Stevens Institute Series on Complex Systems and Enterprises
Series Editor: William B. Rouse
Universities as Complex Enterprises: How Academia Works, Why It Works These Ways, and Where the University Enterprise Is Headed • William B. Rouse
Modeling Human‐System Interaction: Philosophical and Methodological Considerations, with Examples • Thomas B. Sheridan
Emergent Behavior in Complex Systems Engineering: A Modeling and Simulation Approach • Saurabh Mittal, Saikou Diallo, and Andreas Tolk
Social-Behavioral Modeling for Complex Systems • Paul K. Davis, Angela O'Mahony, and Jonathan Pfautz
Complexity Challenges in Cyber Physical Systems: Using Modeling and Simulation (M&S) to Support Intelligence, Adaptation and Autonomy • Saurabh Mittal and Andreas Tolk
Healthcare System Engineering Access: Measurement, Inference, and Intervention • Nicoleta Serban
Edited by
This edition first published 2019
© 2019 John Wiley & Sons, Inc.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.
The right of Dr. Paul K.Davis, Dr. Angela O'Mahony and Dr. Jonathan Pfautz are to be identified as the editors of the editorial material in this work has been asserted in accordance with law.
Registered Office
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
Editorial Office
111 River Street, Hoboken, NJ 07030, USA
For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.
Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.
Limit of Liability/Disclaimer of Warranty
While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.
Library of Congress Cataloging‐in‐Publication Data
Names: Davis, Paul K., 1943- editor. | O'Mahony, Angela, editor. | Pfautz, Jonathan, editor.
Title: Social-behavioral modeling for complex systems / edited by Dr. Paul K. Davis, Dr. Angela O'Mahony, Dr. Jonathan Pfautz.
Description: Hoboken, NJ, USA : Wiley, [2019] | Includes bibliographical references and index. |
Identifiers: LCCN 2018046976 (print) | LCCN 2018047389 (ebook) | ISBN 9781119484981 (Adobe PDF) | ISBN 9781119484974 (ePub) | ISBN 9781119484967 (hardcover)
Subjects: LCSH: Social psychology–Data processing. | Collective behavior–Simulation methods. | System analysis.
Classification: LCC HA29 (ebook) | LCC HA29 .S6736 2019 (print) | DDC 302/.011–dc23
LC record available at https://lccn.loc.gov/2018046976
Cover Design: Wiley
Cover Image: © agsandrew/Shutterstock
Trying to understand human behavior has probably been a human passion since the first cavewoman tried to figure out why her mate could never remember to wipe his feet before coming inside. We're still at it, working with ever more sophisticated approaches and for increasingly important outcomes.
My own experience in this area started out when I was working at NASA to help design an autopilot for the Space Shuttle vehicle. It worked great when there was no astronaut in the loop, but when they started wrangling the controls themselves, the astronauts were not exactly enamored with my design and the way it behaved. Dismay would be a word that might apply here, but I went back to school to figure out why and soon became enmeshed in trying to understand human self‐motion perception and control, by bringing together established theory, controlled experimentation, and computational modeling. It was an eye‐opener on how complicated even the simplest of human behaviors could be, as well as the beginning of a long foray into developing and using computational models in this arena.
Of course, as I later discovered, there's a long history of this, perhaps going as far back as the 1880s with Ernst Mach and his pioneering work in visual and vestibular psychophysics and certainly at least the 1940s to Norbert Weiner with his introduction of cybernetics and the mathematical modeling of both humans and machines. Since then, many other disciplines have contributed and elaborated on this idea, from the basic sciences of neurophysiology and cognitive science to the more applied efforts in human systems engineering and robotics. Moving into the domain of computational representation has forced many of us to sharpen our approach to describing behaviors, formalizing our theories, and validating them against real data. As one of my mentors once told me (and I paraphrase a bit here): “The rubber hits the road when you start hacking code.”
This movement by the research community has been documented in a number of efforts. In 1998, the National Research Council (NRC) published a review of potential models that might be usefully embedded in existing military simulations to provide greater realism by including the human element (National Research Council 1998). The report concluded that there was no single framework or architecture that could meet all the simulation needs of the services, but it did provide an extensive review of computational behavioral models, in‐depth discussions of different functional areas (e.g. attention, memory, learning, etc.), and considerations for small unit representation (i.e. groups of individuals). A follow‐on NRC study in 2008 provided a somewhat broader review, covering models of not only individuals but also organizations and societies (National Research Council 2008). This study also discussed different categories of models – both formal and informal – and common challenges across the community (e.g. interoperability, inconsistent frameworks, verification, etc.). And, like almost all NRC reports of this ilk, there were a number of recommendations proposed, in this case, covering areas from basic theory development to data collection methods and tools for model building.
Although many insights from these and other studies remain relevant, much has happened in the last decade, in terms of new basic research results, new applications afforded by the acceleration of technology (particularly in sensing, networking, computation, and memory), and, not least, a resurgence of a general interest in natural and artificial cognition, with the recent reemergence of artificial intelligence and machine learning. For example, on the basic research side, a revolution in neuroimaging methods is linking the underpinning of human thinking – across individual and societal levels. On the applied side, masses of data on human behavior are now being collected to describe and predict activity in a huge variety of applications spanning everyday consumer devices, socio‐commercial networks, and population monitoring systems installed by local and national governments, to name a few. Online populations can now support crowd‐sourced and A/B experiments that drive how corporations interact with their customers and governments with their citizens.
A reexamination of the issues addressed by the earlier studies is clearly called for, in light of what's happened over the last decade. This volume does just that and is a particularly welcome addition to the research community. It is structured to address issues of science, modeling, and relationships among theory, modeling, empirical research, and computational social science. It candidly emphasizes past shortcomings in these relationships and current progress that's been made in improving those relationships. One can sense a good deal of excitement among the contributors – both established researchers in their chosen fields and fresh PhDs – because so much is happening on so many fronts, including theory development, data collection, and computational methods of inquiry. It's also a delight to see chapters bringing to bear new insights from the study of nonhuman social systems, neuropsychology, psychology, and anthropology, among other disciplines. And, having spent much of my career concerned about real‐world problems needing insights from the behavioral sciences, I was pleased to see several chapters addressing the frontier problem of how to develop and use social‐behavioral modeling to assist human decision‐making regarding profound social issues, some complicated by equally profound issues of privacy and ethics.
This book represents a carefully curated set of contributions that aim to inspire the current and next generation of researchers – and to encourage how the act of challenging current conceptual boundaries is needed to advance science. Social‐behavioral modeling will continue to be beset by uncertainties because the social‐behavioral phenomena occur in complex adaptive systems into which we have imperfect and sometimes contradictory insight. Nonetheless, such modeling – if undertaken appropriately – can help humans who must plan, operate, and adapt in their complex worlds. One of the editors' themes is most welcome: complexity need not be paralyzing, especially if we take a multidisciplinary use‐centric approach to working on real‐world problems.
All in all, this volume is a welcome contribution that will be valuable to diverse audiences in schoolhouses, research laboratories, and the workplace. It is a collection, not a textbook or single‐author monograph, and it conveys an excellent sense of the current state of the art and the exciting opportunities that are now being exploited. The editors deserve credit for bringing about, organizing, and personally contributing heavily to this volume that merits a prominent place in the libraries of human behavior researchers, as well as those interested in helping solve some of the larger socio‐technical issues facing us. I heartily recommend this volume to all of you.
22 July 2018
Greg L. Zacharias
Weston, Massachusetts