DOEpro - Analysis of Designed Agricultural Experiments
A 'shiny' application and supporting functions for the
analysis of designed agricultural experiments, following the
procedures set out by Gomez and Gomez (1984,
ISBN:9780471870920). Handles completely randomised, randomised
complete block, Latin square, factorial (up to four factors),
split-plot and strip-plot designs, and pooled (combined)
analysis over environments including factorial treatments,
after Yates and Cochran (1938) <doi:10.1017/S0021859600050978>;
analyses several response variables simultaneously; recommends
and applies variance-stabilising transformations using the
profile likelihood of Box and Cox (1964)
<doi:10.1111/j.2517-6161.1964.tb00553.x>; reports standard
errors and critical differences for every legitimate
comparison, including the four distinct comparisons of a split
plot, the mixed ones using the approximation of Satterthwaite
(1946) <doi:10.2307/3002019>; and produces publication-format
tables of means, diagnostic plots and a written interpretation.
