Stability of Individual Differences in Heart Rate Activity During the Newborn Period

ABSTRACT Reliability of 14 measures of newborn heart rate activity was estimated by intercorrelations among repeated daily testings and by the coefficient of concordance. Subjects were 16 newborns tested for three days and a subsample of 10 newborns tested for five days. Thirteen of the measures showed some significant stability although they differed in the size of correlations and in the degree of concordance. The two measures of temporal characteristics of response were among the least reliable and measures taken at the point of minimal activity were among the most reliable. The only measure for which there was no evidence of significant reliability was the within‐individual regression of poststimulus on prestimulus scores. This finding has obvious implications for the argument that individual regressions should be used instead of group regressions to adjust for prestimulus differences.