Supplementary material for Spatial Regression Models

This webpage contains replication material and data for Michael Ward and Kristian Skrede Gleditsch. 2018. Spatial Regression Models, Second Edition. Thousand Oaks, CA: Sage.

About the book

  • Link to the book at Sage

    About the authors

    Michael D. Ward (1948-2021) was professor emeritus at Duke University, and a previous faculty member at the University of Washington, University of Colorado, and Northwestern University. For an overview of his interest in geography and space, see this symposium in the journal Political Geography. He played a key role in disseminating insights on the role of geography and spatial concepts and methods to the social sciences. This book is a testament to his legacy, and we hope it can inspire other scholars.

    Kristian Skrede Gleditsch is Regius Professor of Political Science at the University of Essex and research associate at the International Peace Research Institute, Oslo (PRIO). His research research interests include conflict and cooperation, democratization, and spatial dimensions of social and political processes.

    Code and data

    The below file provides updated R code for the examples discussed in Ward and Gleditsch 2018. This code has been updated from the original code in July-August 2023 by Kristian Skrede Gleditsch and Andrine Fogstad. It uses the R Groundhog package for version control

    Much of the original code for the 2018 edition of this book was written by the late Michael D. Ward, ca 2017. In some cases it relied extensively on R packages and other sources that have changed substantially since then. Unfortunately, R packages are often revised, and reported code that is correct at one point in time may not work at a later time, following revisions and their implications. Groundhog makes R code replicable for the future by using the specific version of the packages available at a specific date.

    For the 2008 version of this book we used a set of precompiled data for the examples. In the revised code for the 2018 edition Mike Ward wanted to showcase the range of options available in R, and in some cases relied on real-time data extraction and methods for multiple imputations of missing data. We have tried to stick as closely as possible to the original code, and primarily made revisions to ensure that code as similar as possible will continue to run with R and package versions as of the time of writing. Mike cared very deeply about this code; the code reflected his manner of working and thinking, and we hope that this can inspire users of the book. However, the reliance on real time and updated data also means that the results in some cases do not match exactly what is printed in the book, and may change with additional revisions to the data. However, the results remain substantively similar. There are also some changes outside R that affect the original examples. For example, one example used Google Maps, which now requires API registration. We note these issues in the text when applicable.

  • Click here for a zipped archive of enhanced R code from the book by chapter (last updated 2023-08-08)
  • Click here for a zipped archive with the data. I recommend unzipping these files to a separate working directory.


  • Install R from CRAN

    Contact details

  • Kristian Skrede Gleditsch, University of Essex, ksg{at}essex{dot}ac{dot}uk

    Return to home page