This 5-day online short course (July 13–17, 2020) introduces the theory and practice of social and ecological network analysis. While standard statistical analyses assume that data on entities (e.g., people, organizations, animals) are independent, network analysis focuses on the relationships among entities to explain emergent properties of complex systems. Network analysis has a long tradition across many disciplines, and this course combines approaches from both the social and natural sciences to inform new strategies for studying socio-environmental systems.
The course starts with an introduction to networks and then covers a variety of established and novel techniques used to analyze social and ecological networks, particularly statistical inference methods. The course includes hands-on coding demonstrations in the R programming language and there is an emphasis on group discussion to help participants develop network-based hypotheses for socio-environmental systems and to provide guidance on testing them effectively. As such, participants are encouraged to bring their own data, and throughout the course, there is the option of meeting one-on-one with an instructor or in small groups with other participants to discuss research and explore overlapping interests.
Participants are expected to be available each day of the course for live virtual classes and meetings from 9 a.m. to 12 p.m. and 2 p.m. to 5 p.m. (U.S. Eastern Time). The registration fee for the course is $50.
This is a foundational course for anyone interested in adding network analysis to their methodological toolkit, regardless of prior experience; however, depending on the pool of applicants, preference may be given to those with prior experience in applications of social network analysis (SNA) or computer programming. Applicants whose research or teaching involves the social or biophysical components of environmental problems will be given preference, but applicants with other areas of interest are also welcome. The course material is intended for students who have basic familiarity of network analysis and want to apply it in their own work, but lack hands-on training. Preliminary readings and educational videos will be made available to participants. The target class size is 12 to 15 participants, so space is limited and competition for places is expected. Some experience with computer programming, ideally with the R language, is expected, as well as an understanding of basic statistics (e.g., probability distributions, hypothesis testing, regression).
- What is network data?
- What are the problems in collecting it?
- What kinds of questions can we use it to answer?
- How is it different from other data?
- Structural & locational properties of actors/locations/resources (centrality, prestige & prominence to determine popular resources, organizations, etc.)
- Structural cohesion (subgroups & cliques)
- Equivalence of actors (structural equivalence & block models to determine niche differentiation or social isomorphism)
- Local analyses (dyadic & triadic analysis, brokerage to determine structural hierarchies and key resources or actors)
- Community detection algorithms
- Matrix permutation tests
- Conditional uniform random graphs
- Network autocorrelation models
- Introduction to statistical global analyses (p1, p*, ERGMs & their relatives)
- Temporal models
Lectures, group discussions, and hands-on coding demonstrations are typically scheduled for the morning (9 a.m. to 12 p.m., U.S. Eastern Time) with one-on-one and small group meetings reserved for the afternoon (2pm to 5pm, US Eastern Time), to facilitate co-ordination across time zones.
- Day 1: Introduction to networks
- Day 2: Describing networks
- Day 3: Network methods and hypothesis testing
- Day 4: Advanced social network analysis
- Day 5: Wrap-up and next steps
Please apply online here.
The University of Maryland is an Equal Opportunity Employer.
Minorities and Women Are Encouraged to Apply.