Hammers and Nails

An Introduction to Bayesian Statistics Through Applications
1. Auflage 2027
ISBN: 978-1-032-88970-2
Verlag: Taylor & Francis Ltd

Buch, Englisch, 392 Seiten, Format (B × H): 178 mm x 254 mm

Hammers and Nails is an example-driven guide to Bayesian statistics that uses small, memorable datasets to show how the well-known flexibility of Bayesian modelling can be used to answer the questions researchers and policy makers may actually ask. The book moves from fundamentals to numerical methods and advanced applications. The latter part covers, among other topics, survival analysis, zero-inflated and other mixture models, and the treatment of missing data. While the basics of regression modelling are inevitably covered as well, examples show how the Bayesian regression framework can be used to answer practical questions beyond the interpretation of model coefficients.

Throughout the book, the research question informs every aspect of the modelling process. The importance of assumptions and diagnostics is made explicit through numerous worked examples, with R code provided. Compact datasets keep ideas visible and place the entire modelling process within a relatable practical context, drawing on the author’s many years of experience as a statistical consultant.

Key Features:

- Contains concrete, real-world case studies spanning safe speed limits, identifying daily temperature extremes, traffic flow, gym attendance, engineering properties of eucalyptus trees, seed germination, the performance of 3D printers, and the treatment of missing data.

- Introduces a comprehensive computational toolkit covering sampling methods, Monte Carlo integration, MCMC methods, particle filters, and approximate Bayesian computation.

- Provides algorithms and R code for Bayesian models applied to a wide range of worked examples.

- Includes end-of-chapter exercises, a consolidated list of algorithms, and appendices on probability theory, probability functions, and key distributions, bridging theory, derivations, and computation.

This book is intended for applied researchers, statisticians, data scientists, and postgraduate students in quantitative fields who are interested in practical, research-question-driven modelling. Readers with a grounding in probability, calculus, and programming will find the material suitable for graduate study, self-study, and professional practice.

Additional materials, including datasets used in the examples, solutions to exercises, and errata, are available at https://github.com/ANOVA-droid/bayesian-hammersand-nails.

Zielgruppe

Postgraduate, Undergraduate Advanced, and Undergraduate Core

Autoren/Hrsg.

Moltchanova, Elena

Fachgebiete

• Mathematik | Informatik Mathematik Stochastik Bayesianische Inferenz

Inhaltsverzeichnis

Part I: The Fundamentals  1. Opening the Tool Box 2. Introducing Bayes’ Formula 3. A Bayesian Sandwich Shop 4. The Best, the Worst and the Irrelevant 5. Capture–Recapture Models and Gibbs Sampling 6. Normal Conjugate Model and Good Enough  Part II: Numerical Methods  7. Monte Carlo Integration and Sampling Methods 8. Approximate Bayesian Computation and Particle Filters 9. The Metropolis-Hastings Algorithm  Part III: Applications  10. Gym Attendance And Zero-Inflated Models 11. Mixture Models and the Plight of Cinderella 12. Old Cars Take a Long Time to Stop 13. Beyond the Coefficients. The Quantities that Actually Matter. 14. Mixed-Effects Models and Seeing the Wood for the Trees 15. A Matter of Survival. Time-to-Event Data, Censoring and Truncation 16. Change Points and Bayesian Model Averaging 17. Missing Data 18. Afterword

Über Autor(innen)

Elena Moltchanova first became interested in Bayesian statistics back in the last century, during her undergraduate studies. She subsequently earned her PhD in Bayesian spatio-temporal modelling from the University of Jyväskylä, Finland. She is currently Professor of Statistics and Head of the Statistical Consulting Unit at the School of Mathematics and Statistics, University of Canterbury, New Zealand.

She has authored or co-authored over 150 journal articles and book chapters applying statistical methods to a wide range of fields, including epidemiology, forestry, ecology and archaeology. Her career has taken her from the National Public Health Institute of Finland to the International Institute for Applied Systems Analysis (Austria), and finally to the University of Canterbury, providing opportunities to collaborate with researchers and stakeholders from many different disciplines.

Having spent over fifteen years helping researchers navigate the complexities of statistical analysis, she believes that good statistics begins not with choosing a method, but with asking the right question. She is passionate about communicating statistical methods and results in ways that are relevant and accessible to researchers, policymakers and the general public. This commitment is also reflected in her development of online courses in statistics for the edX platform.

Hammers and Nails: An Introduction to Bayesian Statistics Through Applications is her first book, born of many years of teaching, research and consulting, and the occasional frustration of seeing perfectly good statistical methods applied to entirely the wrong questions.