Bayesian analysis of agricultural field experiments

• Published in: Journal of the Royal Statistical Society. Series B, Statistical methodology Statistical methodology, 1999, Vol.61 (4), p.691-746
• Main Authors: Besag, J., Higdon, D.
• Format: Article
• Language: English
• Subjects: Agricultural economics; Agricultural field trials; Agriculture; Analytical estimating; Applications; Bayesian analysis; Bayesian computation; Bayesian inference; Bayesian method; Binary data; Combining information; Crop experiments; Crops; Design analysis; Exact sciences and technology; Experiment design; Experimentation; Field experiments; Field work; Frequentism; Insurance, economics, finance; Intrinsic autoregressions; Markov chain Monte Carlo methods; Markov random fields; Markovian processes; Mathematical foundations; Mathematics; Modeling; Monte Carlo simulation; Prior distributions; Probability and statistics; Ranking and selection; Regression analysis; Sciences and techniques of general use; Spatial models; Spatial statistics; Statistics; Variety trials
• ISSN: 1369-7412; 1467-9868
• DOI: 10.1111/1467-9868.00201

The paper describes Bayesian analysis for agricultural field experiments, a topic that has received very little previous attention, despite a vast frequentist literature. Adoption of the Bayesian paradigm simplifies the interpretation of the results, especially in ranking and selection. Also, complex formulations can be analysed with comparative ease, by using Markov chain Monte Carlo methods. A key ingredient in the approach is the need for spatial representations of the unobserved fertility patterns. This is discussed in detail. Problems caused by outliers and by jumps in fertility are tackled via hierarchical-t formulations that may find use in other contexts. The paper includes three analyses of variety trials for yield and one example involving binary data; none is entirely straight-forward. Some numerical comparisons with frequentist analyses are made.