Skip to main content

About this book

This volume introduces a series of different data-driven computational methods for analyzing group processes through didactic and tutorial-based examples. Group processes are of central importance to many sectors of society, including government, the military, health care, and corporations. Computational methods are better suited to handle (potentially huge) group process data than traditional methodologies because of their more flexible assumptions and capability to handle real-time trace data.
Indeed, the use of methods under the name of computational social science have exploded over the years. However, attention has been focused on original research rather than pedagogy, leaving those interested in obtaining computational skills lacking a much needed resource. Although the methods here can be applied to wider areas of social science, they are specifically tailored to group process research.
A number of data-driven methods adapted to group process research are demonstrated in this current volume. These include text mining, relational event modeling, social simulation, machine learning, social sequence analysis, and response surface analysis. In order to take advantage of these new opportunities, this book provides clear examples (e.g., providing code) of group processes in various contexts, setting guidelines and best practices for future work to build upon.
This volume will be of great benefit to those willing to learn computational methods. These include academics like graduate students and faculty, multidisciplinary professionals and researchers working on organization and management science, and consultants for various types of organizations and groups.

Table of Contents


Chapter 1. Introduction

Computational social science has emerged as a response to bigger, complex, and newer types of data available. Although a growing body of research has taken advantage of this new approach, there have been few outlets available that pedagogically give researchers a resource to learn these types of methods. The purpose of this book is to provide such a resource for group researchers seeking to exploit methods falling under the domain of computational social science. Methods in the volume include response surface analysis, machine learning, relational event modeling, semantic network analysis, social sequence analysis, and simulation techniques.
Andrew Pilny, Marshall Scott Poole

Chapter 2. Response Surface Models to Analyze Nonlinear Group Phenomena

Because of the large amount of group data available today (e.g., online groups, archival records, electronic emails, etc.), response surface methodology (RSM), mostly common in the natural and physical sciences, can be a useful tool to analyze nonlinear group phenomena. RSM typically requires multiple observations on each level of each independent variable, which has mainly been the reason for its nonexistent use in experimental group research. However, the main goal of RSM is optimization, a very common line of inquiry in functional group research. In other words, RSM attempts to locate the precise values of independent variables that will predict an optimal (or minimal) response in the dependent variable. RSM does this by fitting a quadratic and interaction-term regression model and uses a canonical analysis to find a solution to the response surface (i.e., the shape of the predicted response). The following chapter describes the basic mathematical logic underlying RSM and provides an example using groups in the online video game EverQuestII. The code using the PROC RSREG feature in SAS is used in the tutorial.
Andrew Pilny, Amanda R. Slone

Chapter 3. Causal Inference Using Bayesian Networks

The availability of new computational technologies, data collection opportunities, and data size has significantly influenced the development of a data driven social scientific approach, popularly known as “computational social science”. Although a slowly growing field within social sciences and largely spearheaded by interdisciplinary scientific teams, computational social science is profoundly changing the nature social scientific analysis. Social scientists are now able to generate predictive results beyond traditional social scientific methods, thus are able to increase the power of social analysis. Machine learning classification algorithms (MLCAs) are a group of techniques utilized by present day computational social scientists. MLCAs became more user friendly for social scientists for the availability of Analytical Graphical User Interfaces (GUI). This article demonstrated the opportunities offered by data driven social scientific approach. Based on an actual research, this article explored a situation where a number of people embedded in teams were working together in a complex environment. This article first, explains the basic ideas of Bayesian Network Analysis – one set of MLCAs. Then it provides step-by-step demonstration on conducting Bayesian Network Analysis using Weka. Finally, the article demonstrates ways of interpreting and presenting social scientific results.
Iftekhar Ahmed, Jeffrey Proulx, Andrew Pilny

Chapter 4. A Relational Event Approach to Modeling Behavioral Dynamics

This chapter provides an introduction to the analysis of relational event data (i.e., actions, interactions, or other events involving multiple actors that occur over time) within the R/statnet platform. We begin by reviewing the basics of relational event modeling, with an emphasis on models with piecewise constant hazards. We then discuss estimation for dyadic and more general relational event models using the relevent package, with an emphasis on hands-on applications of the methods and interpretation of results. Statnet is a collection of packages for the R statistical computing system that supports the representation, manipulation, visualization, modeling, simulation, and analysis of relational data. Statnet packages are contributed by a team of volunteer developers, and are made freely available under the GNU Public License. These packages are written for the R statistical computing environment, and can be used with any computing platform that supports R (including Windows, Linux, and Mac).
Carter T. Butts, Christopher Steven Marcum

Chapter 5. Text Mining Tutorial

A growing challenge facing scholars who study group processes is textual data overload. The immense amount of text generated by group members’ interactions via email, text messages, and social media can be a barrier during data collection and analysis. Instead of scaling back textual data collection, group process scholars can make use of text mining, a computational approach to finding patterns within and extracting information of interest from textual datasets. This tutorial provides an entry-level introduction to the text mining approach in terms of how it works, its underlying assumptions, the basic steps of analysis, and decisions that must be made during the text mining process from data collection to final interpretation. The approach is demonstrated using a real-world dataset consisting of transcriptions of medical consultation conversations among groups of emergency department physicians. The results demonstrate the potential benefits of a data-driven approach to analysis of textual datasets.
Natalie J. Lambert

Chapter 6. Sequential Synchronization Analysis

One of the oldest questions in group research is “What makes a group more than just a collection of individuals?” This chapter posits that as group members interact, their activities can become socially entrained, constituting the group as an entity beyond the individual members. Capturing social entrainment provides a unique marker on when and how unique properties emerge at the group level. Sequential synchronization analysis is a method for assessing the type and degree of entrainment in groups and teams based on member communication and behavior. It first defines meaningful sequences of actions for each team member and then analyzes how those sequences are synchronized over time. The chapter provides a step-by-step guide on the new approach and an example.
Toshio Murase, Marshall Scott Poole, Raquel Asencio, Joseph McDonald

Chapter 7. Group Analysis Using Machine Learning Techniques

Analysis of performance of groups or teams is of a primary importance in field of social group studies. In this article we are targeting group performance analysis using computational techniques from machine learning. In order to understand the feature space, we make use of a combination of machine learning methods: decision trees, feature selection as well as correlation analysis. These models are chosen for their easy interpretability. Alongside we also propose methodology to build group level metrics from individual level data. This helps us interpret the feature space at group level and understand how things like attribute variety among group members affects performance. We propose a full methodology that employs machine learning models taking various group level metrics as input, finally providing a thorough analysis of the feature space. In this research we employ the NATO dataset collected using the game-based test-bed called SABRE. We give a hands-on experience by performing a four phase exhaustive group analysis on the SABRE dataset using Weka software, which is a user friendly GUI based machine learning tool.
Ankit Sharma, Jaideep Srivastava

Chapter 8. Simulation and Virtual Experimentation: Grounding with Empirical Data

Some examinations of group processes may require intensive data collection to answer initial research questions. Findings, however, often lead to more research questions that could be answered if additional data were available. In cases where collecting additional data is cost prohibitive, researchers may benefit from formulating “what-if” questions that can be answered via simulation and virtual experimentation. This chapter presents a step-by-step guide to demonstrate (1) how simulation procedures can be developed and validated with existing empirical data and (2) how these procedures can be executed to conduct virtual experiments. To demonstrate these steps, we provide a tutorial based on potential what-if questions about two different aspects of the relationship between team cohesion and team effectiveness using continuous and discrete empirical data, respectively, along with Matlab code for the simulation, validation, and virtual experimentation. We then present two more complex examples from our own published papers.
Deanna Kennedy, Sara McComb
Additional information

Premium Partner

    Image Credits