*3.4. Analysis Performed*

The lack of randomisation has led to non-parametric analyses to measure the effects. *Wilcoxon (Z) t*-tests—a non-parametric test equivalent to the Student's *t*-test—were carried out to determine whether the programme had an effect on the adolescents. This test was used because the distributions failed to meet the criteria of normality and homoscedasticity, in addition to being based on small subsamples. Most of the measurements that were taken correspond to ordinal variables with few ranges. For each of the *Z* tests, Cohen's *d* was calculated as the effect size coefficient [57] (p. 12) [58].
