Untitled Document

Experimental Design: Randomisation

Randomisation is the simplest way to eliminate the effect of localized environmental variation. For example, a particular patch of the trial site might be less fertile or drier than the rest. A field site is almost never completely uniform. The idea is to separate the replicates of each line, so that any environmental variation affects all lines equally (on average).

There are software programs that can help you generate random numbers for your plots (for example, Microsoft Excel, TFPlan [Bowman 2000]).