In Reply.—In response to Dr MacMahon's letter, I would like to address the following points:
Subjects were randomly assigned to treatment groups. All subjects were recruited from the same sports teams, and all were interested in strength training. The experimental group underwent strength training, while control subjects were told that they would begin strength training at the end of the 14-week experimental period. (They did train after the strength-training study was over.) This is mentioned in the "Pubertal Status" section of the article.
I am sensitive to the concern relative to initial group inequity. The possibility that the groups might differ prior to experimental treatment motivated the inclusion of a pretest in the experimental design.
Although my colleagues and I considered the use of an analysis of covariance (ANCOVA) in this case, we preferred to use its repeated-measures analysis of variance (ANOVA) counterpart. The conflicts regarding which