Notes on designing an experiment

What measurements are to be recorded?

Write down everything to be recorded, for example
  • weight in kg at 15 days old
  • proportion of diseased plants in each of 10 samples of 100 plants from each plot (so the samples are the observational units)

    It is a good idea to have a data sheet, either on paper or a spreadsheet, prepared in advance, with one row for each observational unit and one column for each measurement.

    Record everything at the proper time; do not go back and make it up later. For example, if you are supposed to record temperature then do so; don't simply look back at someone else's temperature records later and copy the figure.

    Do not make `neat' copies: you are likely to introduce errors.

    protocol purpose treatments experimental units method observational units
    measurements design justification randomization plan analysis

    Page maintained by R. A. Bailey

    Page updated 17/11/01