People are social entities. They build complex relationships with others around them, form communities and social circles, belong to organizations. Every decision to make a connection to others is based a large variety of variables (called attributes). Every connection, in turn, affects people’s attitudes, behavior, and actions. This relationship between the structure of people’s connections to others and everything that this structure affects is called social dynamics.
Social dynamics is the focus of social network analysis. In this course, we will introduce this exciting field, starting with the very basics – the definitions of network concepts. You will quickly learn that network analysis allows to answer questions and find insights not available with any other approaches.
In business, where relationships are essential to efficiency and effectiveness of an organization, it is crucial that analysts know how to analyze these relationships. Therefore, we will not only show you the network concepts, but apply them immediately to real-life business datasets.
The possibilities of network analysis are quite broad. In this course, we divide the complex field according to the three major theoretical concepts in social relations: social selection, social influence, and community building. Models of social influence help explain why networks can affect individual behavior. Models of social selection help us understand how people create their network. Community detection models allow us to find the communities that people build, to better understand the structure of such communities.
Taken together with network statistics, these models are being demonstrated on real-life datasets collected in real companies. Learners can immediately see how much more powerful relational analysis (networks) are relative to standard statistics alone.
They are designed to illustrate some of the specific state-of-the-art approaches within the broader areas.
This Course is part of HSE University Master of Data and Network Analytics degree program. Learn more about admission into the program and how your Coursera work can be leveraged if accepted into the program here https://inlnk.ru/WMKM6.
What are Networks?
The first lecture is designed to familiarize the learners with the idea of networks. First, networks are highly visual, so the lecture introduces the network graphs. Then, we talk about what is social network analysis (SNA), the role of networks in our lives, and applications of SNA in a variety of settings. Finally, we talk about network theory in organizations and application of SNA in organization. A session in R is dedicated to demonstrating networks.
Network Analysis as a Method
This lecture introduces all the foundational network concepts. We start with important terminology, then move to network study design, data collection and descriptive statistics. Then, we examine everything learned in the lecture on a real-life dataset. R session has several segments: loading and manipulating network data, drawing graphs using different packages, interpreting graphs.
Foundational Network Measures
This lecture starts with analyzing networks on the very basic units: dyads and triads. We learn how to interpret triadic census. Then, we move on to one of the most important concepts in network analysis: centrality. We explore these foundational network measures on a real-life dataset. R sessions are dedicated to local analyses (dyads, triads and other measures) and calculating centrality measures.
Social Influence Models
In this lecture, we introduce the idea of modeling on networks. The idea is simple in principle: we use the network measures, which we’ve learned in previous lectures, as predictors in regular statistical models, such as regression. First, we discuss the theories of social influence. Then, we discuss how social influence models are built. We discuss best practices in social influence network models and apply them to a real-life dataset in an R session.
Social Selection Modeling
This lecture takes network analysis to a next level – the models of social selection. We start by talking about the very idea of statistical network models – why do we need them? Then, we talk about social processes and the theory behind network formation. Next we discuss the role of random graphs in the analytic process, which leads us to exponential random graph models (or the models of social selection). We discuss how to build ERGMs and apply this knowledge to a real-life dataset.
Community Detection Approaches
One of the most important application of network analysis is community detection. We start by talking about communities: what are they? Then, we discuss various approaches to community detection and look at a network-level method: blockmodels. We discuss theory in blockmodeling, roles and positions, and learn how to build blockmodels in R on a real-life dataset.