The equivalence test was used to calculate the 95% confidence intervals (CIs) for each statistic (the difference between the means) by bootstrapping using the percentile method (with an equivalence test, the null hypothesis is the converse of H0, i.e. the null hypothesis is non-equivalence).