Using the Social Network Data From Add Health

1
Using the Social Network Data From Add Health
James Moody
2001 Add Health Users Workshop August 9
10 Bethesda Maryland
2

• Introduction What and Why
• Levels of Network Data
• Composition Pattern
• Networks on both sides of the equation
• Network Data structures
• Adjacency Matricies
• Adjacency Lists
• Network Analysis Programs
• Network Data in Add Health
• In School Friendship Nominations
• In Home Friendship Nominations
• Constructing Networks
• Total Networks
• Local Networks
• Peer Groups

3
Levels of Network Data
Best Friends
ego
ego
Local Network
Peer Group
4
(No Transcript)
5
(No Transcript)
6
Measuring Network Context Patterns

• Pattern measures capture some feature of the
  distribution of relations across nodes in the
  network. These include
• Density of all possible ties actually made
• Reciprocity likelihood that given a tie from i
  to j there will also be a tie from j to i.
• Transitivity extent to which friends of friends
  are aslo friends
• Hierarchy Is there a status order to
  nominations? How is it patterned?
• Clustering Are there significant groups? How
  so?
• Segregation Do attributes (such as race) and
  nominations corespond?
• Distance How many steps separate the average
  pair of persons in the school? Is this larger or
  smaller than expected?
• Block models What is the implied role strucutre
  underlying patterns of relations?
• These features (usually) require having
  nomination data from each person in the network.

7
Measuring Network Context Composition

• Compostion measures capture characteristics of
  the population of people within a given network
  level. These include
• Heterogeneity How dispersed are actors with
  respect to a given attribute?
• Means What is the mean GPA of egos friends? How
  likely is it that most of egos friends will go
  to college?
• Dispersion What is the age-range of people ego
  hangs out with?
• These features can often be measured from the
  simple ego network.

8
Analysis with Social Network data

• Networks as Dependant Variables
• Interest is in explaining the observed patterns
  of relations.
• Examples
• Why are some schools segregated and others not?
• What accounts for differences in hierarchy across
  schools?
• What accounts for homophily in friendship choice?
• Tools
• Descriptive tools to capture properties
• Standard analysis tools at the level of networks
  to explain the measures
• p and other specialized network statistical and
  simulation models

9
Analysis with Social Network data

• Networks as independent Variables
• Interest is in explaining behavior with network
  context (Peer influence/ context models)
• Examples
• Is egos probability of smoking related to the
  smoking levels of those he/she hangs out with?
  (compositional context)
• Is the transition to first intercourse affected
  by the peer context?
• Are isolated students more likely to cary weapons
  to school than those in dense peer groups?
  (positional context)
• Tools
• Depends on dependant variable
• Peer influence models
• Dyad models
• Contextual models, with network level as nested
  context (students within peer groups)

10
Network Data Structures
Adjacency Matrix
Graph
Arc List
Node List
11
Network Analysis Programs

• 1) UCI-NET
• Genearl Network analysis program, runs in Windows
• Good for computing measures of network topography
  for single nets
• Input-Output of data is a little clunky, but
  workable.
• Not optimal for large networks
• Availiable from
• Analytic Technologies
• Borgatti_at_mediaone.net
• 2) STRUCTURE
• A General Purpose Network Analysis Program
  providing Sociometric Indices, Cliques,
  Structural and Role Equivalence, Density Tables,
  Contagion, Autonomy, Power and Equilibria In
  Multiple Network Systems.
• DOS Interface w. somewhat awkward syntax
• Great for role and structural equivalance models
• Manual is a very nice, substantive, introduction
  to network methods
• Availiable from a link at the INSNA web site
• http//www.heinz.cmu.edu/project/INSNA/soft_inf.ht
  ml

12
Network Analysis Programs

• 3) NEGOPY
• Program designed to identify cohesive sub-groups
  in a network, based on the relative density of
  ties.
• DOS based program, need to have data in arc-list
  format
• Moving the results back into an analysis program
  is difficult.
• Availiable from
• William D. Richards
• http//www.sfu.ca/richards/Pages/negopy.htm
• 4) PAJEK
• Program for anlayzing and plotting very large
  networks
• Intuitive windows interface
• Used for all of the real data plots in this
  presentation
• Mainly a graphics program, but is expanding the
  analytic capabilities
• Free
• Availiable from

13
Network Analysis Programs

• 5) SPAN - Sas Programs for Analyzing Networks
  (Moody, ongoing)
• is a collection of IML and Macro programs that
  allow one to
• a) create network data structures from the Add
  Health nominations
• b) import/export data to/from the other network
  programs
• c) calculate measures of network pattern and
  composition
• d) analyze network models
• Allows one to work with multiple, large networks
• Easy to move from creating measures to analysing
  data
• All of the Add Health data are already in SAS
• Availiable by sending an email to
• Moody.77_at_osu.edu

14
Network Data Collected in Add Health
In -School Network Data

• Complete Network Data collected in every school
• Each student was asked to name up to 5 male and 5
  female friends
• These data provide the basic information needed
  to construct network context measures.
• Due to response rates, we computed data on 129 of
  the 144 total schools.
• Variable is named MFltgtAID form male friend,
  FFltgtAID for female friends.

15
Network Data Collected in Add Health
In -School Network Data

• Nomination Categories
• Matchable People Inside Egos School or Sister
  School
• People who were present that day
• ID starting with 9 and are in the sample
• People who were absent that day
• ID starting with 9, but not in the school sample
• People in egos school, but not on the directory
• Nomination appears as 99999999
• People in egos sister school, but not on the
  director
• Nomination appears as 88888888
• People not in egos school or the sister school
• Nomination appears as 77777777
• Other Special Codes
• Nominations Appears as 99959995
• Nominator Categories
• Matchable Nominator
• Person who was on the roster, ID starts is 9.
• Unmatchable Nominator

16
Network Data Collected in Add Health
In -School Network Data
17
Network Data Collected in Add Health
In -School Network Data
Example 1. Ego is a matchable person in the
School
Out
Un
Out
Out
Un
Un
M
Ego
M
Ego
M
M
M
M
M
M
True Network
Observed Network
18
Network Data Collected in Add Health
In -School Network Data
Example 2. Ego is not on the school roster
M
M
M
Un
M
Un
M
M
M
M
M
M
Un
Un
Un
True Network
Observed Network
19
Network Data Collected in Add Health
In -School Network Data
20
Network Data Collected in Add Health
In -School Network Data
21
Network Data Collected in Add Health
In -Home Network Data

• Network Data were collected in both Wave1 and
  Wave 2 Surveys
• There were two procedures
• Saturated Settings
• Attempted to survey every student from the
  In-School sample.
• 2 large schools, and 10 small schools.
• Was supposed to replicate the in-school design
  exactly.
• Unsaturated Settings
• Each person was only asked to name one other
  person
• In both cases, the design was not always carried
  out. As such, some of the students in the
  saturated settings were alowed to name only one
  male and one female friend, while some students
  who were in the non-saturated settings were asked
  to nominate a full slate of 5 and 5.

22
Network Data Collected in Add Health
In -Home Network Data

• Data Usage Notes
• Romantic Relation Overlap
• For the W1 and W2 friendship data, any friendship
  that was also a romantic relation was recoded to
  55555555, to protect the romantic relation
  nominations.
• Bad Machine on Wave 2 Data
• Data on from one school in wave 2 seems to be
  corrupted. We have no way to show this for
  certain, but it seems to be the case that data
  from machines 200065 or 200106 gave incorrect
  data. We suspect this is so, because almost
  everyone who used these two machines nominated
  the same person multiple times. This results in
  one person having an abnormally large in-degree.
• All nomination s are now valid
• Unlike the in-school data, Ids starting with
  something other than 9 can be nominated.
• Same out-of-sample special codes
• All other special codes for these data are the
  same as in the in-school data.

23
Network Data Collected in Add Health
In -Home Network Data
Descriptive Statistics for Saturated Settings
24
Constructing Network Measures
Total Network
To construct the social network from the
nomination data, we need to integrate each
persons nominations with every other nomination.
Methods 1) Export the Nomination data to
construct network in other program MOST of the
other programs require you to pre-process the
data a great deal before they can read them. As
such, it is usually easier to create the files in
SAS first, then bring them into UCINET or some
such program. 2) Construct the network in
SAS The best way to do this is to combine IML
and the MACRO language. SAS IML lets you work
with matricies in a (fairly) strait forward
language, the SAS MACRO language makes it easy to
work with all of the schools at once. Programs
already set up to do this are availiabel in
SPAN.
25
Constructing Network Measures
Adjacency Matricies
The key to analyzing / measuring the total
network is constructing either an adjacency
matrix or an adjacency list. These data
structures allow you to directly identify both
the people ego nominates and the people that
nominate ego. Thus, the first step in any
network analysis will be to construct the
adjacency matrix.
To do this you need to 1) Identify the universe
of possible people in the network. This is
usually the same as the set of people that
you have sampled. However, if you want to
include ties to non-sampled people you may make
the universe include all people named by
anyone. 2) create a blank matrix with n rows and
n columns. 3) loop over all respondents, placing
a value in the column that corresponds to the
persons they nominate. This can be binary (named
or not) or valued (number of activities they do
with alter).
26
Constructing Network Measures
Total Network
Data for 12th grade males in a small school.
not in 12 grade male sub-sample
27
Constructing Network Measures
Total Network
Program for creating a network and exporting it
to PAJEK
0 proc iml 1 include 'c\moody\sas\programs\modu
les\adj.mod' 2 include 'c\moody\sas\programs\mo
dules\pajwrite.mod' 3 include
'c\moody\sas\programs\modules\pajpart.mod' 4
use work.d 5 read all varaidr into id 6
read all varmf1aid mf2aid mf3aid mf4aid mf5aid
into noms 7 adjmatadj(id,noms) / adj() is
a pre-programed module / 8 adj_idadjmat,1 9
insampj(nrow(adj_id),1,0) / identify people
who are also in the sub-sample / 10 do
i1 to nrow(insamp) 11 ilocloc(idadj_idi)
12 if type(iloc)'N' then do 13
insampi1 14 end 15 free iloc 16
end 17 adjmatadjmat,2ncol(adjmat) 18
file 'c\moody\conferences\add_health\ptp15_paj.ne
t' 19 call pajwrite(adjmat,adj_id,2) 20
file 'c\moody\conferences\add_health\ptp15_paj.cl
u' 21 call pajpart(insamp) 22 quit
28
Constructing Network Measures
Resulting network as displayed by PAJEK.
Total Network
Senior Male subsample in Red
29
Constructing Network Measures
Local Networks.

• To create and calculate measures based only on
  the people ego nominates, you can work directly
  from the nomination list (dont need to construct
  the adjacency matrix).
• To create and calculate measures based on the
  received or reciprocated ties, you need to have a
  list of people who nominate ego, which is easiest
  to get given the adjacency matrix.
• To calculate positional measures (density,
  reciprocity, etc.) all you need is the nomination
  data.
• To calculate compositional data, you need both
  the nomination data and matching attribute data.

30
Constructing Network Measures
Local Networks.
31
Constructing Network Measures
An example network All senior males from a small
(n350) public HS.
Local Networks.
Adjacency Matrix
32
Constructing Network Measures
Local Networks.
Example 2 Suppose you want to identify egos
friends, calculate what proportion of egos
female friends are older than ego, and how many
male friends they have (this example came up in a
model of fertility behavior).

• You need to
• Construct a dataset with
• (a) ego's id (aid1 - make it a number instead of
  a character),
• (b) age of each person,
• (c) the friendship nominations variables.
• Write a macro that loops over each
  community/School
• For each community, do
• a) Identify ego's friends
• b) Identify their age
• c) compare it to ego's age
• d) count it if it is greater than ego's.
• An example SAS program to do this is in the
  handouts.

33
Constructing Network Measures
Peer Groups.
Identifying cohesive peer groups requires first
specifying what a cohesive peer group is.
Potential defintions could be a) all people
within k steps of ego (extended ego-network) b)
a set of people who interact with each other
often (relative density) c) a set of people with
a particular pattern of ties (a closed loop, for
example) UCINET, STRUCTURE, NEGOPY and SPAN all
provide methods for identifying cohesive groups.
They all differ on the underlying definition of
what constitutes a group. The FACTIONS
algorithm in UCINET and NEGOPYs algorithm use
relative density. The CROWD algorithm is SPAN
uses a combination of relative density and
pattern. Once you have constructed the
adjacency matrix, you can export to these other
programs fairly easily. However, most of them
are QUITE time consuming (FACTIONS, for example,
is a bear) and take a good deal of time to run,
so be sure you have identified exactly what you
want before you start processing.
34
Constructing Network Measures
Peer Groups Characteristics.
Identifying Cohesive Sub-Groups

• Cohesion The group is difficult to separate the
  connection of the group does not depend on one
  relation or person.
• Groupness Relative to the rest of the network,
  a cohesive sub - group has high relational
  volume.
• Inclusion Some people are not in groups while
  others bridge groups.

35
Examples of Peer groups within Add Health High
Schools Crowds Algorithm
36
Observed Clustering within Adolescent Social
Networks
Network Characteristics of Sub Groups

• On average, 65 of a schools adolescents are in
  cohesive sub-groups.
• 87 of all relations are within sub-groups.
• The average sub-group has 22 members.
• The average diameter for a sub-group is 3 steps.
• The mean segregation index is .96 (1Complete,
  0Random)

37
Observed Clustering within Adolescent Social
Networks
Distribution of Characteristic within groups,
relative to school distribution
38
Constructing Network Data School Level
39
Constructing Network Data School Level
Inter-Group Relations
40
Analysis Using Network Data Nets as Dependent
Variable Racial Segregation
41
Analysis Using Network Data Nets as Dependent
Variable Modeling the network
Network Model Coefficients, In school Networks
42
Analysis Using Network Data Nets as Independent
Variable Suicide
Relational Structures and Forms of Suicide
Regulation
Low
High
High
Anomic
Altruistic
Integration
Low
Egotistic
Fatalistic
43
Analysis Using Network Data Nets as Independent
Variable Suicide
Measuring Isolation and Anomie.
44
Analysis Using Network Data Nets as Independent
Variable Suicide
45
Analysis Using Network Data Nets as Independent
Variable Weapons
Probability of Carrying a Weapon by Race and
Gender
0.14
0.12
0.1
Probability of carrying a weapon
0.08
Males
Females
0.06
0.04
0.02
0
White
Black
Hispanic
Asian
Native American
Other
Race/Ethnicity
a) Figure represents predicted probabilities
model 6 of table 5, holding all other variables
at the full sample mean.
46
Analysis Using Network Data Nets as Independent
Variable Weapons
Network Effects on Weapon Carrying
0.18
0.16
Peer Group Deviance
0.14
0.12
0.1
Probability of carrying a weapon to school
Social Outsiders
0.08
0.06
School Oriented Peer Group
0.04
0.02
0
Positive
0.08
0.19
0.3
0.41
0.52
0.63
0.74
0.85
Negative
0
1
2
3
4
5
6
7
character of peer context
a) Figure represents predicted probabilities for
an average white male student based on models 2
and 3 in table 6. b) Outsiders do not vary on
peer context, they are presented here as a
continuous line simply for comparison with
otherwise similar adolescents
47
Analysis Using Network Data Nets as Independent
Variable Sexual Debut
48
Analysis Using Network Data Nets as Independent
Variable Pregnancy
49
Analysis Using Network Data Nets as Independent
Variable Delinquency
Haynie (2001) AJS 1061013-57
50
Analysis Using Network Data Nets as Independent
Variable Delinquency
Susanne Bunn, Adolescent Substance Use The
Interactive Influence of Parents and Peers
(MAOSU)