To the Editor: The article by Burke et al, "Variability of Threshold Shift in SAL Technique," which appeared recently in the Archives of Otolaryngology (80:155 [August] 1964), contains a very curious piece of logic. The authors found that the threshold shifts obtained by the Sensorineural Acuity Level (SAL) technique in normal listeners were normally distributed, with a standard deviation of about 6 db. They concluded that "... this variability precludes the use of the SAL as a replacement for conventional bone-conduction audiometry and raises serious question as to its value as a diagnostic audiologic procedure.
These authors seem unaware that the norms for certain other audiometric procedures, notably conventional air-conduction and bone-conduction audiometry, are derived from data with a very similar distribution. The figure, for example, compares the standard deviations of the distributions of AC, BC, and SAL in normal listeners. The standard deviations for AC are taken from Corso1