We appreciate your attention to our article1 and your comments on it. The description of the statistical methods may not be clear.
The statistical methodology used was as follows.
Normality was tested with the Kolmogorov-Smirnov test. The eight SF-36 dimensions and the two summary measures at the four established times were compared with the nonparametric Friedman test, and multiple comparisons were then made using one-way analysis of variance. Before applying one-way analysis, we performed a rank transformation of the data, which accounts for the issues associated with variables with non-normal distribution.
In our paper, we omitted to state that we used the Friedman test and a rank transformation of the data prior to application of one-way ANOVA. Rank transformation versions of parametric tests closely mimic the properties of the best-known rank randomization tests (e.g. Spearman's rho, Mann-Whitney U or Wilcoxon T) with moderate to large sample sizes.2
In another article we published on the same subject, the reviewers had the same doubt you had, and we changed the section on statistical methodology accordingly.3 They accepted the justification that we provide in the previous paragraph.
Conflicts of interestThe authors have no conflicts of interest to declare.