Deck 7: Non-Parametric Models

Question 1

View the anxiety data (Sport Competitive Anxiety Test - SCAT) collected from a group of middle-distance runners. All the runners completed the questionnaire before being split into either a control group (who received no imagery intervention) or an experimental group (who received an imagery intervention). The anxiety scores were then measured again after 6 weeks. If we wanted to compare the control group results with the experimental group results, what type of research design would we use?

A) Independent-groups design
B) Repeated-measures design
C) Dependent-groups design
D) Matched-pairs design

Answer 1

A

Question 2

Analyse the group data by using a Kolmogorov-Smirnov test. Which data sets are not normally distributed ?

A) Control group pre-test score and experimental group post-test score
B) Control group post-test score and experimental group pre-test score
C) Control group pre-test score and experimental group pre-test score
D) Control group post test score and experimental group post-test score

Answer 2

A

Question 3

What factors are likely to influence the effect size? (You may select more than one.)

A) Sample size
B) Significance
C) Population demographic
D) Sampling method

Answer 3

A AND B

Question 4

Which of the following descriptions best explains non-parametric tests?

A) Assumption-free
B) More likely to protect against Type II error
C) Robust
D) Normally distributed

Question 5

Calculate the effect size for the test conducted on the experimental group data (Q13).

A) r = -.82
B) r = .72
C) r = -.62
D) r = .52

Question 6

See the week 12 data in Q5 for the imagery group. Run an appropriate test to examine the variance between weeks 0, 6 and 12. Which is the correct answer?

A) χ²(2) = 26.96, p < .001
B) χ²(2) = 26.96, p = .000
C) χ²(2) = 0.001, p < 26.96
D) χ²(2) = 26.96, p < .005

Question 7

Which of the following are non-parametric tests?

A) Willcoxon test
B) Mann-Whitney test
C) Kruskal-Wallis test
D) All of the above

Question 8

Examine the following Q-Q plot from the pre-test control group. What can be assumed from the graph?

A) Data may not be normally distributed because the points deviate from the diagonal line.
B) Data are definitely not normally distributed because the points deviate from the diagonal line.
C) Data may be normally distributed because the points deviate from the diagonal line.
D) Data are definitely normally distributed because the points deviate from the diagonal line.

Question 9

If we wanted to compare the pre-test results with the post-test results, what type of research design would we use?

A) Independent-groups design
B) Repeated-measures design
C) Mixed model groups design
D) Between-groups design

Question 10

What were the effect sizes for the between group pre and post tests (see results from Q18)?

A) r = .11 and r = .35
B) r = .15 and r = .42
C) r = .21 and r = .30
D) r = .25 and r = .30

Question 11

Suppose you were interested in the longer-term effects of the imagery interventions and so you decide to get the experimental group to measure their anxiety scores using the SCAT questionnaire again, but this time 12 weeks later. What type of test might you use if you were interested in examining the variance at 0, 6 and 12 weeks?

A) Kruskal-Wallis test
B) Friedman's test
C) Sign test
D) McNemar's test

Question 12

Which test can be considered the non-parametric equivalent to an independent t-test?

A) Willcoxon test
B) Mann-Whitney test
C) Kruskal-Wallis test
D) Friedman's ANOVA

Question 13

Run an appropriate test between the pre-test and post-test data of the experimental group. What is the result of the test?

A) W_s = 1.5, z = -4.514, p < .001
B) W_s = -4.514, z = 1.5, p < .001
C) W_s = 1.5, z = -4.514, p = .000
D) W_s = -4.514, z = 1.5, p = .000

Question 14

When using non-parametric tests, which measure of central tendency is preferred when summarizing a set of data?

A) Median
B) Mean
C) Mode
D) None of the above are appropriate.

Question 15

Based on the answer to Q13, what should you do?

A) Reject the null hypothesis.
B) Accept the null hypothesis.
C) Reject the research hypothesis.
D) None of the answers provided are acceptable.

Question 16

Based on the results of the initial analysis, what would be the most appropriate test to run if you wanted to compare the pre-test results with the post-test results of each group (i.e. control group pre versus post test or experimental pre versus post-test)?

A) Willcoxon test
B) Mann-Whitney test
C) Paired-samples t-test
D) Independent t-test

Question 17

Suppose you wanted to see if there was a difference between the participants in the control group and the experimental group for the pre-test and the post-test data. What is the correct expression for the test results?

A) U = 506, z = -0.830, p = .406 and U = 269.5, z = -2.682, p = .007
B) U = -0.830, z = 394, p = .406 and U = -2.682, z = 269.5, p = .007
C) U = 0.406, z = -0.830, p = 394 and U = 0.007, z = -2.682, p = 269.5
D) U = 394, z = 0.406, p = -.830 and U = 269.5, z = 0.007, p = -2.682

Question 18

If we wanted to compare the control group results with the experimental group results for the pre and post conditions, what type of research design would we use?

A) Independent-groups design
B) Repeated-measures design
C) Mixed model groups design
D) Within-groups design

Question 19

Run an appropriate test between the pre-test and post-test data of the control group. What is the result of the test?

A) W_s = 163.5, z = 1.205, p = .228
B) W_s = 1.205, z = 163.5, p = .228
C) W_s = .228, z = 1.205, p = 163.5
D) W_s = 163.5, z = .228, p = 1.205

Question 20

Which is the correct expression for the Kolmogorov-Smirnov test result for the control group pre-test score?

A) D(30) = .181, p = .014
B) D(30) = .140, p = .140
C) D(30) = .122, p > .2
D) D(30) = .203, p = .003

Question 21

Examine the following Q-Q plot from the pre-test control group. What can be assumed from the graph?

A) Data may not be normally distributed because the points deviate from the diagonal line.
B) Data are definitely not normally distributed because the points deviate from the diagonal line.
C) Data may be normally distributed because the points deviate from the diagonal line.
D) Data are definitely normally distributed because the points deviate from the diagonal line.

Question 22

Based on the answer to Q13, what should you do?

A) Reject the null hypothesis.
B) Accept the null hypothesis.
C) Reject the research hypothesis.
D) None of the answers provided are acceptable.

Question 23

Suppose you wanted to see if there was a difference between the participants in the control group and the experimental group for the pre-test and the post-test data. What is the correct expression for the test results?

A) U = 506, z = -0.830, p = .406 and U = 269.5, z = -2.682, p = .007
B) U = -0.830, z = 394, p = .406 and U = -2.682, z = 269.5, p = .007
C) U = 0.406, z = -0.830, p = 394 and U = 0.007, z = -2.682, p = 269.5
D) U = 394, z = 0.406, p = -.830 and U = 269.5, z = 0.007, p = -2.682

Question 24

If we wanted to compare the control group results with the experimental group results for the pre and post conditions, what type of research design would we use?

A) Independent-groups design
B) Repeated-measures design
C) Mixed model groups design
D) Within-groups design

Question 25

What is the parametric equivalent of a Kruskal-Wallis test?

A) One-way ANOVA
B) Repeated-measures ANOVA
C) Two-way ANOVA
D) Three-way ANOVA

Question 26

Which of the following descriptions best explains non-parametric tests?

A) Assumption-free
B) More likely to protect against Type II error
C) Robust
D) Normally distributed

Question 27

Which of the following are non-parametric tests?

A) Willcoxon test
B) Mann-Whitney test
C) Kruskal-Wallis test
D) All of the above

Question 28

Which is the correct expression for the Kolmogorov-Smirnov test result for the control group pre-test score?

A) D(30) = .181, p = .014
B) D(30) = .140, p = .140
C) D(30) = .122, p > .2
D) D(30) = .203, p = .003

Question 29

Where were differences between the data sets? (You may select more than one option.)

A) Pre and Post
B) Pre and Post + 12
C) Post and Post + 12
D) No differences were identified.

Question 30

Run an appropriate test between the pre-test and post-test data of the control group. What is the result of the test?

A) W_s = 163.5, z = 1.205, p = .228
B) W_s = 1.205, z = 163.5, p = .228
C) W_s = .228, z = 1.205, p = 163.5
D) W_s = 163.5, z = .228, p = 1.205

Question 31

If we wanted to compare the pre-test results with the post-test results, what type of research design would we use?

A) Independent-groups design
B) Repeated-measures design
C) Mixed model groups design
D) Between-groups design

Question 32

Calculate the effect size for the test conducted on the experimental group data (Q13).

A) r = -.82
B) r = .72
C) r = -.62
D) r = .52

Question 33

Based on the results of the initial analysis, what would be the most appropriate test to run if you wanted to compare the pre-test results with the post-test results of each group (i.e. control group pre versus post test or experimental pre versus post-test)?

A) Willcoxon test
B) Mann-Whitney test
C) Paired-samples t-test
D) Independent t-test

Question 34

Run an appropriate test between the pre-test and post-test data of the experimental group. What is the result of the test?

A) W_s = 1.5, z = -4.514, p < .001
B) W_s = -4.514, z = 1.5, p < .001
C) W_s = 1.5, z = -4.514, p = .000
D) W_s = -4.514, z = 1.5, p = .000

Question 35

Which test can be considered the non-parametric equivalent to an independent t-test?

A) Willcoxon test
B) Mann-Whitney test
C) Kruskal-Wallis test
D) Friedman's ANOVA

Question 36

Since there are three sets of data, how will the differences within the data sets be identified?

A) Multiple Wilcoxon tests
B) Multiple Mann-Whitney tests
C) Multiple sign tests
D) None of the tests are appropriate.

Question 37

When using non-parametric tests, which measure of central tendency is preferred when summarizing a set of data?

A) Median
B) Mean
C) Mode
D) None of the above are appropriate.

Question 38

Analyse the group data by using a Kolmogorov-Smirnov test. Which data sets are not normally distributed ?

A) Control group pre-test score and experimental group post-test score
B) Control group post-test score and experimental group pre-test score
C) Control group pre-test score and experimental group pre-test score
D) Control group post test score and experimental group post-test score

Question 39

What factors are likely to influence the effect size? (You may select more than one.)

A) Sample size
B) Significance
C) Population demographic
D) Sampling method

Question 40

View the anxiety data (Sport Competitive Anxiety Test - SCAT) collected from a group of middle-distance runners. All the runners completed the questionnaire before being split into either a control group (who received no imagery intervention) or an experimental group (who received an imagery intervention). The anxiety scores were then measured again after 6 weeks. If we wanted to compare the control group results with the experimental group results, what type of research design would we use?

A) Independent-groups design
B) Repeated-measures design
C) Dependent-groups design
D) Matched-pairs design

Question 41

Where were differences between the data sets? (You may select more than one option.)

A) Pre and Post
B) Pre and Post + 12
C) Post and Post + 12
D) No differences were identified.

Question 42

See the week 12 data in Q5 for the imagery group. Run an appropriate test to examine the variance between weeks 0, 6 and 12. Which is the correct answer?

A) χ²(2) = 26.96, p < .001
B) χ²(2) = 26.96, p = .000
C) χ²(2) = 0.001, p < 26.96
D) χ²(2) = 26.96, p < .005

Question 43

What were the effect sizes for the between group pre and post tests (see results from Q18)?

A) r = .11 and r = .35
B) r = .15 and r = .42
C) r = .21 and r = .30
D) r = .25 and r = .30

Question 44

Suppose you were interested in the longer-term effects of the imagery interventions and so you decide to get the experimental group to measure their anxiety scores using the SCAT questionnaire again, but this time 12 weeks later. What type of test might you use if you were interested in examining the variance at 0, 6 and 12 weeks?

A) Kruskal-Wallis test
B) Friedman's test
C) Sign test
D) McNemar's test

Question 45

What is the parametric equivalent of a Friedman's test?

A) Repeated-measures ANOVA
B) One-way ANOVA
C) Two-way ANOVA
D) Three-way ANOVA

Question 46

What is the parametric equivalent of a Kruskal-Wallis test?

A) One-way ANOVA
B) Repeated-measures ANOVA
C) Two-way ANOVA
D) Three-way ANOVA

Question 47

Since there are three sets of data, how will the differences within the data sets be identified?

A) Multiple Wilcoxon tests
B) Multiple Mann-Whitney tests
C) Multiple sign tests
D) None of the tests are appropriate.