Mann Whitney U test

Question 1

When is this test used for?

Answer 1

Independent measures experimental design
Ordinal level data

Question 2

Ranking positions

Answer 2

Create new table:
Place scores for each group in order from smallest to largest in a separate column
Colour code the score based on what group it was from
Rank the positions for each score

Question 3

Theory behind this test

Answer 3

Analyse the rank position of both groups scores, if one group had higher ranks or if the ranks are evenly distributed etc

Question 4

Caclulate U values

Answer 4

Per condition:
[sum of all rank positions of this group]
-
[number of people in group x (number of ppl in group-1)
DIVIDED BY 2]

Question 5

N means

Answer 5

Number of participants in group

Question 6

R₂ means

Answer 6

ranks of position 2

Question 7

R₁ means

Answer 7

Sum of all Ranks of group 1

Question 8

Calculated value of U

Answer 8

Choose whichever U is lower from the 2 calculated U values from each condition

Question 9

Finding critical value of U

Answer 9

Refer to table of either one or two tailed hypothesis
If calc value is lower than crit at p<0.05: Choose table of critical values of best level of significance the calculated value of U beats critical value of U

Question 10

If calculated value of U is lower than critical value

Answer 10

Then at this level of significance the results are statistically significant

Question 11

How to remember to use this test

Answer 11

OIMU
Ordinal level data
Independent measures
Mann Whitney U test