Ordinal Data: Nonparametric Statistical Analyses And Spss Applications Info

This shows the spread of the middle 50% of your data. Effect Size: For Mann-Whitney or Wilcoxon, calculate to show the magnitude of the difference. Conclusion

You risk Type I errors (finding an effect where none exists). This shows the spread of the middle 50% of your data

Parametric tests (e.g., t-test, ANOVA) assume: This shows the spread of the middle 50% of your data

Assess if a treatment or time interval caused a consistent shift in rankings. This shows the spread of the middle 50% of your data