–436.
Kossinets, G., & Watts, D. J. (2009). Origins of homophily in an evolving social network. American Journal of Sociology, 115, 405–450.
Laumann, E. O., & Senter, R. (1976). Subjective social distance, occupational stratification, and forms of status and class consciousness: A cross-national replication
and extension. American Journal of Sociology, 81, 1304–1338.
Lazer, D., Pentland, A., Adamic, L., Aral, S., Barabási, A. L., Brewer, Devon, … Van
Alstyne, M. (2009). Computational social science. Science, 323, 721–723.
Lewis, K., Gonzalez, M., & Kaufman, J. (2012). Social selection and peer influence in
an online social network. Proceedings of the National Academy of Sciences of the United
States of America, 109, 68–72.
Lyons, R. (2011). The spread of evidence-poor medicine via flawed social-network
analysis. Statistics, Politics, and Policy, 2(1), Article 2.
Martin, J. L., & Yeung, K.-T. (2006). Persistence of close personal ties over a 12-year
period. Social Networks, 28, 331–362.
McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily
in social networks. Annual Review of Sociology, 27, 415–444.
Mercken, L., Snijders, T. A. B., Steglich, C., Vartiainen, E., & de Vries, H. (2010).
Dynamics of adolescent friendship networks and smoking behavior. Social
Networks, 32, 72–81.
Moody, J. (2004). The structure of a social science collaboration network: Disciplinary
cohesion from 1963 to 1999. American Sociological Review, 69, 213–238.
Newcomb, T. M. (1961). The acquaintance process. New York, NY: Holt, Rinehart and
Winston.
Noel, H., & Nyhan, B. (2011). The “unfriending” problem: The consequences of
homophily in friendship retention for causal estimates of social influence. Social
Networks, 33, 211–218.
12
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
Pachucki, M. A., & Breiger, R. L. (2010). Cultural holes: Beyond relationality in social
networks and culture. Annual Review of Sociology, 36, 205–224.
Padgett, J. F., & Ansell, C. K. (1993). Robust action and the rise of the medici,
1400–1434. American Journal of Sociology, 98, 1259–1319.
Rivera, M. T., Soderstrom, S. B., & Uzzi, B. (2010). Dynamics of dyads in social
networks: Assortative, relational, and proximity mechanisms. Annual Review of
Sociology, 36, 91–115.
Robins, G., Pattison, P., Kalish, Y., & Lusher, D. (2007). An introduction to exponential
random graph (p*) models for social networks. Social Networks, 29, 173–191.
Robins, G., Snijders, T., Wang, P., Handcock, M., & Pattison, P. (2007). Recent developments in exponential random graph (p*) models for social networks. Social
Networks, 29, 192–215.
Rogers, E. M. (1987). Progress, problems and prospects for network research: Investigating relationships in the age of electronic communication technologies. Social
Networks, 9, 285–310.
Rosenfeld, M. J., & Thomas, R. J. (2012). Searching for a mate: The rise of the internet
as a social intermediary. American Sociological Review, 77, 523–547.
Sampson, S. F. (1968). A novitiate in a period of change: An experimental and case study of
social relationships. PhD thesis. Cornell University.
Schaefer, D. R., Kornienko, O., & Fox, A. M. (2011). Misery does not love company:
Network selection mechanisms and depression homophily. American Sociological
Review, 76, 764–785.
Shalizi, C. R., & Thomas, A. C. (2011). Homophily and contagion are generically confounded in observational social network studies. Sociological Methods & Research,
40, 211–239.
Skopek, J., Schulz, F., & Blossfeld, H.-P. (2011). Who contacts whom? Educational
homophily in online mate selection. European Sociological Review, 27, 180–195.
Snijders, T. A. B., van de Bunt, G., & Steglich, C. (2010). Introduction to stochastic
actor-based models for network dynamics. Social Networks, 32, 44–60.
Snijders, T. A. B. (2011). Statistical models for social networks. Annual Review of Sociology, 37, 131–153.
Srivastava, S. B., & Banaji, M. R. (2011). Culture, cognition, and collaborative networks in organizations. American Sociological Review, 76, 207–233.
Steglich, C., Snijders, T. A. B., & Pearson, M. (2010). Dynamic networks and behavior:
Separating selection from influence. Sociological Methodology, 40, 329–393.
Wimmer, A., & Lewis, K. (2010). Beyond and below racial homophily: ERG models
of a friendship network documented on Facebook. American Journal of Sociology,
116, 583–642.
Yeung, K.-T. (2005). What does love mean? Exploring network culture in two network settings. Social Forces, 84, 391–420.
FURTHER READING
Blau, P. M. (1977). Inequality and heterogeneity: A primitive theory of social structure. New
York, NY: Free Press.
How Networks Form: Homophily, Opportunity, and Balance
13
Freeman, L. C. (2004). The development of social network analysis: A study in the sociology
of science. Vancouver, BC: Empirical Press.
McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily
in social networks. Annual Review of Sociology, 27, 415–444.
Rivera, M. T., Soderstrom, S. B., & Uzzi, B. (2010). Dynamics of dyads in social
networks: Assortative, relational, and proximity mechanisms. Annual Review of
Sociology, 36, 91–115.
Snijders, T. A. B. (2011). Statistical models for social networks. Annual Review of
Sociology, 37, 131–153.
KEVIN LEWIS SHORT BIOGRAPHY
Kevin Lewis is an Assistant Professor of Sociology at the University of
California, San Diego and a Faculty Associate at the Berkman Center for
Internet & Society at Harvard University. Lewis’ research focuses on the
formation and evolution of social networks and attempts to identify the
underlying micromechanisms responsible for the generation of observed
patterns. To do this, he has employed a number of large-scale datasets
diverse in nature—from Facebook friendships among college students to
messages sent among online dating site users to recruitment ties among
online activists—and utilized recent advances in network modeling
techniques for cross-sectional and longitudinal data. His work has been
published in the American Journal of Sociology, the Proceedings of the National
Academy of Sciences, Social Networks, Sociological Science, and the Journal of
Computer-Mediated Communication.
Personal webpage: http://sociology.ucsd.edu/faculty/KevinLewis.shtml
RELATED ESSAYS
Global Economic Networks (Sociology), Nina Bandelj et al.
Problems Attract Problems: A Network Perspective on Mental Disorders
(Psychology), Angélique Cramer and Denny Borsboom
Migrant Networks (Sociology), Filiz Garip and Asad L. Asad
Interdependence, Development, and Interstate Conflict (Political Science),
Erik Gartzke
The Development of Social Trust (Psychology), Vikram K. Jaswal and Marissa
B. Drell
Herd Behavior (Psychology), Tatsuya Kameda and Reid Hastie
Emerging Trends in Social Network Analysis of Terrorism and Counterterrorism (Sociology), David Knoke
Network Research Experiments (Methods), Allen L. Linton and Betsy Sinclair
Culture, Diffusion, and Networks in Social Animals (Anthropology), Janet
Mann and Lisa Singh
14
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
Gender and Women’s Influence in Public Settings (Political Science), Tali
Mendelberg et al.
The Role of School-Related Peers and Social Networks in Human Development (Psychology), Chandra Muller
Social Relationships and Health in Older Adulthood (Psychology), Theodore
F. Robles and Josephine A. Menkin
How Do Labor Market Networks Work? (Sociology), Brian Rubineau and
Roberto M. Fernandez
Creativity in Teams (Psychology), Leigh L. Thompson and Elizabeth Ruth
Wilson