Of science, meta-science, and clinical practice: the generalization of a generalization to a particular

Although science is the most powerful method for advancing knowledge and scientific and statistical formalisms are generally much sounder than impressionistic judgment, advanced methodological training (as it is presently comprised) is often of little direct use to practicing clinicians. Practitioners should be able to identify grossly deficient knowledge claims and clear scientific winners, but beyond this, the trick usually is to determine which scientific authority or authorities on whom to depend, a decision task that is ultimately to be tackled through the type of meta-scientific studies that Faust and Meehl have proposed. In applying scientific knowledge, the clinician usually must choose between the generalization of a scientific generalization to a particular versus the application of a clinical generalization to a particular. Determining the relative power of scientific generalizations developed outside, versus clinical generalizations developed within the setting of application, can be very difficult, and there are few established guides. A more advanced science of generalization, another meta-scientific problem, would allow sounder predictions about application to new domains and could greatly aid the practitioner.