Deck 18: Nonparametric Tests: Tests for Ordinal Data

Question 1

Which of the following gives the best explanation for why it would be advantageous to convert the following scores to ordinal data?

A) because the data are normally distributed
B) because the data are split into two groups
C) because there is an outlier in the data for Group 2
D) because parametric tests are always preferred to nonparametric tests

Answer 1

because there is an outlier in the data for Group 2

Question 2

A researcher finds that the number of positive ranks is 13 and the number of negative ranks is 4 using a sample of 17 participants.Based on these data,what is the decision for a related-samples sign test at a .05 level of significance?

A) retain the null hypothesis
B) reject the null hypothesis
C) there is not enough information to answer this question

Answer 2

reject the null hypothesis

Question 3

For the Wilcoxon signed-ranks T,when T is less than the critical value,we decide to ______.

A) retain the null hypothesis
B) reject the null hypothesis
C) not enough information

Answer 3

reject the null hypothesis

Question 4

Tests for ordinal data meaningfully convey information regarding ______.

A) variance
B) differences
C) variance and differences

Question 5

A researcher computes the Wilcoxon signed-ranks T test.She finds that the sum of positive ranks is 56 and the sum of negative ranks is 22.What is the value of the test statistic,T?

A) 22
B) 34
C) 56
D) 78

Question 6

A researcher computes the Wilcoxon signed-ranks T test.She computes T = 68 using a sample of 24 participants.What is the decision for a two-tailed test at a .05 level of significance?

A) retain the null hypothesis
B) reject the null hypothesis
C) there is not enough information to answer this question

Question 7

The key advantage of measuring ranked data is that it minimizes ______ caused by outliers.

A) significance
B) effect size
C) variability
D) all of these

Question 8

A researcher computes a one-sample sign test in which the number of pluses is 8,the number of minuses is 2,and the number of scores equal to the median is 1.The test statistic (x)for the sign test is equal to ______.

A) 1
B) 2
C) 8
D) there is not enough information to answer this question

Question 9

Tests for ordinal data can ______ the power to detect an effect when data are skewed.

A) decrease
B) increase
C) have no effect on
D) categorize

Question 10

Which of the following is not a step to average tied ranks?

A) compute the mean of the tied ranks
B) assume the ranks are in order
C) rank each score by its effect size
D) shift the remaining ranks accordingly

Question 11

A researcher finds that the number of scores above the median is 11 and the number of scores below the median is 5,using a sample of 16 participants.Based on these data,what is the decision for a one-sample sign test at a .05 level of significance?

A) retain the null hypothesis
B) reject the null hypothesis
C) there is not enough information to answer this question

Question 12

A researcher computes a related-samples sign test in which the number of positive ranks is 9 and the number of negative ranks is 3.The test statistic (x)is equal to ______.

A) 3
B) 9
C) 12
D) There is not enough information to answer this question.

Question 13

The sign test is used as a nonparametric alternative to which parametric test?

A) one-sample t test
B) two-independent-sample t test
C) related-samples t test
D) one-sample t test and related-samples t test

Question 14

Tests for ordinal data ______.

A) are nonparametric tests
B) do not measure variance to test for significance
C) minimize problems associated with having outliers in a data set
D) all of these

Question 15

Which of the following scales of measurement are analyzed using a nonparametric test?

A) interval and ratio data
B) ordinal and interval data
C) nominal and ordinal data
D) ordinal and ratio data

Question 16

The Wilcoxon signed-ranks T test is used as a nonparametric alternative to which parametric test?

A) one-sample t test
B) two-independent-sample t test
C) related-samples t test
D) one-sample t test and related-samples t test

Question 17

Which of the following best explains how to find the test statistic for the one-sample sign test?

A) Scores above the median are assigned a plus sign (+); scores below the median are assigned a negative sign (-); the test statistic is the number of pluses or minuses, whichever occurs most often.
B) Scores are correlated as either positive (+) or negative (-); the test statistic is the extent to which scores are positively or negatively correlated.
C) Scores are coded as being consistent (+) or inconsistent (-) with the null hypothesis; the test statistic is the number of pluses or minuses, whichever occurs most often.
D) The variability of scores is compared with the variance stated by the null hypothesis; larger variability is considered a sign that the data are significant.

Question 18

Each of the following is an appropriate test for ordinal data,except ______.

A) the Mann-Whitney U test
B) the chi-square goodness-of-fit test
C) the one-sample sign test
D) the Friedman test

Question 19

The test statistic (x)for the sign test is equal to ______.

A) variability attributed to the groups divided by variability attributed to error
B) the number of pluses or minuses, whichever occurs most often
C) the number of scores that are less than or equal to the median value
D) the number of positive scores divided by the number of negative scores

Question 20

Which of the following best explains how to interpret the test statistic for the Wilcoxon signed-ranks T?

A) The test statistic is interpreted the same as that for the related-samples t test.
B) The larger the value of T, the more likely we are to reject the null hypothesis.
C) The smaller the value of T, the more likely we are to reject the null hypothesis.
D) We reject the null hypothesis when T is less than .05.

Question 21

For the Mann-Whitney U test,we reject the null hypothesis when ______.

A) values of U are greater than or equal to the critical value
B) values of U are less than or equal to .05
C) values of U are less than or equal to the critical value
D) values of U are closer to ± 1.0

Question 22

The test statistic for the Friedman test is largest when ______.

A) the ranks for each participant are consistent across groups
B) the mean difference between groups is largest
C) the ranks for each participant are not consistent across groups
D) the value of the test statistic is less than or equal to .05

Question 23

The Friedman test is used as a nonparametric alternative to which parametric test?

A) related-samples t test
B) two-independent-sample t test
C) one-way between-subjects ANOVA
D) one-way within-subjects ANOVA

Question 24

What is the null hypothesis for a Kruskal-Wallis H test?

A) The number of scores above and below the median are equal.
B) The median differences between two groups are equal.
C) The sum of ranks in each group does not differ between groups.
D) The ranks in two groups are equal to 0.

Question 25

A professor ranks the grades of students in each of three sections of a statistics course.He computes H = 6.83 for the Kruskal-Wallis H test to test for differences between the sections.What is the decision for this test?

A) retain the null hypothesis
B) reject the null hypothesis
C) there is not enough information to answer this question

Question 26

Which of the following dispersions of scores for three groups (A,B,and C)best illustrate the null hypothesis for a Kruskal-Wallis H test?

A) The number of scores above and below the median are equal.
B) The mean differences between two groups are equal.
C) The sum of ranks in each group does not differ between groups.
D) The ranks in two groups are equal to 0.

Question 27

The Mann-Whitney U test is used as a nonparametric alternative to which parametric test?

A) one-sample t test
B) two-independent-sample t test
C) related-samples t test
D) one-sample t test and related-samples t test

Question 28

To compute the test statistic for the Mann-Whitney U test,we sum the ranks in two groups.The test statistic is ______.

A) the larger total
B) the smaller total
C) the average of the two totals
D) the product of the two totals

Question 29

A researcher compares the ranks in four groups and computes H = 6.55 for the Kruskal-Wallis H test.What is the decision for this test?

A) retain the null hypothesis
B) reject the null hypothesis
C) there is not enough information to answer this question

Question 30

Each of the following hypothesis tests has a normal approximation formula,except for the ______.

A) Mann-Whitney U test
B) related-samples sign test
C) Wilcoxon signed-ranks T test
D) Friedman test

Question 31

A researcher computes and .What is the value of the test statistic for a Mann-Whitney U test?

A) 17
B) 38
C) There is not enough information to answer this question.

Question 32

When the null hypothesis is true and n is greater than or equal to 5 per group,the test statistic for the Friedman test is associated with ______ degrees of freedom.

A) no
B) n - 1
C) k - 1
D) .05

Question 33

We compare the test statistic of a Kruskal-Wallis H test to the critical value given in the table for which hypothesis test?

A) the analysis of variance
B) the t test
C) the chi-square test
D) the Pearson correlation

Question 34

The Friedman test compares differences in the sum of ranks in two or more ______ groups.

A) independent
B) related
C) significant
D) mature

Question 35

Which of the following dispersions of scores for two groups,A and B,best illustrates the null hypothesis for a Mann-Whitney U test?

A) A, B, A, B, A, B, A, B
B) A, A, B, B, A, B, B, A
C) B, A, B, A, A, B, A, B
D) A, A, A, A, B, B, B, B

Question 36

The Kruskal-Wallis H test is used as a nonparametric alternative to which parametric test?

A) related-samples t test
B) two-independent-sample t test
C) one-way between-subjects ANOVA
D) one-way within-subjects ANOVA

Question 37

Which of the following nonparametric tests can be computed using a normal appropriation formula?

A) related-samples sign test
B) Wilcoxon signed-ranks T test
C) Mann-Whitney U test
D) all of these

Question 38

What is the null hypothesis for a Mann-Whitney U test?

A) The number of scores above and below the median are equal.
B) The ranks in two groups are equally dispersed.
C) The mean differences between two groups are equal.
D) The ranks in two groups are equal to 0.

Question 39

When the null hypothesis is true and n is greater than or equal to 5 per group,the test statistic for the Friedman test is approximately distributed as a ______.

A) t distribution
B) F distribution
C) normal distribution
D) chi-square distribution

Question 40

For the Friedman test,when is less than the critical value,we decide to ______.

A) retain the null hypothesis
B) reject the null hypothesis
C) not enough information

Question 41

For a two-sample sign test,count the number of positive and negative difference scores.The larger number is the test statistic.

Question 42

Ordinal data indicate that one score is larger or smaller than another.

Question 43

When the test statistic for the Mann-Whitney U test is equal to or smaller than the critical value,we reject the null hypothesis.

Question 44

Nonparametric tests are associated with less power to detect an effect when the data are skewed in the population.

Question 45

When the test statistic for the Kruskal-Wallis H test is equal to or smaller than the critical value,we reject the null hypothesis.

Question 46

Variance can't be used to analyze ordinal scale data,because these data do not meaningfully convey distance.

Question 47

The null hypothesis for the Friedman test is that the sum of ranks in each group does not differ between groups.

Question 48

When the null hypothesis is true and n is greater than or equal to 5 per group,the test statistic is approximately distributed as a chi-square distribution.

Question 49

The null hypothesis for the Kruskal-Wallis H test is that scores above and below the median are equal.

Question 50

The one-sample sign test can be used for ordinal data in place of the one-sample t test.

Question 51

The Wilcoxon signed-ranks T is a statistical procedure used to determine whether the total ranks in two related groups are significantly different.

Question 52

When the test statistic for the Friedman test is equal to or smaller than the critical value,we reject the null hypothesis.

Question 53

The sign test can be used as an alternative to the one-sample t test and the two-independent-sample t test.

Question 54

Tests for ordinal data are regarded as nonparametric tests.

Question 55

Nonparametric tests are used when the measured data are on an ordinal scale of measurement.

Question 56

A researcher computes Wilcoxon T = 12 and finds that this value is larger than the critical value.The decision is to reject the null hypothesis.

Question 57

When the null hypothesis is true and n is greater than or equal to 5 per group,the test statistic H is approximately distributed as a chi-square distribution.

Question 58

For a Mann-Whitney U test,the more lopsided the dispersion of scores between groups,the more likely we are to reject the null hypothesis.

Question 59

For a Wilcoxon signed-ranks T test,we decide to reject the null hypothesis when T is larger than the critical value.

Question 60

The Mann-Whitney U is a statistical procedure used to determine whether the dispersion of ranks in two or more independent groups is equal.