Deck 7: Non-Parametric Models

Question 1

When the groups have unequal numbers of participants in them then the test statistic (Ws) for the Wilcoxon rank-sum test is:

A)The sum of the ranks for the two groups
B)The largest rank subtracted by the smallest rank
C)The sum of ranks in the group that contains the fewer people
D)The value of the smaller summed rank

Answer 1

C

Question 2

Which of the following non-parametric tests does not rely on the calculation of ranks?

A)Wilcoxon rank-sum test
B)Kruskal-Wallis test
C)Wilcoxon signed-rank test
D)They all rely on the calculation of ranks.

Answer 2

D

Question 3

The _______ test is based on calculating the difference between two sets of scores, making a note of the sign of the difference (positive or negative) and then ranking the differences from lowest to highest ignoring whether they are positive or negative.

A)Wilcoxon rank-sum test
B)Mann-Whitney U test
C)Dependent t
D)Wilcoxon signed-rank test

Answer 3

D

Question 4

You have carried out a Kruskal-Wallis test. There are significant differences between the three groups you are testing. How might you conduct your pairwise comparisons?

A)None of these.Post hoc analyses cannot be carried out using non-parametric tests.
B)Perform Mann-Whitney tests
C)Perform Wilcoxon rank-sum tests
D)Perform t-tests.

Question 5

What is the parametric equivalent of the Wilcoxon signed-rank test?

A)Pearson's r correlation
B)Independent ANOVA
C)The paired samples t-test
D)The independent t-test

Question 6

Which of the following tests would you use to assess whether there is a significant difference between the mean ranks of two independent conditions?

A)Wilcoxon rank-sum test
B)Kruskal-Wallis test
C)Friedman's test
D)Spearman's rho

Question 7

What symbol represents the test statistic for the Wilcoxon signed-rank test?

A)U
B)W_s
C)T
D)H

Question 8

What symbol represents the test statistic for the Kruskal-Wallis test?

A)W_s
B)U
C)H
D)T

Question 9

What can be done to reduce the inflated familywise error rate that occurs when performing multiple pairwise comparisons to follow up a significant Kruskal-Wallis test?

A)Use a lower probability as the threshold for significance.
B)Use a higher probability as our threshold for significance.
C)Use the Jonckheere-Terpstra test.
D)None of these are correct.

Question 10

A researcher measured people's physiological reactions to horror films. He split the data into two groups: males and females. The resulting data were significantly skewed and men and women had equal variances. What test should be used to analyse the data?

A)Wilcoxon signed-rank test
B)Mann-Whitney test
C)Paired-samples t-test
D)Independent t-test

Question 11

Imagine we wanted to investigate whether the reason for quitting smoking affects success rate. We recruited a sample of 120 smokers. Half wanted to quit to improve their appearance (yellow teeth and bad breath, etc.), and half wanted to quit to improve their health (life expectancy, etc.). After six months we measured how successful they had been at quitting smoking. The data were skewed, so we ran a Mann-Whitney test on the data (output below). Which of the following sentences correctly reports the results from the Mann-Whitney test?

A)People who wanted to quit smoking to improve their appearance did not significantly differ from those who wanted to quit to improve their health, U = 1415.50, z = −2.04, p = .04.
B)People who wanted to quit smoking to improve their appearance (M = 66.91) were significantly more successful at quitting than those who wanted to quit to improve their health (M = 54.09), U = −2.04, z = 188.57, p = .04.
C)People who wanted to quit smoking to improve their appearance were significantly more successful at quitting than those who wanted to quit to improve their health, U = 1415.50, z = −2.04, p = .04.
D)People who wanted to quit smoking to improve their health were significantly more successful at quitting than those who wanted to quit to improve their appearance, U = 1415.50, z = −2.04, p = .04.

Question 12

What is the problem with doing multiple pairwise comparisons to follow-up a significant Kruskal-Wallis test?

A)It decreases the significance of the Kruskal-Wallis test
B)None of these.
C)It inflates the familywise error rate.
D)It potentially masks important differences between variables.

Question 13

Non-parametric tests are used when:

A)The assumptions of parametric tests have not been met.
B)You want to increase the power of your experiment.
C)You have more than the maximum number of tied scores in your data set.
D)All of these.

Question 14

A researcher measured people's physiological reactions while watching horror films and compared them to when watching erotic films, and a documentary about wildlife. Different people viewed each type of film. The resulting data were skewed. What test should be used to analyse the data?

A)Kruskal-Wallis test
B)Independent analysis of variance
C)Repeated-measures analysis of variance
D)Friedman's ANOVA

Question 15

The _______ and _______ tests are calculated by ranking all of the scores from the two independent conditions that you are comparing from lowest to highest and adding up the ranks separately for each condition. The lowest of these summed ranks is the test statistic.

A)Friedman's ANOVA, Wilcoxon signed-rank
B)Mann-Whitney U, Wilcoxon rank-sum
C)Wilcoxon signed-rank, Mann-Whitney U
D)Wilcoxon signed-rank, Jonckheere-Terpstra

Question 16

What is the non-parametric equivalent of the independent t-test?

A)Friedman's ANOVA
B)Wilcoxon signed-rank test
C)Spearman's rho
D)Mann-Whitney U test

Question 17

What is the Jonckheere-Terpstra test used for?

A)To control for the familywise error rate.
B)To test for an ordered pattern to the medians of the groups you're comparing.
C)To test whether the variances in your data set are approximately equal.
D)To test for an ordered pattern to the means of the groups you're comparing.

Question 18

What is the Kruskal-Wallis test based upon?

A)Ranks
B)Means
C)Intervals
D)Residuals

Question 19

Imagine we wanted to investigate whether the reason for quitting smoking affects success rate. We recruited a sample of 120 smokers. Half wanted to quit to improve their appearance (yellow teeth and bad breath, etc.), and half wanted to quit to improve their health (life expectancy, etc.). After six months we measured how successful they had been at quitting smoking. The data were skewed so we ran a Wilcoxon rank-sum test on the data (output below). Which of the following sentences correctly reports the results from the Wilcoxon rank-sum test?

A)People who wanted to quit smoking to improve their health were significantly more successful at quitting than those who wanted to quit to improve their appearance, Ws = 3245.5, z = -2.04, p = .04.
B)People who wanted to quit smoking to improve their appearance were significantly more successful at quitting than those who wanted to quit to improve their health, T = -2.04, z = 188.57, p < .05.
C)This output relates to the Mann-Whitney test and tells us nothing about the Wilcoxon rank-sum test.
D)People who wanted to quit smoking to improve their appearance were significantly more successful at quitting than those who wanted to quit to improve their health, Ws = 3245.5, z = -2.04, p = .04.

Question 20

A psychologist was interested in whether there was a gender difference in the use of email. She hypothesized that because women are generally better communicators than men, they would spend longer using email than their male counterparts. To test this hypothesis, the researcher sat by the computers in her research methods laboratory and when someone started using email, she noted whether they were male or female and then timed how long they spent using email (in minutes). How should she analyse the differences in males and females (use the output below to help you decide)?

A)Paired t-test
B)Wilcoxon signed-rank test
C)Independent t-test
D)Mann-Whitney test

Question 21

When the groups have equal numbers of participants in them then the test statistic (Ws) for the Wilcoxon rank-sum test is:

A)The value of the smaller summed rank
B)The summed value of the ranks for the two groups
C)The largest rank subtracted by the smallest rank
D)The smallest rank subtracted by the largest rank