Question 1

For which of the following would phi and Cramer's V be appropriate as measures of association?

Accepted Answer

A)  for variables measured at the nominal level 
B)  for variables measured at the interval-ratio level 
C)  for variables measured at the ordinal level 
D)  for variables measured at the continuous level 
A)  for variables measured at the nominal level 
B)  for variables measured at the interval-ratio level 
C)  for variables measured at the ordinal level 
D)  for variables measured at the continuous level A

Question 2

What is the major difference between Cramer's V and phi?

Accepted Answer

A)  Cramer's V is suitable for tables larger than 2 × 2, and phi is not. 
B)  Phi is suitable for tables larger than 2 × 2, and Cramer's V is not. 
C)  Cramer's V gives us the direction of an association, and phi does not. 
D)  Phi gives us the direction of an association, and Cramer's V does not. 
A)  Cramer's V is suitable for tables larger than 2 × 2, and phi is not. 
B)  Phi is suitable for tables larger than 2 × 2, and Cramer's V is not. 
C)  Cramer's V gives us the direction of an association, and phi does not. 
D)  Phi gives us the direction of an association, and Cramer's V does not. A

Question 3

-Refer to the table. Which of these statements is the best way to describe the relationship between type of workplace and job satisfaction?

Accepted Answer

A)  There is a weak positive association. 
B)  There is a moderate negative association. 
C)  There is an association, but it is impossible to determine its direction. 
D)  There is no association. 
A)  There is a weak positive association. 
B)  There is a moderate negative association. 
C)  There is an association, but it is impossible to determine its direction. 
D)  There is no association. D

Question 4

Which of the following best describes how measures of proportional reduction in error (PRE) quantify association?

Accepted Answer

A)  by comparing two predictions 
B)  by comparing two bivariate tables 
C)  by comparing expected and observed values in a bivariate table 
D)  by comparing two critical values 
A)  by comparing two predictions 
B)  by comparing two bivariate tables 
C)  by comparing expected and observed values in a bivariate table 
D)  by comparing two critical values

Question 5

-Refer to the table. How should we calculate the conditional distribution of job satisfaction among those in unionized workforces?

Accepted Answer

A)  (75/220) × 100 = 34% 
B)  (75/150) × 100 = 50% 
C)  ((75 + 220)/440) × 100 = 67% 
D)  ((75 + 220)/370) × 100 = 80% 
A)  (75/220) × 100 = 34% 
B)  (75/150) × 100 = 50% 
C)  ((75 + 220)/440) × 100 = 67% 
D)  ((75 + 220)/370) × 100 = 80%

Question 6

In a bivariate table, Cramer's V is 0.21 and lambda is 0. Which of the following should be our conclusion?

Accepted Answer

A)  that the null hypothesis cannot be rejected; we have no evidence of an association 
B)  that the null hypothesis should be rejected, but we may have made a computation error when calculating lambda 
C)  that although V suggests a moderate relationship, lambda corrects our interpretation to indicate there is no relationship 
D)  that the variables may be associated even if lambda is zero, and disregard lambda if the row marginals are very unequal 
A)  that the null hypothesis cannot be rejected; we have no evidence of an association 
B)  that the null hypothesis should be rejected, but we may have made a computation error when calculating lambda 
C)  that although V suggests a moderate relationship, lambda corrects our interpretation to indicate there is no relationship 
D)  that the variables may be associated even if lambda is zero, and disregard lambda if the row marginals are very unequal

Question 7

If the independent (X) variable is located in the columns, and the dependent (Y) variable in the rows of a bivariate table, how would we inspect the table to understand a bivariate association?

Accepted Answer

A)  Read down the rows to compare the conditional distributions of X. 
B)  Look across the columns to compare the conditional distributions of Y. 
C)  Look for the cell with the largest frequency value and compare it to the cell with the lowest frequency value. 
D)  Add the frequency values together and divide by the total frequency. 
A)  Read down the rows to compare the conditional distributions of X. 
B)  Look across the columns to compare the conditional distributions of Y. 
C)  Look for the cell with the largest frequency value and compare it to the cell with the lowest frequency value. 
D)  Add the frequency values together and divide by the total frequency.

Question 8

What must be calculated before phi and Cramer's V can be computed?

Accepted Answer

A)  lambda 
B)  maximum difference 
C)  clustered bar charts 
D)  chi square test 
A)  lambda 
B)  maximum difference 
C)  clustered bar charts 
D)  chi square test

Question 9

A researcher is interested in the association between gender (measured in two categories) and income (measured in three ordered categories: low, medium, high). What variable should dictate what measure of association is used?

Accepted Answer

A)  gender 
B)  income 
C)  low income 
D)  Measures of association cannot be computed with these variables. 
A)  gender 
B)  income 
C)  low income 
D)  Measures of association cannot be computed with these variables.

Question 10

Describe the difference between the "dependent-only prediction rule" and the "relational prediction rule" when applying a proportional reduction in error (PRE) approach.

Accepted Answer

The dependent-only rule is the first ste

Question 11

Why are measures of association helpful in testing the validity of a theory?

Accepted Answer

A)  Theories do not specify what the independent or dependent variable is in a relationship, so measures of association help clarify this point. 
B)  Theories seldom articulate cause-and-effect relationships, so measures of association help determine causality. 
C)  Theories stipulate that one variable causes another, so we need to know how strongly the independent variable is related to the dependent variable. 
D)  Measures of association are not helpful in testing a theory; they are used for research purposes where theory is not yet developed. 
A)  Theories do not specify what the independent or dependent variable is in a relationship, so measures of association help clarify this point. 
B)  Theories seldom articulate cause-and-effect relationships, so measures of association help determine causality. 
C)  Theories stipulate that one variable causes another, so we need to know how strongly the independent variable is related to the dependent variable. 
D)  Measures of association are not helpful in testing a theory; they are used for research purposes where theory is not yet developed.

Question 12

-Refer to the table. What is E2?

Accepted Answer

A)  30 
B)  45 
C)  55 
D)  70 
A)  30 
B)  45 
C)  55 
D)  70

Question 13

Which percentages should be computed in order to display the effects of the X variable on the Y variable in a bivariate table when the independent variable has been arranged in the columns?

Accepted Answer

A)  those using n as the denominator 
B)  those within each column 
C)  those within each row 
D)  those between each column 
A)  those using n as the denominator 
B)  those within each column 
C)  those within each row 
D)  those between each column

Question 14

A historical study of dietary patterns in the Middle Ages finds that, with increased wealth, people consumed more calories. How would this relationship be summarized statistically?

Accepted Answer

A)  There was a negative association between wealth and caloric intake. 
B)  There was a positive association between wealth and caloric intake. 
C)  There was a perfect association between wealth and caloric intake. 
D)  There was either a moderate or a strong association between these variables, and there could not be a weak relationship. 
A)  There was a negative association between wealth and caloric intake. 
B)  There was a positive association between wealth and caloric intake. 
C)  There was a perfect association between wealth and caloric intake. 
D)  There was either a moderate or a strong association between these variables, and there could not be a weak relationship.

Question 15

What is the basis of our prediction for the first prediction in a proportional reduction in error (PRE) measure?

Accepted Answer

A)  the most typical value of the dependent variable 
B)  the most typical value of the independent variable 
C)  the value of a dependent variable at a given value of the independent variable 
D)  the least common value from the dependent or independent variable-whichever is lower 
A)  the most typical value of the dependent variable 
B)  the most typical value of the independent variable 
C)  the value of a dependent variable at a given value of the independent variable 
D)  the least common value from the dependent or independent variable-whichever is lower

Question 16

A researcher has found a non-random relationship in a large random sample using hypothesis testing. Why would it be a good idea to also compute measures of association?

Accepted Answer

A)  The researcher will want to assess the strength and direction of the relationship. 
B)  Hypothesis tests do not have sufficient power to detect important relationships in large samples. 
C)  Measures of association can sometimes contradict hypothesis test results, so both are needed for complete information. 
D)  Measures of association are not necessary if the null hypothesis has been rejected. 
A)  The researcher will want to assess the strength and direction of the relationship. 
B)  Hypothesis tests do not have sufficient power to detect important relationships in large samples. 
C)  Measures of association can sometimes contradict hypothesis test results, so both are needed for complete information. 
D)  Measures of association are not necessary if the null hypothesis has been rejected.

Question 17

As age increases, memorization speed declines. What kind of relationship is this?

Accepted Answer

A)  a negative relationship 
B)  a positive relationship 
C)  a non-linear relationship 
D)  a null relationship 
A)  a negative relationship 
B)  a positive relationship 
C)  a non-linear relationship 
D)  a null relationship

Question 18

A researcher wants to predict fitness club membership, with income level as the independent variable. When ignoring income, the researcher's prediction results in 102 errors. When taking income into account, the number of errors falls to 56. What should be the conclusion?

Accepted Answer

A)  Income causes differences in fitness club membership. 
B)  There is an association between income and fitness club membership. 
C)  There is no association between income and fitness club membership. 
D)  Phi will equal the value of lambda. 
A)  Income causes differences in fitness club membership. 
B)  There is an association between income and fitness club membership. 
C)  There is no association between income and fitness club membership. 
D)  Phi will equal the value of lambda.

Question 19

What is the association between variables if they vary in opposite directions?

Accepted Answer

A)  neutral 
B)  contradictory 
C)  negative 
D)  positive 
A)  neutral 
B)  contradictory 
C)  negative 
D)  positive

Question 20

Which one of the following can occur to lambda when the row marginals are extremely unequal?

Accepted Answer

A)  Lambda may exceed 1.00. 
B)  Nothing, since lambda is not affected by row marginals. 
C)  Lambda may be less than zero. 
D)  Lambda may be zero even though there is an association between the variables. 
A)  Lambda may exceed 1.00. 
B)  Nothing, since lambda is not affected by row marginals. 
C)  Lambda may be less than zero. 
D)  Lambda may be zero even though there is an association between the variables.

For which of the following would phi and Cramer's V be appropriate as measures of association?

What is the major difference between Cramer's V and phi?

-Refer to the table. Which of these statements is the best way to describe the relationship between type of workplace and job satisfaction?

Which of the following best describes how measures of proportional reduction in error (PRE) quantify association?

-Refer to the table. How should we calculate the conditional distribution of job satisfaction among those in unionized workforces?

In a bivariate table, Cramer's V is 0.21 and lambda is 0. Which of the following should be our conclusion?

If the independent (X) variable is located in the columns, and the dependent (Y) variable in the rows of a bivariate table, how would we inspect the table to understand a bivariate association?

What must be calculated before phi and Cramer's V can be computed?

A researcher is interested in the association between gender (measured in two categories) and income (measured in three ordered categories: low, medium, high). What variable should dictate what measure of association is used?

Describe the difference between the "dependent-only prediction rule" and the "relational prediction rule" when applying a proportional reduction in error (PRE) approach.

Why are measures of association helpful in testing the validity of a theory?

-Refer to the table. What is E₂?

Which percentages should be computed in order to display the effects of the X variable on the Y variable in a bivariate table when the independent variable has been arranged in the columns?

A historical study of dietary patterns in the Middle Ages finds that, with increased wealth, people consumed more calories. How would this relationship be summarized statistically?

What is the basis of our prediction for the first prediction in a proportional reduction in error (PRE) measure?

A researcher has found a non-random relationship in a large random sample using hypothesis testing. Why would it be a good idea to also compute measures of association?

As age increases, memorization speed declines. What kind of relationship is this?

A researcher wants to predict fitness club membership, with income level as the independent variable. When ignoring income, the researcher's prediction results in 102 errors. When taking income into account, the number of errors falls to 56. What should be the conclusion?

What is the association between variables if they vary in opposite directions?

Which one of the following can occur to lambda when the row marginals are extremely unequal?

Introduction

Basic Descriptive Statistics: Percentages, Ratios and Rates, Tables, Charts, and Graphs

Measures of Central Tendency and Dispersion

The Normal Curve

Introduction to Inferential Statistics: Sampling and the Sampling Distribution

Estimation Procedures for Sample Means and Proportions

Hypothesis Testing With Nominal and Ordinal Variables: CHI Square

Measures of Association for Variables Measured at the Ordinal Level

Hypothesis Testing With Means and Proportions: The One-Sample Case

Hypothesis Testing With Means and Proportions: The Two-Sample Case

Hypothesis Testing With More Than Two Means: One-Way Analysis of Variance

Hypothesis Testing and Measures of Association for Variables Measured at the Interval-Ratio Level

Partial Correlation and Multiple Regression and Correlation

Filters

Exam 8: Measures of Association for Variables Measured at the Nominal Level

For which of the following would phi and Cramer's V be appropriate as measures of association?

What is the major difference between Cramer's V and phi?

-Refer to the table. Which of these statements is the best way to describe the relationship between type of workplace and job satisfaction?

Which of the following best describes how measures of proportional reduction in error (PRE) quantify association?

-Refer to the table. How should we calculate the conditional distribution of job satisfaction among those in unionized workforces?

In a bivariate table, Cramer's V is 0.21 and lambda is 0. Which of the following should be our conclusion?

If the independent (X) variable is located in the columns, and the dependent (Y) variable in the rows of a bivariate table, how would we inspect the table to understand a bivariate association?

What must be calculated before phi and Cramer's V can be computed?

A researcher is interested in the association between gender (measured in two categories) and income (measured in three ordered categories: low, medium, high). What variable should dictate what measure of association is used?

Describe the difference between the "dependent-only prediction rule" and the "relational prediction rule" when applying a proportional reduction in error (PRE) approach.

Why are measures of association helpful in testing the validity of a theory?

-Refer to the table. What is E2?

Which percentages should be computed in order to display the effects of the X variable on the Y variable in a bivariate table when the independent variable has been arranged in the columns?

A historical study of dietary patterns in the Middle Ages finds that, with increased wealth, people consumed more calories. How would this relationship be summarized statistically?

What is the basis of our prediction for the first prediction in a proportional reduction in error (PRE) measure?

A researcher has found a non-random relationship in a large random sample using hypothesis testing. Why would it be a good idea to also compute measures of association?

As age increases, memorization speed declines. What kind of relationship is this?

A researcher wants to predict fitness club membership, with income level as the independent variable. When ignoring income, the researcher's prediction results in 102 errors. When taking income into account, the number of errors falls to 56. What should be the conclusion?

What is the association between variables if they vary in opposite directions?

Which one of the following can occur to lambda when the row marginals are extremely unequal?

Introduction

Basic Descriptive Statistics: Percentages, Ratios and Rates, Tables, Charts, and Graphs

Measures of Central Tendency and Dispersion

The Normal Curve

Introduction to Inferential Statistics: Sampling and the Sampling Distribution

Estimation Procedures for Sample Means and Proportions

Hypothesis Testing With Nominal and Ordinal Variables: CHI Square

Measures of Association for Variables Measured at the Ordinal Level

Hypothesis Testing With Means and Proportions: The One-Sample Case

Hypothesis Testing With Means and Proportions: The Two-Sample Case

Hypothesis Testing With More Than Two Means: One-Way Analysis of Variance

Hypothesis Testing and Measures of Association for Variables Measured at the Interval-Ratio Level

Partial Correlation and Multiple Regression and Correlation

Filters

-Refer to the table. What is E₂?