In the research planning stages of a project, sometimes replication of treatments is impractical, prohibitively expensive, or impossible. Perhaps the researcher does not have enough material to fully randomize and replicate, or there are physical space and/or time limitations, or too many barriers to complete the data collection required for a fully replicated experiment to answer the proposed research question. In these scenarios, it can be advantageous to use unreplicated or augmented experimental designs. These designs reduce costs with an acceptable loss of precision. Other (though not all) important consequences are: (i) not all treatments are compared with the same precision; (ii) experimental error is partially estimated; (iii) in factorial experiments, some interaction effects cannot be estimated and some are confounded; (iv) standard experimental designs can (and should) be modified to deal with unreplicated treatments; and v) the statistical analysis approach is usually different to the usual analysis of variance.

Maize CRP FP2 - Novel tools, technologies and traits for improving genetic gains and breeding efficiency

Wheat CRP FP3 - Global partnership to accelerate genetic gain in farmers field

EiB

Text in English