Deck 15: More Nonparametric Tests

Question 1

The tests described in Chapter 15 are null hypothesis statistical testing techniques.

Answer 1

True

Question 2

With scores of 2, 2, 3, 3, 4, 5, a score of 3 has a rank of 3.

Answer 2

False

Question 3

Nonparametric tests assume that the populations being sampled from are normally distributed and have equal variances.

Answer 3

False

Question 4

Controversy exists over the relative merits of parametric and nonparametric tests with respect to power.

Question 5

The Wilcoxon-Wilcox multiple comparisons test is used for independent samples.

Question 6

The number of independent variables in a Wilcoxon matched-pairs signed-ranks T test is two.

Question 7

For the Wilcoxon matched-pairs signed-ranks T test, one difference of zero causes a pair of score to be dropped from the analysis.

Question 8

For large samples, the sampling distribution for the Mann-Whitney U test and the Wilcoxon matched-pairs signed-ranks T test is the normal curve.

Question 9

Spearman r_s values can be less than zero.

Question 10

Chapter 15 of your textbook explained how to predict individual scores using a nonparametric regression technique.

Question 11

The null hypotheses of parametric and nonparametric tests are identical.

Question 12

With scores of 1, 2, 3, 4, 4, 5, a score of 4 has a rank of 4.

Question 13

Nonparametric tests do not assume that the populations being sampled from are normally distributed nor that they have equal variances.

Question 14

Nonparametric tests can never exceed the power that parametric tests have.

Question 15

The Mann-Whitney U test is used for independent samples.

Question 16

The number of independent variables in a Wilcoxon matched-pairs signed-ranks T test is one.

Question 17

For a Spearman r_s, a difference of zero means that an error has been made.

Question 18

For large samples, the sampling distribution for the Mann-Whitney U test and the Wilcoxon matched-pairs signed-ranks T test is the F distribution.

Question 19

A Spearman r_s value of 1.96 is not possible.

Question 20

Chapter 15 of your textbook explained the calculation of three kinds of correlation coefficients.

Question 21

The tests described in Chapter 15 use sampling distributions based on ranks.

Question 22

With scores of 2, 2, 4, 4, 5, a score of 4 has a rank of 3.

Question 23

The null hypothesis in nonparametric tests is that the populations being sampled from have identical distributions.

Question 24

Nonparametric statistical tests do not require the assumptions about the populations that are required of parametric tests.

Question 25

The Mann-Whitney U test is used for paired samples.

Question 26

The number of independent variables in a Wilcoxon-Wilcox multiple comparisons test is one.

Question 27

For a Spearman r_s, a difference of zero causes a pair of score to be dropped from the analysis.

Question 28

For large samples, the sampling distribution for the Mann-Whitney U test and the Wilcoxon matched-pairs signed-ranks T test is the chi square distribution.

Question 29

A Spearman r_s value is never greater than 1.00.

Question 30

Chapter 15 of your textbook explained how to analyze two-group designs that have two dependent variables.

Question 31

Data Set 15-1: The scores below are ranks. They were obtained by taking each raw score and giving it a rank of 1 to 4.
$$\begin{array} { c c } \text { Group 1 } & \text { Group 2 } \\\hline 1 & 3 \\2 & 4\end{array}$$

-Given what has already been done to the data in Data Set 15-1, the most likely analysis would be a

A) Mann-Whitney test
B) Wilcoxon matched-pairs signed ranks test
C) Spearman's r_s
D) unknown not enough information is given.

Question 32

Data Set 15-1: The scores below are ranks. They were obtained by taking each raw score and giving it a rank of 1 to 4.
$$\begin{array} { c c } \text { Group 1 } & \text { Group 2 } \\\hline 1 & 3 \\2 & 4\end{array}$$

-Refer to Data Set 15-1. The rationale of nonparametric tests says that if the null hypothesis is true,

A) the mean score of the population from which Group 1 comes will be equal to that ofGroup 2
B) the mean score of the population from which Group 1 comes will not be equal to that ofGroup 2
C) the sum of the ranks of those in Group 1 will be equal to the sum of those in Group 2
D) the sum of the ranks of those in Group 1 will not be equal to the sum of those in Group 2.

Question 33

The date that is closest to the dates when nonparametric tests were introduced is

A) 1700
B) 1800
C) 1900
D) 1950.

Question 34

The nonparametric tests in the text are based on sampling distributions of

A) means
B) mean differences
C) ranks
D) variances.

Question 35

The thing that all four nonparametric methods in your text have in common is that the statistic analyzed is a

A) mean
B) median
C) raw score
D) rank score.

Question 36

According to the rationale for nonparametric tests, the null hypothesis is that the sum of the ranks of Population A and Population B will be

A) equal
B) unequal, with A > B
C) unequal, with B > A
D) unequal, with the direction being chosen by the researcher.

Question 37

According to the rationale for nonparametric tests, the null hypothesis says that the sum of the ranks of Population A will be _______ to the sum of the ranks of Population B.

A) equal
B) unequal, with A < B
C) unequal, with B > A
D) unequal, with the direction being chosen by the researcher.

Question 38

The basic rationale of nonparametric statistics is that an overall ranking of samples

A) all drawn from the same population will produce sums of sample ranks that are equal
B) all drawn from the same population will produce sums of sample ranks that are unequal
C) all drawn from different populations will produce sums of sample ranks that are equal
D) all drawn from different populations will produce sums of sample ranks that are unequal and the differences can be determined mathematically.

Question 39

The null hypothesis for nonparametric tests is that the population

A) means are identical
B) variances are identical
C) distributions are identical
D) none of the descriptive alternatives are correct.

Question 40

The null hypothesis for the Mann-Whitney U test is that the

A) sample means are the same
B) population means are the same
C) sample distributions are the same
D) population distributions are the same.

Question 41

The null hypothesis for the Wilcoxon matched-pairs signed-ranks T test is that the

A) sample means are the same
B) sample distributions are the same
C) population distributions are the same
D) population means are the same.

Question 42

The null hypothesis for testing the significance of r_s is that the population correlation coefficient is

A) .00
B) 1.00
C) the statistic, r_s, calculated from the sample data
D) none of the descriptive alternatives are correct.

Question 43

Your text discussed two issues for which hard and fast rules don't exist on the choice of para-metric or nonparametric tests. The issues were

A) random assignment and scales of measurement
B) sample size and power
C) random assignment and sample size
D) scales of measurement and power.

Question 44

If the populations being compared are normally distributed and have equal variances, the most powerful statistical test is a test.

A) descriptive
B) parametric
C) nonparametric
D) orthogonal.

Question 45

Parametric and nonparametric tests differ from each other when it comes to

A) whether a null hypothesis is tested
B) whether a sampling distribution is used
C) which test has more power in a particular situation
D) all of the descriptive alternatives are correct.

Question 46

Power is the likelihood of

A) rejecting H₀ when it is true
B) rejecting H₀ when it is false
C) retaining H₀ when it is true
D) retaining H₀ when it is false.

Question 47

The more powerful test is the one that is more likely to

A) retain H₀ when it is true
B) reject H₀ when it is false
C) both descriptive alternatives are correct
D) neither descriptive alternatives is correct.

Question 48

The likelihood of retaining the null hypothesis when it is false is most closely associated with

A) power
B) Type I error
C) nonparametric tests rather than parametric tests
D) Mann-Whitney U rather than Wilcoxon matched-pairs signed-ranks T.

Question 49

Which of the terms below does not go with nonparametric statistics?

A) normally distributed dependent variable
B) hypothesis testing
C) can be used to analyze small data sets
D) assume that samples are randomly assigned.

Question 50

In comparing parametric and nonparametric tests, which of the following is(are) the same for both tests?

A) require random assignment
B) based on the logic of testing a null hypothesis
C) both of the descriptive alternatives are correct
D) neither of the descriptive alternatives is correct.

Question 51

If the populations that are being tested are normally distributed and have equal variances,

A) parametric tests have more power
B) nonparametric tests have more power
C) unknown this issue has not yet been resolved.

Question 52

require that the variances be equal for the dependent variable scores of the populations in question.

A) parametric tests
B) nonparametric
C) both of the descriptive alternatives are correct
D) neither of the descriptive alternatives are correct.

Question 53

On the question of when to use a parametric test and when to use a nonparametric test, your text discussed two issues. (This discussion has not yet resulted in a recommendation.) The two issues were

A) null hypothesis and hypothesis testing
B) scale of measurement and power
C) random assignment and alternative hypotheses
D) independent data and correlated data.

Question 54

Nonparametric tests are used rather than a t test or an ANOVA when

A) the researcher does not know the specific value of the population parameters
B) the data are in the form of ranks
C) the assumption of random assignment is not justified
D) a null hypothesis cannot be specified.

Question 55

Controversy over the relative merits of parametric and nonparametric tests exists with respect to

A) whether to use a random sample
B) power
C) both whether to use a random sample and power
D) neither whether to use a random sample nor power.

Question 56

require that the dependent variable scores be normally distributed.

A) parametric tests
B) nonparametric tests
C) both of the descriptive alternatives are correct
D) neither of the descriptive alternatives is correct.

Question 57

Both parametric and nonparametric tests require you to assume that the dependent variable scores are

A) normally distributed
B) have equal variances
C) both of the descriptive alternatives are correct
D) neither of the descriptive alternatives is correct.

Question 58

are based on the idea that subjects are randomly assigned to groups.

A) parametric tests
B) nonparametric tests
C) both of the descriptive alternatives are correct
D) neither of the descriptive alternatives is correct

Question 59

The Wilcoxon matched-pairs signed-ranks T test is most similar to

A) ANOVA followed by HSD tests
B) independent-samples t test
C) paired-samples t test
D) chi square.

Question 60

Which of the following statistical tests is most similar to the Mann-Whitney U test?

A) t test
B) ANOVA
C) $\chi$ ² test
D) r_s.

Question 61

The descriptive (rather than inferential) statistic(s) in Chapter 15 is (are) the

A) Mann-Whitney U test
B) Wilcoxon matched-pairs signed-ranks T test
C) both of the descriptive alternatives are correct
D) neither of the descriptive alternatives is correct.

Question 62

Wilcoxon-Wilcox multiple comparisons test is to Wilcoxon matched-pairs signed-ranks T test as

A) Tukey HSD tests are to paired-samples t test
B) t tests are to F tests
C) paired samples are to independent samples
D) a priori is to post hoc.

Question 63

Mann-Whitney U is to Wilcoxon matched-pairs signed-ranks T asis to

A) independent-samples t is to paired-samples t
B) paired-samples t is to independent-samples t
C) random samples is to biased samples
D) biased samples is to random samples.

Question 64

The Mann-Whitney U test is to the Wilcoxon matched-pairs signed-ranks T as

A) retain is to reject
B) r is to F
C) a priori is to post hoc
D) independent-samples t is to paired-samples t.

Question 65

The number of independent variables in the Wilcoxon-Wilcox multiple comparisons test is

A) one
B) two
C) two or more.

Question 66

Suppose 25 cardinals and 25 bluejays were ranked for resistance to air pollution with the rank of one going to the most resistant bird. The smaller sum of ranks, 738, was for cardinals. The conclusion from these data is that

A) a Wilcoxon matched-pairs signed-ranks test is appropriate
B) cardinals are significantly more resistant than bluejays
C) bluejays are significantly more resistant than cardinals
D) bluejays and cardinals are not significantly different on resistance to air pollution.

Question 67

Suppose a Wilcoxon matched-pairs signed-ranks test with N = 60 produced a sum of thenegative ranks of 643. A z value of leads to the conclusion that the null hypothesisshould be .

A) 2.00, rejected at the .05 level
B) 0.02, retained
C) 7.82, rejected at the .001 level

Question 68

A value of 0 is obtained for U when

A) the samples are identical
B) the samples are identical, on the average
C) an arithmetic mistake has been made
D) none of the descriptive alternatives are correct

Question 69

The _____ the value of U and the ____ the value of T, the more likely to reject the null hypothesis.

A) larger, larger
B) larger, smaller
C) smaller, smaller
D) smaller, larger.

Question 70

Given the scores: 7, 7, 8, 8, 8, 9, a score of 8 is given the rank

A) 8
B) 3.5
C) 3
D) 4

Question 71

What rank does a score of 3 receive? Scores: (score of 1 ranks 1) 1, 1, 2, 2, 3, 3, 3

A) 2
B) 3
C) 4
D) 5
E) 6

Question 72

For the numbers 1, 2, 3, 3, 3, 4, 4, 4, 4, 5, the rank for 4 is (score of 1 ranks 1)

A) 7
B) 7.5
C) 8
D) none of the descriptive alternatives are correct.

Question 73

What rank would a score of 4 have in the following distribution? 1, 2, 3, 4, 4, 4, 4, 5

A) 4
B) 4.5
C) 5
D) 5.5

Question 74

What rank would a score of 4 have in the following distribution? 1, 2, 3, 4, 4, 4, 5, 5

A) 4
B) 5
C) 5.5
D) 6

Question 75

What rank does a score of 5 receive? Scores: (score of 1 ranks 1) 1, 2, 3, 4, 4, 4, 4, 5, 5

A) 5
B) 8
C) 8.5

Question 76

What is the rank of 4 in the following distribution? 1, 2, 2, 3, 3, 4, 4, 5, 5

A) 4
B) 5.5
C) 6
D) 7
E) 6.5

Question 77

What rank would a score of 4 have in the following distribution? 1, 2, 3, 3, 3, 4, 4, 4, 6

A) 4
B) 6
C) 7
D) 7.5

Question 78

What rank would a score of 4 have in the following distribution? 1, 2, 3, 3, 4, 4, 5, 5, 5

A) 4
B) 4.5
C) 5
D) 6

Question 79

The normal distribution is used as a sampling distribution for

A) large sample Spearman r_s
B) large sample Mann-Whitney U
C) small sample Wilcoxon matched-pairs signed-ranks T
D) none of the descriptive alternatives are correct.

Question 80

Nonparametric methods for large sample tests use the distribution to provide probability figures.

A) t
B) F
C) normal
D) U.