A quantitative hypothetico-deductive approach has continued to contribute greatly to advances in biological and medical science. Quantitative methods are adopted over other approaches primarily because they contribute the most new knowledge about biological processes. Nevertheless, investigators make many assumptions when testing a biological hypothesis quantitatively. These assumptions may become invalid unless experiments are designed with great care. Problems arise in relation to formulating appropriate hypotheses, using volunteer samples, controlling the experimental intervention and potentially interfering behaviors, reaching an acceptable level of proof, excluding alternative hypotheses, and generalizing findings beyond the immediate experimental sample. When biologists are aware of these issues, they can take appropriate countermeasures and reach valid conclusions. However, the issues become more critical and resolution is less clear-cut when the same methods are extended from biology to psychology and the social sciences, and from general to special populations. In such situations, case studies and single-subject designs may have continuing relevance.
Roy J. Shephard is with the Faculty of Physical Education and Health at the University of Toronto. Direct all correspondence to PO Box 521, Brackendale, BC, VON 1H0, Cananda.