Deck 10: A Synopsis of Selected Statistical Tests for Examining Ordinal Data

Question 1

Spearman rank correlation coefficient provides you with a coefficient for data measured at the:

A) nominal level
B) interval level
C) ratio level
D) ordinal level

Answer 1

D

Question 2

When you have an ordinal variable such as order of preferences for an event and another variable such as gender that is dichotomous from independent samples, you will choose which one of the following for your statistical analysis?

A) the one-sample t Test
B) Pearson's product moment correlation
C) the Mann-Whitney U Test
D) Spearman rank correlation coefficient

Answer 2

C

Question 3

When you have a variable such as birth order ranked and the other is nominal such as ethnic background that has more than two categories with the data drawn from independent samples, you will employ the:

A) the one-sample t Test
B) Pearson's product moment correlation
C) the Mann-Whitney U test
D) Kruskal-Wallis one-way analysis of variance

Answer 3

D

Question 4

Spearman rank correlation generates a coefficient:

A) that can range from 1-2 and can be either positive or negative
B) that can range from 0 to 1.0 and can be either positive or negative
C) that can range from 0 to 1.0 and cannot be negative
D) that can range from 0 to -1.0 and cannot be positive

Question 5

The Spearman correlation coefficient can be used in an analysis of an interval level variable and an ordinal variable, the ideal application is for a study of the relationship between:

A) two nominal variables
B) two ordinal variables
C) two interval variables
D) two ratio variables

Question 6

In a study where school teachers were asked to rate themselves on an ordinal scale for their job satisfaction and the teachers were also asked to provide the rating for their supervisors on an ordinal scale, in order to establish if there is a correlation between these two variables you will use:

A) the binomial test
B) Pearson's product moment correlation
C) Spearman's rank correlation
D) Kruskal-Wallis one-way analysis of variance

Question 7

When you have a correlation coefficient of .88 (p < .001), it means:

A) there is a weak relationship between the variables that is statistically significant
B) there is a moderate relationship between the variables that is not statistically significant
C) there is a strong relationship between the variables that is statistically significant
D) there is a moderately strong relationship between the variables

Question 8

The Mann-Whitney U test determines:

A) whether there is a significant relationship between a dichotomous ordinal variable and an interval variable when the data are drawn from independent samples
B) whether there is a significant relationship between a dichotomous nominal variable and an ordinal variable when the data are drawn from independent samples
C) whether there is a significant relationship between a dichotomous interval variable and an ordinal variable when the data are drawn from independent samples
D) whether there is a significant relationship between a dichotomous nominal variable and an ordinal variable when the data are drawn from dependent samples

Question 9

The Mann-Whitney U test tells you whether two groups differ on their responses to an:

A) interval variable
B) nominal variable
C) ordinal variable
D) both nominal and interval level variables

Question 10

If you would like to determine whether mothers and fathers differ on their level of agreement on a five-point ordinal scale with the statement "children should be given primacy in a household," you would employ:

A) Spearman's rank correlation
B) Pearson's product moment correlation
C) the Mann-Whitney U test
D) Kruskal-Wallis one-way analysis of variance

Question 11

The Mann-Whitney U test calculates a value for U from which:

A) a p value can be derived
B) a coefficient value can be derived
C) a chi square value can be derived
D) a value of r can be derived

Question 12

The Wilcoxon matched-pairs signed ranks test is a means of examining the relationship between a dichotomous nominal variable and an ordinal variable:

A) when the data are drawn from independent samples
B) when the data are drawn from same sample
C) when the data are drawn from related samples
D) when the data are drawn from different samples

Question 13

The test that can be used in a one-group pretest-posttest study in which the data on the client's behavior which is considered dependent variable is measured at the ordinal level is:

A) the Mann-Whitney U test
B) Spearman's rank correlation
C) Kruskal-Wallis one-way analysis of variance
D) Wilcoxon matched-pairs signed ranks test

Question 14

If you are interested in determining whether there is a significant relationship between an ordinal variable and a nominal variable with more than two categories when the data are drawn from independent samples, you will employ:

A) the Mann-Whitney U test
B) Kruskal-Wallis one-way analysis of variance
C) Spearman's rank correlation
D) Wilcoxon matched-pairs signed ranks test

Question 15

When you want to know whether teachers, school Principals, and school psychologists differ in their level of commitment to test based performance, and if you have measured their commitment at the ordinal level, you have:

A) a nominal variable with three categories of an ordinal variable
B) an ordinal variable with three categories and an ordinal variable
C) an interval variable with three categories and an ordinal variable
D) a nominal variable with three categories and an ordinal variable

Question 16

When you are comparing three groups of people rather than comparing two sets of scores from the same people, you have:

A) data that are independent
B) data that are dependent
C) data that are linked
D) grouped data

Question 17

The Spearman rank correlation coefficient ranks the study subjects on each variable and compares these ranks to compute the coefficient.

Question 18

If you wanted to test the effectiveness of a program of peer mentoring in which youth at risk are matched with mentors to pursue STEM (Science, Engineering, Technology & Math) majors and stay college bound, you can measure the effectiveness of this program by asking the respondents to indicate the level of their interest in pursuing college degree. You could compute a mean rank to that interest in pursuing a college degree in the STEM field was higher after participation in the mentoring services than before. The Wilcoxon test would then determine whether the increased interest in pursuing a degree in the STEM field as indicated in the posttest ranks is significantly different from those of the pretest.

Question 19

The Kruskal-Wallis test does not generate a p value to determine the level of statistical significance.

Question 20

When you want to compare groups on a given variable such as happiness, you could first compute the mean for each group for happiness to see which group was highest and then determine whether the differences among them are significant by employing the Kruskal-Wallis one-way analysis of variance.

Question 21

Mann-Whitney U test is helpful in determining whether two groups differ on their responses to an ordinal variable.

Question 22

Non-parametric statistical tests such as Mann-Whitney U Test are useful in analyzing data for variables measured at the interval level.

Question 23

For categorical data, people are placed into categories that have no order.

Question 24

Spearman rank correlation coefficient is used when the data are drawn from dependent samples.

Question 25

Statistical significance can be computed for Kruskal-Wallis one-way analysis of variance.

Question 26

Explain the importance of statistical significance in Kruskal-Wallis one-way analysis of variance.

Question 27

An evaluative researcher wants to know whether three different groups of professionals employed by a child welfare agency differ with regard to their level of job satisfaction, and he/she has measured job satisfaction at the ordinal level. In this situation, you have a nominal variable with three categories and an ordinal variable. Please explain the test you will use and the rationale for the choice of your statistical test.

Question 28

When a researcher is interested in measuring whether two groups differ on their responses to an ordinal variable, which test is the apt one to be used? Provide an example.