Figure 2.

Sample size required in each group to detect a certain partition coefficient difference with a two group design of 80% power and an alpha error of 0.05. The X axis values corresponding to the partition coefficient difference need to be detected like the first column in Table 4. The three curves corresponding to case 1, 2 and 3 of Table 4. The smaller partition coefficient difference and higher inter-study SD, the larger the sample size needed. The dashed line corresponding to the sample size needed to detect a 0.063 partition coefficient difference for the three cases as showed in Table 4.