Introduction > Planning Your Research > Designing Your Experiment > Replication

Replication A replicate is a repetition of your experiment. You should never conduct an experiment using just one sample because, even with a control, you could never be truly certain that the recorded effects were due to your treatment. For example, if a pharmaceutical company gave a drug to one volunteer, and that volunteer then suffered a 10 minute sneezing fit, can you be sure that the sneezing fit is attributed to the effects of the drug? The volunteer could have reacted to the drug, but equally so they could have reacted to pollen in the air, the dust mites in the carpet and so on. If the experiment was replicated, and the drug was given to 100 volunteers, you can more safely deduce that the volunteers were reacting to the drug. The more replication you have, the more accurate your results will be and the less likely that any effects observed are occurring by 'chance' alone. Statistically, the minimum number of replications you should use is three. The maximum number is often determined by factors that are outside your control, such as the time it takes to collect your data and resources available to you.

<<< Previous Page >>><<< Next Page >>>