From Experimental Network to Meta-analysis

Chapter 1. Introduction and examples<br> Objectives of the analysis of experimental networks and meta-analysis<br> Data<br> The type of data<br> The data collection<br> Data validation<br> Analysis<br> Main steps<br> Presentation of the tested hypotheses<br> Collection of data<br> Data validation<br> Data analysis<br> Validation of the analysis<br> Communication of results<br> Objective of the book<br> A simple example of a mixed model<br> Definition<br> Data<br> Model definition<br> Estimate<br> Comparison with the model without random effect<br> References<br> <br> Part I. Analysis of experimental networks<br> <br> Chapter 2. Basic Concepts<br> Agronomic experimentation<br> Experimental network<br> Definition<br> Example of experiment network<br> Environmental concept<br> Objectives of network of experiments<br> Concept of population of environments<br> Interaction concept<br> References<br> <br> Chapter 3. Analysis of network of experiments in blocks of complete randomness as a studied factor<br> Objective of the chapter<br> Example "wheat"<br> Modelization<br> Model with a random experiment effect<br> Model with a fixed experimental effect<br> Example<br> How to choose between a model with a fixed experimental effect and a model with a random experiment effect?<br> Model evaluation<br> Normality<br> Homoscedasticity<br> Independence<br> Suspicious data<br> Average comparisons<br> Hypothesis tests: equality tests<br> Confidence intervals<br> Hypothesis tests: equivalence tests<br> Example<br> Example "wheat": R script and commented analysis<br> References<br> <br> Chapter 4. Advanced Methods for Network Analysis<br> <br> Analysis of average data<br> Step 1: Analysis of individual experiments to estimate treatment averages<br> Step 2: Analysis of the average data<br> Example<br> A variant: analysis of average data with a fixed model<br> Estimation of the interaction variance treatment-experimentation<br> R script<br> Experiments with heterogeneous variances<br> Introduction<br> Example "wheat"<br> For further<br> Missing data<br> Origin of missing data<br> Adjusted averages<br> The factors place and year<br> Goal<br> Example "wheat_pluri"<br> Model for analyzing average data<br> Variance estimation of the treatment-year-place interaction<br> Variance of the difference between two treatments<br> Analysis of the example "wheat_pluri" and script R<br> References<br> <br> Chapter 5. Planning an Experimental Network<br> Goal<br> Comparison of two treatments<br> Case of a multilocal network<br> Case of a multi-local and multi-year network<br> Other contrasts<br> Average comparison of several witnesses<br> Comparison to the overall average<br> References<br> <br> Part II. The meta-analysis<br> <br> Chapter 6. Basics for meta-analysis<br> Definition, origin and main stages of the meta-analysis<br> Estimated average effect size<br> Goal<br> Systematic search of studies, selection of references and data extraction<br> Estimation of the average effect size with a model without random effect<br> Estimation of the average effect size with a random effects model<br> Meta-regression<br> Goal<br> Example<br> Regression models with and without random effect<br> Example (continued)<br> Critical analysis of results<br> References<br> <br> Chapter 7. Specific statistical problems for the meta-analysis<br> Setting the effect size<br> Correction of the bias related to the use of ratios<br> Difference between observation means<br> Effect sizes for binary data<br> Correlation coefficient<br> Effect sizes based on variance<br> Generalized linear models for discrete data analysis<br> Binomial logit model with random effects to analyze the effect of a treatment<br> Example<br> Mixed nonlinear models<br> Interest and definition<br> Example<br> Bayesian models<br> Definition<br> Example: meta-analysis with MCMCglmm<br> References<br> <br> Annex. R resources to implement the methods of analysis networks and meta-analysis<br> KenSyn package: R code and datasets of the examples presented in the different chapters<br> Installation<br> Content and use<br> Implement the mixed model under R<br> Adjust a mixed model<br> Manipulate the results of mixed models under R<br> The metafor package, dedicated to performing meta-analyzes under R<br> Bayesian approach with the mixed model<br> MCMCglmm package<br> <br> Coda package<br> References<br>