Premium
MULTIRIATE NORMAL FREQUENCY DISTRIBUTIONS FOR THE ANALYSIS OF SCALP HAIR MEASUREMENTS
Author(s) -
SIMS R. T.,
KNOLLMEYER H. H. F.
Publication year - 1970
Publication title -
british journal of dermatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.304
H-Index - 179
eISSN - 1365-2133
pISSN - 0007-0963
DOI - 10.1111/j.1365-2133.1970.tb15045.x
Subject(s) - scalp , medicine , dermatology , mathematics
SUMMARY.— Observations were made on samples of 10 scalp hairs from each of 20 male undergraduates who were all 20 years of age. The greatest diameters of the hairs were measured proximally just above the hair root and distally 40 mm. away from the root when the hair was dry and wet. It was found that the paired dry proximal and distal measurements could have been a random sample from a bivariate normal distribution. The dry proximal and wet proximal measurements did not fit a bivariate normal distribution until a square root transformation was made on the wet measurements. Analysis of variance of the proximal dry measurements showed that the observations on each subject were not random samples from a homogeneous population. The mean and variance of the observations differed for each subject. In a subsidiary experiment the diameter of the hair shaft was found to be constant for distances up to 40 mm. from the hair root. It was concluded that measurements of hair shaft diameters can be described by a multivariate normal distribution for an uncountably infinite number of variables.