Add power analysis to hypothesis tests

[krajcsi]

Sensitivity analysis. Before presenting the hypothesis test result, display the minimum effect size that might have appropriate power with the given sample size.

Retrospective interval analysis. Consider adding a a retrospective analysis based not only on a point estimation of the effect size but on an interval estimation of the effect size. The result would be an interval of possible retrospective power.

[krajcsi]

Already implemented sensitivity analyses:

1. • Shapiro-Wilk test
2. • One-sample t-test
3. • Wilcoxon signed-rank test
4. • Pearson correlation hypothesis test
5. • Spearman correlation hypothesis test
6. • Chi-squared test
7. • paired t-test
8. • paired Wilcoxon test
9. • repeated measures ANOVA
10. • Mauchly's sphericity test (a)
11. • Holm-Bonferroni corrected post-hoc tests
12. • Friedman test
13. • Durbin-Conover test
14. • McNemar's test
15. • Cochran Q-test
16. • modified t-test
17. • Mann-Whitney test
18. • Levene's test (a)
19. • two-sample t-test
20. • Welch's t-test
21. • one-way ANOVA
22. • post hoc Tukey's HSD tests
23. • Kruskal-Wallis test
24. • post hoc Dunn's test
25. • Two-way ANOVA