Issues concerning the design and analysis of comparative field experiments
Abstract:In scientific publications, a clear report of an appropriate statistical analysis allows independent assessment of the results and adds weight to the conclusions. Poorly described statistical analyses of field experiments give little justification for the conclusions drawn. However, to gain maximum benefit, statistical analysis should always be considered at the planning stage of the experiment, in the choice of treatments applied and the experimental design. Good choices at this stage can enable a suitable statistical analysis and reduce uncertainty in treatment estimates. This paper gives guidelines on the issues that should be considered at the design and analysis stages to facilitate good experimentation and reporting. These are principles and guidelines rather than rules, as the best choices for any particular experiment will depend on the context and aims of the experiment. Further general information on the design and analysis of field crop experiments aimed at practitioners can be found in Steel & Torrie (1990), Mead, Curnow & Hasted (1993) or Snedecor & Cochran (1989). However, these texts may be less helpful to scientists unused to mathematical notation. A better solution may be an early approach for advice to a statistical consultant with experience in the design and analysis of agricultural field plot experiments.
Document Type: Research Article
Publication date: August 1, 2006
More about this publication?