A Two-Part Generalized Linear Mixed Modelling Approach to Analyze Physical Activity Outcomes

Andy H. Lee, Liming Xiang and Fumi Hirayama
Curtin University of Technology, Perth, WA, Australia, and others

Series: Mathematics Research Developments, Public Health in the 21st Century
BISAC: MAT002050

Clear

$0.00

eBook

Digitally watermarked, DRM-free.
Immediate eBook download after purchase.

Product price
Additional options total:
Order total:

Quantity:

Details

Physical activity (PA) is a modifiable lifestyle factor for many chronic diseases and its health benefits are well known. PA outcomes are often measured and assessed in many clinical and epidemiological studies. This book first reviews the problems and issues regarding the analysis of PA outcomes. These include outliers, presence of many zeros and correlated observations, which violate the statistical assumptions and render standard regression analysis inappropriate. An alternative two-part generalized linear mixed models (GLMM) approach is proposed to analyze the heterogeneous and correlated PA data. At the first part, a logistic mixed regression model is fitted to estimate the prevalence of PA and factors associated with PA participation. (Imprint: Nova)

ABSTRACT

1. INTRODUCTION

2. PHYSICAL ACTIVITY ASSESSMENT

3. TWO-PART GENERALIZED LINEAR MIXED MODELS

4. EXAMPLE

5. DISCUSSION

ACKNOWLEDGEMENT

APPENDIX

REFERENCES

INDEX

You have not viewed any product yet.