An Application to a Laboratory problem of Discriminant Function Analysis involving more than two Groups

Abstract

We have recently attempted to assess the value of a modification of the Wassermann reaction, based upon the suggestion of Richardson (1940). The results are being described in a pathological journal (Jennison, Penfold, and Roberts, 1949). The present paper deals with certain points which emerged during the statistical analysis, and which have a rather more general interest in regard to the application of statistical methods to biological data.