Epidemiological investigation designed to test a hypothesized cause-effect relationship by intentional change of a supposed causal factor in a population.