THE KLUGE‐KERFOOT PHENOMENON—A STATISTICAL ARTIFACT

Abstract

Kluge and Kerfoot (1973) reported that for seven vertebrate data sets the relative amount of interlocality differentiation for a given morphometric variable was positively correlated with the amount of its within-population variation. Since that time this empirical relationship (now called the Kluge-Kerfoot phenomenon, Sokal, 1976) has been demonstrated in a number of data sets for a wide variety of organisms (insects, mollusks, fish, protozoa, and birds). In almost every study in which this relationship was examined some positive association was found. The underlying premise of Kluge and Kerfoot (1973) was that, if there were a positive relationship between the measures of variation within and among populations, it must be attributable at least in part to the variability of