Word Doc Week 9

Question 1

Kruskal-Wallis Test

Answer 1

• Non- Parametric ANOVA for single factor between subjects designs
• Is there a significant difference among different groups on the ranks associated with reported scores
• Tells us there is a significant difference but does not reveal the source
• post-hoc tests can help but must be done manually

Question 2

Steps involved in running post-hoc for Kruska-Wallis

Answer 2

1. Run every possible pairwise comparison among all groups
2. Use a non-parametric equivalent of independent samples t-test to run these tests
   
   • called Mann-Whitney U test
3. Adjust the critical p value such as a level manually
4. a = .05 divided by the number of post-hoc tests you run
5. formula - .05/(no. post-hoc tests) = U
6. Each Mann-Whitney U test must have an observed .05/(no. post-hoc tests) = U to be significant
7. This is lcalled a Bonferroni correction

Question 3

Friedman Test

Answer 3

• Non-Parametric Single-Factor Within Subjects ANOVA
• Null hypothesis says there is no difference between data reported at different times

Question 4

Steps involved in running post-hoc tests - Friedman Test

Answer 4

1. Run every possible pairwise comparison among all groups
2. Use non-parametric equivalent of paired samples t-test
3. Test is called a Wilcoxon signed rank sum test
4. Adjust the critical p value manually
5. .05/3=.017
6. Each Wilcoxon test must have an observed p < .017 in order to be significant
   
   • This is call a Bonferroni correction