Askovitz' Study on Paired Comparisons

In Holland, the one-side treatment (simultaneous paired comparison method, Sulzberger)* had already been applied before World War I (Siemens, 1938, 1939).4,5 Especially, attention has been paid to the clinical application of this method (Siemens, 1952, 1958).6-8 In other countries, however, scientists hardly considered this kind of research, in spite of the fact that Sulzberger has frequently pointed out the importance of the method (Sulzberger, 1940, 1946, and other places).10,11 This is the reason why Askovitz' investigations (1958)1 have drawn our special attention. In his study a changed statistical evaluation of the results of the paired comparisons has been proposed. Askovitz' argumentation for this changed evaluation is briefly the following: Very often in paired comparison testing, patients are recorded "in whom the response to the new drug is no different clinically from the response to treatment with a blank control." The number, ND, of these patients in whom