EVALUATION OF RESULTS OF THE TESTING OF SMALLPOX VACCINES. A REPORT ON STATISTICAL METHODS.

A report is given on the statistical methods used in the evaluation of results in the international assay of smallpox vaccines, and on special statistical problems connected with the counting of lesions (pocks and plaques).Variances due to different sources of error are estimated. These estimates are used to determine minimal testing programmes for detecting certain differences in relative vaccine potencies. Thus for comparison between three vaccines-each given in five dilutions-a minimum of two rabbits in the scarification test or 15 eggs in the pock counts (one egg per dilution) is required.To obtain the same variance in the scarification test as in the usual pock count programme (six eggs per dilution-i.e., 90 eggs for the three vaccines) 13 rabbits will be needed.