5. MANN WHITNEY AND WILCOXON

Question 1

What is the Mann-Whitney U test?

Answer 1

A non-parametric test comparing two independent groups.

Question 2

What is the Wilcoxon Signed-Rank test?

Answer 2

A non-parametric test comparing two related groups.

Question 3

When should you use the Mann-Whitney U test?

Answer 3

When comparing two independent groups with skewed or ordinal data.

Question 4

When should you use the Wilcoxon Signed-Rank test?

Answer 4

When comparing two related groups with skewed or ordinal data.

Question 5

What type of design is Mann-Whitney U used for?

Answer 5

Between-subjects design.

Question 6

What type of design is Wilcoxon Signed-Rank used for?

Answer 6

Within-subjects design.

Question 7

What is the null hypothesis for Mann-Whitney U?

Answer 7

The two groups have the same distribution.

Question 8

What is the null hypothesis for Wilcoxon Signed-Rank?

Answer 8

The median difference between the two conditions is zero.

Question 9

What kind of data do these tests use?

Answer 9

Ordinal or continuous data with skewed distribution.

Question 10

Why is Mann-Whitney U used with skewed data?

Answer 10

It ranks the data, reducing the impact of outliers.

Question 11

What does the Mann-Whitney U test compare?

Answer 11

The sum of ranks between two groups.

Question 12

What does the Wilcoxon Signed-Rank test compare?

Answer 12

The ranks of the differences between two conditions.

Question 13

What is a tied rank?

Answer 13

When two or more values have the same rank.

Question 14

What is the test statistic for Mann-Whitney U?

Answer 14

U

Question 15

What is the test statistic for Wilcoxon Signed-Rank?

Answer 15

V

Question 16

What is the effect size measure for these tests?

Answer 16

Vargha-Delaney’s A.

Question 17

What does Vargha-Delaney’s A measure?

Answer 17

The probability that one condition outperforms the other.

Question 18

What is the range of Vargha-Delaney’s A?

Answer 18

0 to 1

Question 19

What does A = 0.5 indicate?

Answer 19

No difference between the groups.

Question 20

What does A > 0.5 indicate?

Answer 20

The second condition outperforms the first.

Question 21

What does A < 0.5 indicate?

Answer 21

The first condition outperforms the second.

Question 22

What does a p-value < .05 indicate in these tests?

Answer 22

The difference between groups is statistically significant.

Question 23

What is the assumption for Mann-Whitney U?

Answer 23

Data is ordinal or non-normally distributed.

Question 24

What is the assumption for Wilcoxon Signed-Rank?

Answer 24

Data is ordinal or non-normally distributed.

Question 25

What is the advantage of using non-parametric tests?

Answer 25

They are robust to violations of normality.

Question 26

What is the limitation of non-parametric tests?

Answer 26

They have less statistical power than parametric tests.

Question 27

What is a Likert scale?

Answer 27

An ordinal scale used to measure attitudes or opinions.

Question 28

Why is the Likert scale considered ordinal?

Answer 28

The intervals between points are not necessarily equal.

Question 29

What is homogeneity of variance?

Answer 29

Equal spread of data across groups.

Question 30

Do non-parametric tests assume homogeneity of variance?

Answer 30

No, they do not require this assumption.

Question 31

What does a significant Mann-Whitney U test mean?

Answer 31

The distributions of the two groups are different.

Question 32

What does a significant Wilcoxon Signed-Rank test mean?

Answer 32

The median difference between conditions is significant.

Question 33

What does a non-significant Mann-Whitney U test mean?

Answer 33

No difference between the group distributions.

Question 34

What does a non-significant Wilcoxon test mean?

Answer 34

No median difference between conditions.

Question 35

What is the difference between wide and long data formats?

Answer 35

Wide format has multiple columns for conditions, long format has one.

Question 36

What is the effect of tied ranks on the test?

Answer 36

It reduces the precision of the test statistic.

Question 37

What is a quasi-experimental design?

Answer 37

A design without random assignment, often used in non-parametric tests.

Question 38

What does a large Vargha-Delaney’s A value indicate?

Answer 38

A large difference between the conditions.