Statistical Methods in Immunology

Abstract

The techniques employed in relation to statistical aspects of the experimental material presented in this report no doubt are unfamiliar to many immunologists. Actually, they are relatively simple but extremely useful tools provided by the statisticians. No doubt the statisticians will provide many more,—that is their proper mission. Judicious application of these techniques requires specialized knowledge of the field of inquiry and is the responsibility of the investigator. Statisticians rightfully can be expected to assist in testing hypotheses; in obtaining reliable answers to questions of concern to immunologists. They cannot be expected to frame those questions. Yet, the very framing of a question so that ultimate statistical treatment is applicable demands a certain understanding and appreciation of statistical methodology. If the foregoing is granted, it appears self-evident that immunologists must learn to think and talk, at least in part, in terms of degrees of freedom, orthogonal comparisons and sigma squared. Perhaps when that time comes we will find that statisticians will lapse into the vernacular and think and talk, at least in part, in terms of “bugs”, mice and milligrams of antibody nitrogen.