Week 5-Wilcoxon and Mann Whitney U

Question 1

What’s the Mann-Whitney U?

Answer 1

a non-parametric version of the independent t-test
■Looks at the difference between two un-related (between-subjects) samples of scores.
■Based on difference ranks of the scores
■Big differences between total rank scores are more likely to produce statistical significance.

Question 2

What’s the assumptions of the MWU?

Answer 2

■If data are continuous / scale (i.e., Interval or Ratio), you would use MWU only if data are skewed
■Alternatively, you would use MWU if data are ordinal

Question 3

State the effect size equation

Answer 3

r=Z/n1+n2(square root)
Z
n1+n2(group 1 size+group 2 size)

Question 4

What’s Wilcoxon signed rank test?

Answer 4

The non-parametric version of the paired samples t-test (within-subjects design)
■Wilcoxon signed ranks test looks at differences between two
related samples. BUT it’s based on ranking the data rather than just
the sign of the difference.

Question 5

What are the assumptions of the WSR?

Answer 5

■If data are continuous / scale (i.e., Interval or Ratio), you would use WSR only if data are skewed
■Alternatively, you would use WSR if data are ordinal

Question 6

How do you rank data?

Answer 6

1.Scores are ordered from lowest to highest.
2.The lowest value is given a rank of 1, the second lowest a rank of 2
etc.
3.Sometimes though there may be values that are the same. In this
instance what the ranks would be (if the numbers were different)
are added together and divided by how many values are the same.
These values all get the same rank. If you have to do it the next
value (which is different) is ranked as normal.

Question 7

How can we interpret effect size?

Answer 7

■ r = .10 small
■ r = .30 medium
■ r = .50 large