When comparing test scores, the fundamental question is, How discrepant must two test scores be for the difference to be significant? Reynolds (1985) offers a regression procedure for determining how large a Difference Score must be to obtain statistical significance. This formula can be used if the intercorrelation between two scores is known and if reliability coefficients for both scores are available. Because both indices are available with respect to the GORT-3 subtests, we were able to use Reynolds's procedure to generate Table 3.2, which provides minimum Difference Scores for comparing GORT-3 subtests.