Social Network Analysis “SNA” for Complete Beginners€¦ · • Social Network Analysis – Map of Network Growth • Social Network Analysis of Disaster Response, Recovery,

SOCIAL NETWORK ANALYSIS “SNA” FOR COMPLETE BEGINNERS

CA BJ KIDD, KCPD

10.01.20 - IACA ONLINE CONFERENCE

WE WILL COVER:

1. What is SNA?

2. How is it differentiated from social media and social networking?

3. Is it relevant?

4. Key terms / concepts

5. Beginner’s Orientation to Gephi & Ora

6. Tips & Resources for self-directed learning/ implementation

GOALS: AFTER THIS PRESENTATION, YOU WILL…

• Have a better understanding of SNA

• Be able to determine if/how SNA fits in your analytical toolbox

• Have a clear path forward for starting your first SNA project

• Have resources to consult as you tackle your first project

• Be equipped to get more out of further SNA training

DISCLAIMERS & CREDITS• Qualifications

• Disclaimers• Co-Learner

• Learn by doing

• NPS CORE LAB

HOUSEKEEPING

1. What is SNA?

2. How is it differentiated from social media and social networking?

3. Is it relevant?

4. Key terms / concepts

5. * Beginner’s Orientation to Gephi & Ora

6. Tips & Resources for self-directed learning/ implementation

New Sections

Questions

* This is not an advertisement. Recommendations

1. WHAT IS SNA?

DEFINITIONS

Social Network: “A finite set of actors, or sets of actors, that share ties with one another.”

Social Network Analysis: “Assumes that the behavior of an actor is profoundly affected by its ties to others, and the networks in which they are embedded.” Stanley Wasserman & Katherine Faust

DEFINITIONS

Social Network Analysis: “A collection of theories and methods that assumes that the behavior of actors* is profoundly affected by their ties to others and the networks in which they’re embedded.”

*Individuals, groups, organizations, etc. Sean Everton, Co-Director CORE Lab

DEFINITIONS

“A social network analysis examines the structure of social relationships in a group to uncover the informal connections between people.” (Inside SNA, p. 1)

SNA IN PRACTICE

Empirical dataPrioritizes dynamic relational ties over

static attributes

Relies heavily on graphic visualizations

to both analyze, explore, and convey the analysis result

Is a fluid reiterative process

Can account for the same actor playing

various roles in various network contexts

Can handle large datasets

2. HOW IS SNA DIFFERENTIATED FROM SOCIAL MEDIA AND SOCIAL

NETWORKING?

SOCIAL NETWORKS

DIFFERENTIATING SNA“SNA differs from conventional approaches to business problems in one very important way: SNA assumes that people are all interdependent. This assumption is radically different from traditional research approaches which assume that what people do, think, and feel, is independent of who they know.” (Inside SNA, p. 2)

3. IS SNA RELEVANT?ONE QUESTION & TWO ANSWERS

DOES YOUR AGENCY STRICTLY DEAL WITH CRIMINALS WHO ARE PERPETUALLY SINGLE AND CHILDLESS ORPHANS WHO LIVE ON UNINHABITED ISLANDS?

1. ATTITUDE NETWORKS & COVID-19http://blogs.cornell.edu/info2040/2020/09/28/attitude-networks-and-covid-19/

ATTITUDE NETWORKS & COVID-19http://blogs.cornell.edu/info2040/2020/09/28/attitude-networks-and-covid-19/

2. ANALYZING “MOVEMENT” POTENTIALThe Power of Habit, Part 3: The Habits of Societies, Ch. 8, “Saddleback Church & the Montgomery Bus Boycott: How Movements Happen.” https://blogs.cornell.edu/info2040/2015/09/21/rosa-parks-strong-ties-and-the-power-of-weak-ties/

Booking Photo 02/22/56

“MOVEMENT POTENTIAL”

Congressional Medal of Freedom in Detroit, in a Nov. 28, 1999

Photos featured on CBS News “Rosa Park’s: Civil Rights Trailblazer” - https://www.cbsnews.com/pictures/rosa-parks/

“Rosa Parks and the Montgomery bus boycott became the epicenter of the civil rights campaign not only because of an individual act of defiance, but also because of social patterns.”

ROSA PARKS HAD STRONG TIES TO INDIVIDUALS AND GROUPS THAT NORMALLY DID NOT ASSOCIATE WITH ONE ANOTHER

Montgomery’s Directory of Civil and Social Organizations

Rosa

Individuals Groups

WEAK TIES

• 1960’s Harvard PhD student Mark Granovetter

• Conducted a study on how 282 men received their current employment (at the time)

• Network measurements are often named after the researcher who developed/advanced the measurement in question.

NETWORK LEVEL MEASURES •Network Topography

•Cohesive Subgroups

•Centrality

•Brokers & Bridges

4 Metric Families

4. KEY TERMS & CONCEPTS

TWO OTHER DISTINCTIONS

•Link Charts •Other social science approaches

TWO OTHER DISTINCTIONS - #1

•Link Charts

TWO OTHER DISTINCTIONS - #1 CONTINUED

•VS SNA /

Sociograms

TWO OTHER DISTINCTIONS - #2

•Other social science approachesTraditional approaches focus on the actor attributes*. However, attributes don’t change when the environmental or situational context changes. Nor do they change when the dynamics of a group to which an actor belongs are changing.

Relational ties, however, tend to remain more static across multiple domains.

Social network analysts and practitioners will argue that understanding interaction patterns is more important to understanding both group dynamics, structure, and the decisions individual actors are more likely to make.

*Race/Sex

SOCIAL NETWORK ANALYSIS INVESTIGATES TWO INTERRELATED ASPECTS OF RELATIONAL DATA:

• Nodes / Vertices / Actors• Actors are not necessarily people.

• i.e., NIIBIN hits, IED signatures

• When actor classes involve people, remember that actors have AGENCY

• Ties / Edges • Unit of measurement

• Ties must be comparable

SOCIAL NETWORK ANALYSIS INVESTIGATES TWO INTERRELATED ASPECTS OF RELATIONAL DATA:

• Ties / Edges = the unit of measurement• “Tie Variance”

• Types – 9 basic types as described by Sean F. Everton (Disrupting Dark Networks)

• Strength

• Direction

TIE TYPES• “Tie Variance”

• Types – 9 basic types as described by Sean F. Everton (Disrupting Dark Networks)

1. Resource

2. Association

3. Behavioral (i.e., communication)

4. Geographical

5. Social Mobility (think Rosa Parks)

6. Physical

7. Formal

8. Biological

9. Sentiment

PRO TIP: This is a good list to check when developing/adjusting your codebook.

TIE STRENGTH & THE POWER OF WEAK TIES

The strength of a tie is determined by number, the frequency, and the intensity of the interactions between actors.

Determining the strength/weakness of a tie can be subjective during the data collection and data structuring process.

However, from the structured data imported into your SNA program of choice, the strength/weakness of ties between actors of this will become naturally apparent from your data.

TIE DIRECTION• “Tie Variance”

• Direction

• Undirected – called Edges

• Directional – also cared Arcs

• One directional & bi-directional

METRIC FAMILIES

NETWORK TOPOGRAPHYNetwork Density

COHESIVE SUBGROUPS

CENTRALITY BROKERS & BRIDGES

METRIC FAMILIES > NETWORK TOPOGRAPHY > NETWORK DENSITY

NETWORK TOPOGRAPHYNetwork Density

Total number of ties / potential number of ties on a range of 0 -1

10

5. BEGINNER’S ORIENTATION TO GEPHI & ORA

“CODEBOOKS”

THE SNA PROGRAM YOU USE WILL DETERMINE HOW YOU NEED TO STRUCTURE YOUR DATA

Project Plan

Codebook: Actor Relationships / Actor Classes

Attributes

https://www.nationalpublicsafetypartnership.org/clearinghouse/Content/ResourceDocuments/Naval%20Postgraduate%20School%20CORE%20Lab%20Codebook.pdf

GEPHI

GEPHI https://gephi.org/

ORA

ORA

IMPORTING DATA

IMPORTING DATA

IMPORTING DATA

6. TIPS & RESOURCES FOR SELF-DIRECTED LEARNING

/ IMPLEMENTATION

POSSIBLE SNA PROJECTS

• Terrorist Networks / Recruitment• NIIBIN data• Robbery crew analysis• Domestic Violence / At risk youth• Contact tracing for COVID-19 exposure• Gang networks / OMGs• Organized retail theft networks (or any active theft

ring)• Admin analysis: information flow within a

unit/department• Narcotics distribution

A PATH FORWARD

• Find some network data and tutorial and work through it. Play around a bit and get comfortable with some of the visualizations.

• Let working with sample data inform the questions you ask of your own data / networked criminals.

• Pick a VERY small network to start with. • You can either track that network over time, or

use it as model for larger projects. • A small project will help you work with your

own data structures while minimizing time “wasted” on the learning curve.

• Take a group you may already have a link chart on and conduct social network analysis on the same subjects.

• Get feedback.

CREATING A PROJECT FRAMEWORK

PROJECT PLAN

CODEBOOK ACTORS - RELATIONSHIPS

CODEBOOK ACTORS – RELATIONSHIPS1 MODE & 2 MODE NETWORKS

EDGELIST

NODE ATTRIBUTES

• Applications of Social Media and Social Network Analysis• Getting a Job: A Study of Contacts and Careers by Mark Granovetter, 2nd

edition• Inside Social Network Analysis by Kate Ehrilich and Inga Carboni. Aug. 10,

2010. • Models and Methods in Social Network Analysis. Structural Analysis in the

Social Sciences Series #27, edited by Peter J Carrington, John Scott, and Stanley Wasserman, Cambridge University Press. *Granovetter, General Editor.

• Sentiment Analysis in Social Networks, by Federico Alberto Pozzi, Elsevier Science, 2016. (From more of a computing standpoint. Interdisciplinary.)

• Social Network Analysis: A handbook, 2nd edition, by John Scott• Social Network Analysis – Map of Network Growth• Social Network Analysis of Disaster Response, Recovery, and Adaptation by

Eric C. Jones and A.J. Fass, Elsevier Science, 2016. • The Sage Handbook of Social Network Analysis, by John Scott and Peter J.

Carrington (PDF Format)• *Understanding Criminal Networks: A Research Guide by Prof. Gisela

Bichler. • Use of Social Network Analysis (SNA) to show growth of networks focused on

common purpose, by the Tutor/Mentor Institute, LLC, 2012. • Social Network Analysis of Terrorist Networks, “Pakistani Journal Article”

uploaded by Mark Lauchs on July 9, 2012

Books. Audio books. Podcasts. Articles. Magazines. Presentations. Syllabi.

* This is not an advertisement.

HIGHLYRECOMMENDED: NPS CORE LAB

* This is not an advertisement.

https://nps.edu/

https://nps.edu/web/core

https://github.com/NPSCORELAB

*Also find them on SLACK.

SELF-STUDY THROUGH THE CENTER FOR HOMELAND DEFENSE AND SECURITY (CHDS)

https://www.chds.us/selfstudy/

https://www.chds.us/selfstudy/courses/social-network-analysis/ - Need a Log In

MIT COURSE – BASIC OVERVIEW SLIDEShttps://ocw.mit.edu/courses/sloan-school-of-management/15-599-workshop-in-it-collaborative-innovation-networks-fall-2011/lecture-notes/MIT15_599F11_lec04.pdf

Disrupting Dark Networks. Structural Analysis in the Social Sciences, #34. By Sean F. Everton. Cambridge University Press, 2012. •

https://www.cambridge.org/core/books/disrupting-dark-networks/1F2BFFEA7C036EC7CFD0ED1FFDAE21D7

FINDING DATA SOURCES TO PRACTICE ON

• Network Specific Repository by the CORE LAB - https://core-dna.netlify.app/post/page/2/

• OPEN DATA - https://www.freecodecamp.org/news/https-medium-freecodecamp-org-best-free-open-data-sources-anyone-can-use-a65b514b0f2d/