Formal experimental designs

These designs are research methods based on statistical principles. This means that in analysing the results not only the result of the experiment can be calculated, but also an estimate can be made of the degree of experimental error in the procedure. Formal experimental designs therefore do for market experimentation what random probability sampling does for sample selection. They allow the experimenter to calculate not only the results of the experiments, but also the probability that the results are good within specified limits of accuracy, at a given confidence level. Since the declared aim of this book is to explain marketing research procedures without going into statistical detail, it would not be appropriate to explain how formal experimental designs are constructed. The non-statistical research user should simply take note of the fact that the use of appropriate statistical techniques in the design of market experiments can increase the number of variables that can be tested in one experiment, allow for calculation of experimental error and assess the interaction effects. The price that must be paid for these benefits is increased design complexity and cost. Four designs are introduced.

Was this article helpful?