CONNECTIONS - INSNA
CONNECTIONS - INSNA
CONNECTIONS - INSNA
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
<strong>CONNECTIONS</strong><br />
Configurations of Homophily<br />
INTRODUCTION<br />
Folk wisdom tells us that “birds of a feather<br />
flock together” yet, at the same time “opposites<br />
attract.” The contradiction between these<br />
familiar adages points to a fundamental feature<br />
of relationships: their multidimensionality. The<br />
forces that attract and repel people operate in<br />
conjunction along several dimensions<br />
simultaneously. This is evident in Blau’s (1977)<br />
notion of multiform heterogeneity, which<br />
implies that the desire for homophily on some<br />
dimensions can make heterophily on other<br />
dimensions inevitable. Consequently,<br />
individuals within relationships are similar in<br />
many respects, while dissimilar in other aspects:<br />
Male and female birds flock together.<br />
This study moves beyond the unidimensional<br />
treatment of homophily that dominates previous<br />
research by developing a multidimensional<br />
conceptualization. Homophily is characterized<br />
as a combination of similarity and dissimilarity<br />
across several dimensions, referred to as<br />
configurations of homophily. These<br />
configurations can be analyzed to better<br />
understand the interrelations between<br />
dimensions that structure interpersonal<br />
associations. Of particular interest are which<br />
configurations of homophily occur more often<br />
than expected by chance.<br />
Introducing multidimensional homophily<br />
proceeds through two steps. First, I define<br />
relations as configurations of homophily with<br />
the graphical assistance of lattices. Just as<br />
scatterplots are valuable for examining<br />
correlational data, lattices provide a useful<br />
representation of configurational data. I present<br />
a lattice visualization technique that conveys<br />
both the distribution of dyads across homophily<br />
configurations and the logical relationship<br />
between configurations. This tool helps portray<br />
the relationships inherent to combinatorial data<br />
and can assist in the evaluation of such data.<br />
Second, to understand which combinations of<br />
homophily are most salient across relations, I<br />
examine ego network data from the 1985<br />
General Social Survey. I use Qualitative<br />
Comparative Analysis (QCA) to identify the<br />
combinations of similarity and dissimilarity that<br />
underlie the configurations of homophily in<br />
observed dyads. This process uncovers how<br />
dimensions of homophily interact to create more<br />
dyads than expected by chance.<br />
Homophily Research<br />
Social scientists have documented homophily as<br />
one of the most persistent features of human<br />
society (McPherson, Smith-Lovin, & Cook,<br />
2001). Homophily is defined as the tendency for<br />
people to associate with others who are similar,<br />
rather than dissimilar, to themselves.<br />
Homophily is important because of its<br />
implications for the movement of ideas and<br />
people through society. While communication<br />
between similar individuals tends to be more<br />
effective (Rogers & Bhowmik, 1970),<br />
homogeneous networks also tend to constrain<br />
attitudes to be more “closed-minded” (Laumann,<br />
1963) or conservative (Bienenstock, Bonacich,<br />
& Oliver, 1990). Beyond attitudes, the presence<br />
of homophily necessitates that any “cultural,<br />
behavioral, genetic, or material information that<br />
flows through networks will tend to be<br />
localized” (McPherson et al., 2001, p. 416).<br />
Homophily also has consequences for social<br />
mobility to the extent that positions of power<br />
and authority are correlated with ascribed<br />
characteristics of their occupants. For instance,<br />
because women are less likely to occupy<br />
powerful positions within organizations, such<br />
individuals must form strategic, heterophilous<br />
ties that are less efficient in order to gain access<br />
to valuable information and resources (Ibarra<br />
1992). Thus, it is important to understand<br />
homophily as a means of ascertaining the types<br />
of social distance that are least often bridged.<br />
Research has documented the persistence of<br />
homophily in groups and interpersonal<br />
relationships across a wide range of dimensions<br />
(for an overview see McPherson et al., 2001).<br />
Race and ethnic homophily is perhaps the<br />
strongest dimension in the United States and is<br />
observed across most forms of relations.<br />
22