BIS 447: Topics in quantitative inquiry

Network analysis & visualization

Joseph J. Ferrare, Ph.D., Assistant Professor, School of Interdisciplinary Arts & Sciences, University of Washington Bothell

OVERVIEW

At the most general level, network analysis is a set of research methods and theories for studying the structure, process, and meaning of relationships. Network analysis has seen wide application across the disciplinary spectrum, including the social and natural sciences, humanities, public health, education, business, and communications. In this course, we will pay close attention to the use of these quantitative techniques in an interdisciplinary context, and thus students should expect to leave this course with a strong foundation of network analysis applicable to a range of substantive problems. 

Throughout the quarter, you will be trained to think like a network analyst who is prepared to identify when certain questions are “network questions,” and to understand how to use the appropriate theories and associated analytic models to address (some of) those questions. For example, how vaccine information spreads through a community is a network question that can be addressed through diffusion models. Or, how one’s friendship ties influence job acquisition is a network question that can be investigated with a social influence model. At the same time, we have to resist the temptation to see everything as a network question. Some questions just do not lend themselves to network thinking. 

Throughout the quarter, we will cover techniques related to network research design, data management/cleaning, software and coding, analysis, visualization, and interpretation. While this is primarily a course in quantitative analysis, we will nevertheless spend some time connecting these methods to relevant theory and engaging with the empirical literature to better understand how researchers use these methods in a variety of disciplinary contexts (e.g., education, health, ecology, business). 

The design of this course assumes no prior knowledge of network analysis. However, it is assumed that students have at least a basic understanding of probability and statistics (BIS 215 or equivalent). 

OBJECTIVES

Students who successfully complete the requirements of this course should expect the following:

  1. To be able to apply and interpret basic and advanced techniques of network analysis to the study of networks across a range of disciplinary contexts

  2. To become competent users of R programming language and packages commonly used in network analysis

  3. To be able to evaluate and identify when research questions are well aligned to the use of network analysis

FORMAT

This course will be offered in a hybrid format involving a mixture of online and in-person activities. A typical week will involve readings and watching video lectures outside of class and in-person meetings in the computer lab. The lectures and readings are designed to provide an introduction to key concepts in network analysis and theory, and to offer examples of these concepts as they apply to the research literature across a variety of disciplines. Our work in the computer lab, meanwhile, will give you opportunities to practice using these techniques with different data sets.

SOFTWARE

This course makes use of R (specifically RStudio) software for statistical computing. We will use a variety of packages, such as igraph, statnet, mass, and others. We will spend considerable time in class getting familiar with the Rstudio environment and the programming language used to make the program do what we want – which is to analyze and visualize networks. With that said, you will need to invest time outside of class to become proficient with R. Fortunately, there are a plethora of resources available for you to practice R, including LinkedIn Learning (available to all UWB students). 

topics covered

  • BASIC CONCEPTS AND MATHEMATICAL FOUNDATIONS IN NETWORK ANALYSIS

  • SOFTWARE: R AND NETWORK-ASSOCIATED PACKAGES

  • NETWORK RESEARCH DESIGN

  • NETWORK VISUALIZATION

  • MEASURES TO DESCRIBE WHOLE NETWORKS

  • MEASURES OF CENTRALITY AND POSITION

  • ANALYZING SUBGROUPS AND COMMUNITY DETECTION

  • TESTING NETWORK HYPOTHESES

  • NETWORK MODELS

  • TWO-MODE NETWORKS

  • SCALING AND CLUSTERING WITH NETWORK DATA