Advertisement

Modern Bayesian Statistics in Clinical Research

  • Ton J. Cleophas
  • Aeilko H. Zwinderman

Table of contents

  1. Front Matter
    Pages i-x
  2. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 1-22
  3. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 33-39
  4. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 41-48
  5. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 49-58
  6. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 59-68
  7. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 69-82
  8. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 83-89
  9. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 91-99
  10. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 101-110
  11. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 111-118
  12. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 119-130
  13. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 131-142
  14. Ton J. Cleophas, Aeilko H. Zwinderman
    Pages 143-173
  15. Back Matter
    Pages 175-188

About this book

Introduction

The current textbook has been written as a help to medical / health professionals and students for the study of modern Bayesian statistics, where posterior and prior odds have been replaced with posterior and prior likelihood distributions. Why may likelihood distributions better than normal distributions estimate uncertainties of statistical test results? Nobody knows for sure, and the use of likelihood distributions instead of normal distributions for the purpose has only just begun, but already everybody is trying and using them. SPSS statistical software version 25 (2017) has started to provide a combined module entitled Bayesian Statistics including almost all of the modern Bayesian tests (Bayesian t-tests, analysis of variance (anova), linear regression, crosstabs etc.).

Modern Bayesian statistics is based on biological likelihoods, and may better fit clinical data than traditional tests based normal distributions do. This is the first edition to systematically imply modern Bayesian statistics in traditional clinical data analysis. This edition also demonstrates that Markov Chain Monte Carlo procedures laid out as Bayesian tests provide more robust correlation coefficients than traditional tests do. It also shows that traditional path statistics are both textually and conceptionally like Bayes theorems, and that structural equations models computed from them are the basis of multistep regressions, as used with causal Bayesian networks. 

Keywords

Bayesian t-tests Clinical Research Bayesian regressions Bayesian crosstabs Bayesian anovas Markov Chain Monte Carlo samplings

Authors and affiliations

  • Ton J. Cleophas
    • 1
  • Aeilko H. Zwinderman
    • 2
  1. 1.Department Medicine Albert Schweitzer HospitalAlbert Schweitzer HospitalSliedrechtThe Netherlands
  2. 2.Department Biostatistics and EpidemiologyAcademic Medical Center Department Biostatistics and EpidemiologyAmsterdamThe Netherlands

Bibliographic information

  • DOI https://doi.org/10.1007/978-3-319-92747-3
  • Copyright Information Springer International Publishing AG, part of Springer Nature 2018
  • Publisher Name Springer, Cham
  • eBook Packages Medicine
  • Print ISBN 978-3-319-92746-6
  • Online ISBN 978-3-319-92747-3
  • Buy this book on publisher's site