In the applied world of product testing the appropriate number of pane
lists (base size) involves technical and business considerations. Base
sizes range from very low (around six; used in expert panelist profil
ing) to high (hundreds; used in product tests by marketing researchers
). Often base sizes are dictated by the requirement that the project i
dentify statistical differences between or among samples. The probabil
istic analysis of differences (significance vs. insignificance) derive
s from statistical theory, with bare size used as a method to influenc
e the sampling error (variability). This paper looks at bare sizes ano
ther way-from the viewpoint of psychophysical scaling. The issue then
can be re-stated as 'what is the necessary base size at which the aver
age rating stabilizes?' Empirical data suggest that base sizes of 40-5
0 panelists generate stable averages, and that beyond the 80 panelists
the average is not particularly affected by the base size. These resu
lts hold for actual data for a variety of products, and for different
types of attributes, specifically sensory (amount of a characteristic)
, and hedonic (liking of a characteristic). (C) 1997 Elsevier Science
Ltd.