Multiple matrix sampling of three subscales of the California Psychological Inventory was used to investigate the effects of four variables on error estimates of the mean (EEM) and variance (EEV). The four variables were examinee population size (600, 450, 300, 150, 100, and 75); number of subtests, (2, 3, 4, 5, 6, and 7), hence the number of items per examinee; skewness; and kurtosis. The relationship of population size, number of subtests, and skewness and kurtosis of each subscale/population combination to the EEM and EEV was examined, by performing an analysis of variance on stepwise multiple linear regression. A total of 324 individual multiple matrix samples were involved. EEM was not significantly affected by any variable studied; but EEV, in contrast to previous studies, was significantly affected by skewness and kurtosis. (CP)