Date of this Version
Farmers are interested in evaluating new agricultural practices on their own farms. To produce results credible to themselves, other farmers, and researchers, a systematic method of testing should be used. If a comparison of agricultural practices results in one practice yielding a few more bushels of crop than another, this does not necessarily mean it is a more valuable practice. The difference may simply be due to field variation or chance.
This circular discusses the on-farm trials and incorporation of the two basic requirements in designing experiments: randomization and replication.