ABSTRACT
In systems which are relatively easy to manipulate, designing experiments with a large number of different treatments for several fixed variables can lead to sophisticated research. However, these experimental designs also require sophisticated analysis techniques. In general, whatever type of experiment or survey, the most elegant research providing the most meaningful results tends to be the result of simple, well-thought-out designs. One way of reducing this edge effect is to incorporate a buffer zone surrounding the plots into the design, where treatments are applied, but measurements are not taken. These randomised block designs are commonly used in field experiments; however, they are also applicable to the way in which many other experiments are designed. Questionnaire design when eliciting information about human activities, or the opinions and attitudes of a section of the population to a particular issue, then some kind of interview technique is appropriate.
