Experimentation

In systems analysis, experimentation is the process of determining the results of a proposed course of action or program by conducting an experiment on a smaller scale in which the course of action is applied to a sample drawn from the future target group. An example would be a test of a new health policy in a restricted region instead of the whole country, or a test on a randomly selected sample of the population. The results are best when the experiment is controlled--i.e., when the test and control groups are chosen before program implementation in such a way that they are as similar as possible. In this way, any differences that are observed during the experiment can be ascribed to the program. Experimentation is used whenever current knowledge and understanding of factors such as social attitudes and group preferences are not sufficient to provide dependable model -based predictions. (IIASA)