Grappling with Grouping III Social Network Analysis
David Henry David Henry
University of Illinois at Chicago Allison Dymnicki
American Institutes for Research
Families and Communities Social Network and N ti I fl P j t
Acknowledgments
Research Group Patrick Tolan Deborah Gorman-SmithNormative Influence Projects Fern Chertok Daneen Deptula Allison Dymnicki Michael Schoeny y Jane Jegerski Christopher Keys Kimberly Kobus
Jennifer Watling Neal Jennifer Watling Neal Zachary Neal
Michael Schoeny
Acknowledgments
• This work was supported by grants from the
Centers for Disease Control and Prevention and the National Institute of Justice. The content of this
presentation is solely the responsibility of the
p y p y
authors and does not necessarily represent the official views of the funders.
• For more information visitFor more information, visit www ihrp uic eduwww.ihrp.uic.edu.
I. Cluster Analysis
Grappling with Grouping
II. Clustering methods for binary variables III. Social Network Analysis
Central theme: Clustering approximates uniqueness in the same way that a sample mean approximates a the same way that a sample mean approximates a population.
N
k A
l i
• Widely used in community psychology research
Network Analysis
• 28 studies since 2000 in just two journals (AJCP, JCP)j j ( , )
– Search terms: “Network Analysis”
• Similar search using the term “cluster analysis” returned 14 studies.
Community Studies Employing Network Analysis
y
p y g
y
Study Variables Type of Analysis
Swindle et al., 2000
Positive and negative social transactions in networks of
HIV+ persons Rating scales Hirsch et al., 2002 Differences in strength of ties by race Ego network Ying, 2002 Social network composition of Taiwanese graduate students Rating scales Langhout, 2003 A single case study using ego networks Ego network Fleisher & Krienert, 2004
Violence among female gang members increases before
pregnancy and decreases afterward Qualitative Interviews Criticism practical support were significant predictors of
Levendosky et al., 2004
Criticism, practical support were significant predictors of
mental health for battered women. Rating scales Toohey et al., 2004
Differences between nearly homeless and housed women
on beliefs about netweok members as housing resources Rating scales Zea et al., 2004
Target‐specific factors were related to the probability of
disclosure. Rating scales Chia, 2006 Sociometric nominations in a work organization Sociometric Knowlton & Latkin, 2007 Ego networks Ego network Dominguez & Maya‐Lariego, 2008 Ego network support characteristics Ego network Pernice Duca 2008 Social Support in clubhouse mental health programs Rating scales
Advancing Health Practice and Policy through Collaborative Research
Community Studies Employing Network Analysis
y
p y g
y
Study Variables Type of
Analysisy
Toro et al., 2008 Social Support in homeless adults ‐ego networks Ego network Campo et al., 2009 Convergent and discriminant validity with other measures Rating scales Latkin et al., 2009
Network drug use contributed to perceptions of neighborhood
disorder. Ego network Neal, 2009 Density, centrality, and relational aggression Informant Trotter & Allen, 2009 Ego networks, qualitative analysis Ego network Crowe, 2010 Personal and Organizational Community Networks Sociometric
L t l 2010 C it C N t k S i t i
Lugue et al., 2010 Community Cancer Network Sociometric Prelow et al., 2010
Social Support buffered ecological risk effects on psychological
distress Rating scales Haines et al., 2011
Network characteristics of an interdisciplinary collaboration
based on multiple types of relationships Sociometric Neal et al (2011) Cohesion vs structural similarity in teacher advice networks Sociometric
Advancing Health Practice and Policy through Collaborative Research
V i t
f N t
k M th d
Variety of Network Methods
in Community Psychology Studies
Type Number Sociometric nominations 5 Informant-based Methods 1 Ego-networks 7 Rating Scales 8 Rating Scales 8 Qualitative 1
Outline
• Overview of methods SociometricsOutline
– Sociometrics – Informants– Ego-networksEgo networks
– Dynamic • For each (-1)
– Theory/method/measures – Software
Advancing Health Practice and Policy through Collaborative Research – Strengths and limitations
• Data source: Relationships
Network Analysis with Sociometrics
• Data source: Relationships
– “Who are your friends?” (Kobus & Henry, 2009)
– “What organizations do you belong to?” (Crowe, a o ga a o s do you be o g o (C o e,
2010)
• Analysis: Matrix and Graph Representations
A B C D E F G A 0 1 0 1 0 0 0 B 1 0 0 1 0 0 0 B E C 0 0 0 0 1 1 0 D 1 1 0 0 0 0 0 E 0 0 1 0 0 1 0 A C D E F
Advancing Health Practice and Policy through Collaborative Research
F 1 0 0 0 0 0 0
G 0 0 0 0 0 1 0 G
Sociometrics: Network Measures
28 . 0 42 / 12 ) 1 (
g g X DENSITY where X = relationships (ties) = 12 ) (g gg = network size (# of potential relationships) = 42
A B C D E F G Σ A 0 1 0 1 0 0 0 2 A 0 1 0 1 0 0 0 2 B 1 0 0 1 0 0 0 2 C 0 0 0 0 1 1 0 2 D 1 1 0 0 0 0 0 2 A B E F D 1 1 0 0 0 0 0 2 E 0 0 1 0 0 1 0 2 F 1 0 0 0 0 0 0 1 G 0 0 0 0 0 1 0 1 A C D F G
Advancing Health Practice and Policy through Collaborative Research
Σ 3 2 1 2 1 3 0 12
G
Sociometrics: Networks Measures
• Mutuality Index 0.77 ) 1 ( ) 1 ( 2 2 2 2 2 2 2 2 L L g L L L M g
M = number of mutual relationships = 5 g = network size = 7
L f th td f th t t l t k 12
)
(g 2
L = sum of the outdegree of the total network = 12
L2 = sum of squares of the outdegree of the total network =22
A B C D E F G Σ A B E F A 0 1 0 1 0 0 0 2 B 1 0 0 1 0 0 0 2 C 0 0 0 0 1 1 0 2 A C D F G D 1 1 0 0 0 0 0 2 E 0 0 1 0 0 1 0 2 F 1 0 0 0 0 0 0 1
Advancing Health Practice and Policy through Collaborative Research
G
G 0 0 0 0 0 1 0 1
Sociometrics: Network Measures
• Boundary Density (Hirsch, 1980)
TTactual = number of actual ties across subgroups = 2
083 . 0
possible actual T T BDTpossible = number of possible ties across subgroups = 24
A B C D E F G Σ A 0 1 0 1 0 0 0 2 B E A 0 1 0 1 0 0 0 2 B 1 0 0 1 0 0 0 2 C 0 0 0 0 1 1 0 2 D 1 1 0 0 0 0 0 2 A C D F G D 1 1 0 0 0 0 0 2 E 0 0 1 0 0 1 0 2 F 1 0 0 0 0 0 0 1 G 0 0 0 0 0 1 0 1
Advancing Health Practice and Policy through Collaborative Research
G Σ 3 2 1 2 1 3 0 12
Sociometrics: Measures of Individuals
Mean Geodesic Distance
j ji j ij D or Dwhere D = Distance and B = Reachability
j ij j ij B o BIn: 1.33 for F, 2.0 for D and 2.16 for G
A B C D E F G Σ A B C E F A 0 1 0 1 0 0 0 2 B 1 0 0 1 0 0 0 2 C 0 0 0 0 1 1 0 2 C D G D 1 1 0 0 0 0 0 2 E 0 0 1 0 0 1 0 2 F 1 0 0 0 0 0 0 1 G 0 0 0 0 0 1 0 1
Advancing Health Practice and Policy through Collaborative Research
G 0 0 0 0 0 1 0 1
Sociometrics: Measures of Individuals
Sociometrics: Measures of Individuals
• Position A B E F – Member – Liaison – Isolate C D G
• Liaisons > Members or Isolates on tobacco and alcohol use (2 studies)
(2 studies)
• Members and Isolates more influenced by peer substance use than Liaisons.
Sociometrics: Statistical Models Example
D d h t d t b t i d
Dyads have a tendency to become triads:
“birds of a feather? or “friends of friends?”
We can model the likelihood of triad closure, but the chance models are complex
Random graph models and double permutation tests provide alternatives suitable for predicting network
Advancing Health Practice and Policy through Collaborative Research provide alternatives suitable for predicting network
N
k A
l i
i h S
i
i
Network Analysis with Sociometrics
• Strengths
– Unbiased assessment of social influence – Patterns of diffusion and communication – Rich measurement and theory
Li it ti
• Limitations
– Missing nominators compromise accuracy
– Costly assessmentCostly assessment
– Complex coding and analysis – Requires bounded social space
Advancing Health Practice and Policy through Collaborative Research
Network Analysis with Sociometrics
Software
• Stand-alone ProgramsStand alone Programs
– UCINET (http://www.analytictech.com/ucinet/) – Krackplot (Freeware – visualization software)
(http://www andrew cmu edu/user/krack/krackplot sh
(http://www.andrew.cmu.edu/user/krack/krackplot.sh tml) • R (http://www.r-project.org/) – iGraph – snasna • Excel N d XL F htt // d l d l /
Advancing Health Practice and Policy through Collaborative Research
Network Analysis from Informants
• Data Source: “Who hangs out together?”
Informants Compilation
Informant Port Kit Tunner Port Kit Tunner
Informant 1
Port Kit Tunner Port Kit Tunner
Port 1 0 1 .5 Kit 0 5 Kit 0 .5 Informant 2 Port 1 1 • Variations Port 1 1 Kit 1
– Cognitive Social Structures (Krakhardt, 1987) – Social Cognitive Mapping (Cairns et al., 1985)
Informants: Examples
• Cairns, Leung, Buchannan, and Cairns (1995) used social cognitive mapping to study the fluidity, reliability,
and interrelations of social networks of 4th and 7th
graders over a 3-week period.
• Neal (2009) used Cognitive Social Structures to study the influence of centrality and density on relational
aggression in a sample of 3rd through 8th grade
children.
Informants: Software
• Cognitive Social Structures • Cognitive Social Structures – consensus aggregation
across k informant matrices
k j i, ijR
R
ac oss o a a cescan be done in Excel or R.
– See Krackhardt (1987) for specific instructions. • Social Cognitive Mapping
– Contact Man-Chi Leung, Ph.D. at UNC (man-chi.leung@unc.edu ) for a copy of the SCM 4.0 program and manual.
Advancing Health Practice and Policy through Collaborative Research p g
Network Analysis from Informants
• Strengths:
– Provides valid estimates of network ties with comparatively few informants.
– Missing data does not decrease accuracy – Economical to administer
• Limitations
Difficult to assess directed relations – Difficult to assess directed relations
– Requires bounded social space (e.g., classrooms, schools, organizations)
Advancing Health Practice and Policy through Collaborative Research
Ego Network Analysis
Ego Network Analysis
• Theory: Best for unbounded networks where • Theory: Best for unbounded networks where
saturation is not possible • Data Source:
– Prompts for different social functions – Demographics, relationships, frequency – Behavior of network members
• Analysis:
Net ork si e densit di ersit bo ndar densit – Network size, density, diversity, boundary density,
heterogeneity, position, behavior of network members
Ego Networks: Measures
Ego Networks: Measures
• Heterogeneity • Heterogeneity
e A ity Heterogene n k iA 1 2 1 n ity Heterogene iA 1where A = a categorical attribute (e.g., gender, race)
Ak = number of individuals with the attribute
e = number of individuals with valid data on A
n = total number of traits of A in the ego network
Ego Networks: Examples
g
p
• Dominguez & Maya-Lariego, 2008
Ego networks of host individuals and immigrants in – Ego networks of host individuals and immigrants in
the U.S. and Spain
– Host individuals had lower centrality than did y
immigrants according to multiple measures.
T l G S ith & H 2003
• Tolan, Gorman-Smith, & Henry, 2003
– Ego network assessments of delinquent involvement in adolescent males
in adolescent males
– Network violence predicted future individual violence.
E
N
k
Ego Networks
• Strengths
– Does not require assessment of entire network – Can provide social network and social support
i f ti
information
– Does not require bounded social space. • Limitations
• Limitations
– Possible bias in the direction of the individual’s behavior
– Ego is central by definition, so meaning of position and centrality are problematic.
E
N
k
S f
Ego Networks: Software
• Like informant-based network data, ego networks populate matrices and graphs of the type we have been discussing.
• Ego network data can be visualized in Krackplot and • Ego network data can be visualized in Krackplot and
other programs and analyzed using any software program you would use to analyze sociometric data. • Because ego networks tend to be smaller than
networks derived from sociometric studies, analyses can often be conducted by hand or using Excel or
can often be conducted by hand or using Excel or SPSS.
Dynamic Social Network Analysis
Dynamic Social Network Analysis
• Theory
– Social relationships are dynamic – Most SNA is static
– Static analysis may miss important characteristics of the social world.
• Examples • Examples
– Is “liaison” a position or a transition state?
– Changes in parent groups over the course of g p g p
intervention.
Group 212: Pre
p
Group 212: Session 4
p
Group 212: Session 9
p
Group 212: Session 14
p
SAFE-E Group 212
0 8 0.9 1 2 5 3 0 4 0.5 0.6 0.7 0.8 1.5 2 2.5 0 0.1 0.2 0.3 0.4 0 0.5 1 # of contacts DensityDynamic SNA
Dynamic SNA
• Methods – Berger-Wolf method • α (Persistence) • β (turnover) • γ (membership) Software – Software• tnet package in R does analysis of time-stamped
ties - http://toreopsahl.com/tnet/p p
• DNA (Discourse Network Analyzer) http://www.philipleifeld.de/
Summary
Saturation Possible? Absentees or non-participants? Multiple Time Points participants? Points N Y N Y N Y Sociometrics - + + - + -Informants - + + + + -Ego-Networks + ? + - + -Dynamic SNA ? ? ? ? - +