<i>z</i>-score distributions of examples of individual subjects.

<p>Left panel shows scatterplots of <i>z</i>-scores versus distribution mean, <i>m</i>, for three typical subjects [the mean <i>z</i>-scores of subjects A, B, and C were the fourth highest (), median (), and fourth lowest () of all subjects, respectively]. Broken lines indicate regression lines of <i>z</i>-scores on the distribution mean for each subject. The correlation coefficients for Subjects A, B, and C were −0.06, 0.07, and −0.05, respectively. Right panels show histograms of <i>z</i>-scores from these three subjects. Each bar has a bin width of 0.5. Thus, the number of observed <i>z</i>-scores (indicated by a bar at <i>z</i>₀) includes . Each curve is a fitted normal distribution (, averaged over all 91 subjects). Mean <i>z</i>-scores significantly differed between subjects (unpaired, two-tailed <i>t</i> tests, versus : , ; versus : , ).