A Computational Model of Public Support for Insurgency and Terrorism

A Computational Model of Public Support for Insurgency and Terrorism: A Prototype for More-General Social-Science Modeling

Paul K. Davis
Angela O’Mahony
Copyright Date: 2013
Published by: RAND Corporation
Pages: 110
  • Cite this Item
  • Book Info
    A Computational Model of Public Support for Insurgency and Terrorism
    Book Description:

    This report builds on earlier RAND research that used qualitative conceptual causal models called “factor trees” to identify the factors that contribute to aspects of terrorism or insurgency and how the factors relate to each other. This report goes beyond the qualitative by specifying a prototype computational social-science model of public support for terrorism and insurgency. The model illustrates designing for reusability and composition.

    eISBN: 978-0-8330-8320-3
    Subjects: Political Science

Table of Contents

  1. Front Matter
    (pp. i-ii)
  2. Preface
    (pp. iii-iv)
  3. Table of Contents
    (pp. v-vi)
  4. Figures
    (pp. vii-viii)
  5. Tables
    (pp. ix-x)
  6. Summary
    (pp. xi-xviii)
  7. Acknowledgments
    (pp. xix-xx)
  8. Abbreviations
    (pp. xxi-xxii)
  9. CHAPTER ONE Introduction
    (pp. 1-6)

    This report describes and largely documents a prototype computational model that demonstrates how a more qualitative social-science model can be fully specified and implemented in a way that highlights context-dependence and accounts also for many other uncertainties. The prototype, named with the abbreviation PSOT, builds on a previously published qualitative model and deals with public support of insurgency and terrorism. Our primary purpose is to illustrate a more-general approach, but the PSOT model itself is intended as a serious straw man suitable for review, debate, and possible iteration as proves necessary. It draws on substantial prior research (Davis, Larson, et...

  10. CHAPTER TWO Specifying the Model
    (pp. 7-28)

    Most of this chapter deals with generic methods that could be useful for a variety of socialscience models. However, part of our generic method is starting with the actual problem and defining terms and concepts. Further, it is often easier to discuss methods by illustrating them with specifics rather than using only abstractions. Thus, we proceed by picking up with the factor tree that our prototype built from.

    As mentioned in Chapter One, our prototype effort built on a prior factor tree for public support for insurgency and terrorism (Figure 1.2), copied here for convenience as Figure 2.1.

    Varied definitions...

  11. CHAPTER THREE Implementation in a High-Level Language
    (pp. 29-36)

    Although we have defined and discussed all of the primary concepts of PSOT in Chapter Two, modelers will know from experience that models change, that documentation is always incomplete or imprecise, and that it is necessary to look in the model itself to know definitively what is there and assumed as defaults.

    To implement PSOT, we sought a language that would be comprehensible to researchers from diverse fields, especially social science, and that would at the same time constitute a specification model for those wishing to implement its substantive content in diverse languages and environments. The primary features that we...

  12. CHAPTER FOUR Looking Ahead to Exploratory Analysis Under Uncertainty
    (pp. 37-44)

    Resources did not permit significant analysis in our prototype effort, but we developed PSOT in part to support exploratory analyses under uncertainty. In what follows, we illustrate exploratory analysis—first at the top level (inputting the top four factors directly) and then for a deeper look at how one of the top-level factors itself depends on lower-level factors.

    To illustrate, Figure 4.1 shows results for Acceptability of Costs as a function of personal risks seen by the public (x-axis), the countervailing pressures in society against public support (e.g., pressures from friends, family, and tribes, as indicated by the colored bars),...

  13. CHAPTER FIVE Using the Model for Knowledge Elicitation, Discussion, and Diagnosis
    (pp. 45-48)

    Our primary purpose in this report is to motivate and document PSOT. Applications must wait for future projects. In this chapter, however, we summarize some hopes, expectations, and results from very limited experiments using PSOT with researchers and students.

    Based on experience prior to our model building we knew the following:

    1. Factor trees themselves have proven quite useful in group discussions: Participants can quickly spot omitted influences and raise important issues about interactions and subtleties. Discussions are more productive than those without such structure and common visual focus.

    2. Participants in group discussions can quickly grasp how to ask...

  14. APPENDIX A Primer on Factor Trees (a reprint)
    (pp. 49-66)
  15. APPENDIX B Verification and Validation
    (pp. 67-70)
  16. APPENDIX C Eliciting Factor Values
    (pp. 71-72)
  17. APPENDIX D Mathematics for “And” and “Or” Relationships
    (pp. 73-84)
  18. Bibliography
    (pp. 85-88)