discriminant function: a comment

by discriminant function (Campos et al 2005) quotes wing lengths exceeding 100 mm for five of 51 females sexed by molecular methods. The occurrence of wings this long is noted by the authors as unusual, whereas they are in fact unprecedented. Cramp & Perrins (1994) detail wing lengths of more than 200 females taken over the species ’ range (excluding Britain and Ireland where birds are known to be smaller) and the largest of these is 98 mm. Based on these data, the probability of obtaining even one female Corn Bunting whose wing exceeds 100 mm is extremely small. Campos et al (2005) discount the possibility that measurement error was involved, but do not say explicitly how this conclusion was arrived at. A common explanation for data showing marked outliers is systematic error