ABSTRACT
In sensory testing, a given problem frequently requires appreciable thought before the appropriate practical tests can be selected (IFT, 1981). This is because the initial conception of the problem may require clarification. It is not unusual for problem and test objectives to be defined and redefined several times before an acceptable design emerges. Sensory tests are expensive, and they often give results which cannot be interpreted. If this happens, the design may be at fault. Pilot tests are often useful as a means of refining a design. It would, for example, be meaningless to carry out a consumer preference test with hundreds of participants without first having shown that a perceptible difference exists; the latter can be established with 10 or 20 tasters, using a difference test. In another example, islands of opposing preference may exist, invalidating a normal preference test; here, the solution may be a pilot study in which various types of customers receive single-sample acceptability tests.
