Chapter 8: Fairness in Employment Decisions

Test Bank

Multiple Choice

1. When the relationship between the predictor and criterion is different for different groups, this is called ______.

A. differential validity

B. differential prediction

C. adverse impact

D. differential weighting

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Knowledge

Answer Location: Chapter Opener
Difficulty Level: Easy

AACSB: Making sound decisions

2. Your organization is using a cognitive ability test to select for employees in an engineering position, but it has led to a lower proportion of minorities workers being hired than the majority. This BEST represents the concept of ______.

A. differential validity

B. differential prediction

C. adverse impact

D. differential weighting

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Application

Answer Location: Chapter Opener
Difficulty Level: Medium

AACSB: Making sound decisions

3. If the distribution of your predictor/criterion scatterplot is similar for group and circular, then this predictor likely has ______.

A. negative validity

B. zero validity

C. positive validity

D. absolute validity

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Analysis

Answer Location: Assessing Differential Validity
Difficulty Level: Hard

AACSB: Making sound decisions

4. Determining if adverse impact could potentially exist based on the selection rate of your majority and minority groups is based on the ______ rule.

A. one-half

B. two-thirds

C. four-fifths

D. five-sixths

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Knowledge

Answer Location: Differential Validity and Adverse Impact
Difficulty Level: Hard

AACSB: Making sound decisions

5. Assume that for a given position, 30 minorities are hired out of 200 minorities applicants and that 100 nonminorities are hired out 400. Does this selection system violate the four-fifths rule?

A. Yes

B. No

C. Not enough information is available

D. The two selection ratios are exactly equal.

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Analysis

Answer Location: Differential Validity and Adverse Impact
Difficulty Level: Medium

AACSB: Making sound decisions

6. This picture shows what type of relationship between groups and validity of the measure?

A. equal validity, unequal predictor means

B. equal predictor means, but positive validity only for the nonminority group

C. positive validity and adverse impact

D. positive validity for entire group; zero validity for each group separately

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Analysis

Answer Location: Differential Validity and Adverse Impact
Difficulty Level: Hard

AACSB: Making sound decisions

7. If a predictor shows no validity only for the nonminority group (but zero validity for the minority group), then the predictor ______.

A. should only be used to predict the minority group

B. should only be used to predict the nonminority group

C. should be used to predict for both groups

D. cannot be used at all

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Comprehension

Answer Location: Differential Validity and Adverse Impact
Difficulty Level: Medium

AACSB: Making sound decisions

8. One of the issues with examining differential validity by using single-group validity is that ______.

A. the minority group typically has a smaller sample size

B. the majority group typically has a smaller sample size

C. the measure is less reliable for the minority group

D. the measure is less reliable for the majority group

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Comprehension

Answer Location: Differential Validity: The Evidence
Difficulty Level: Medium

AACSB: Making sound decisions

9. Since sample sizes are typically inadequate in any one differential validity study, ______ are often used to determine if differential validity might exist for certain variables.

A. dominance analyses

B. regression lines

C. meta-analyses

D. multivariate analyses

Learning Objective: 8-8: Use the latest knowledge regarding test fairness to inform HR practices and public policy.

Cognitive Domain: Comprehension

Answer Location: Differential Validity: The Evidence
Difficulty Level: Medium

AACSB: Making sound decisions

10. Based on recent research, which of these groups has the lowest validity (corrected) coefficient for the relationship between cognitive ability tests and performance?

A. Whites

B. Asian-Americans

C. Hispanic Americans

D. African Americans

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Comprehension

Answer Location: Differential Validity: The Evidence
Difficulty Level: Medium

AACSB: Making sound decisions

11. Which type of variable depresses the size of the predictor–criterion relationship depending on what group of the variable the person belongs to?

A. moderator

B. mediator

C. suppressor

D. extraneous variable

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Knowledge

Answer Location: Assessing Differential Prediction and Moderator Variables
Difficulty Level: Easy

AACSB: Making sound decisions

12. Which of these can show you predictive bias among groups when using moderated multiple regression?

A. slope differences between subgroups only

B. intercept differences between subgroups only

C. both slope and intercept differences

D. neither slope nor intercept differences can tell you this

Learning Objective: 8-8: Use the latest knowledge regarding test fairness to inform HR practices and public policy.

Cognitive Domain: Comprehension

Answer Location: Assessing Differential Prediction and Moderator Variables
Difficulty Level: Medium

AACSB: Making sound decisions

13. In the moderated multiple regression equation, a represents the ______.

A. predicted value for the criterion

B. the least-squares estimate of the population regression coefficient for the moderator

C. least-squares estimate of the population regression coefficient for the predictor

D. the least squares estimate of the intercept

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Knowledge

Answer Location: Assessing Differential Prediction and Moderator Variables
Difficulty Level: Easy

AACSB: Making sound decisions

14. Before the late 2000s, evidence seemed to suggest what about differential prediction for cognitive ability?

A. an overall lack of differential prediction based on ethnicity and gender

B. differential prediction for ethnic minorities, but only for Hispanics, not African Americans

C. differential prediction for gender, namely women were underrepresented

D. differential prediction for all ethnic minorities

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Comprehension

Answer Location: Assessing Differential Prediction and Moderator Variables
Difficulty Level: Medium

AACSB: Making sound decisions

15. Because regression coefficients are tied to specific measures, it is preferable for those conducting meta-analyses to collect ______ when testing for differential prediction, which can be compared across measures.

A. standard deviations

B. correlation coefficients

C. standard errors

D. chi squares

Learning Objective: 8-2: Design and conduct a state-of-the-science differential prediction study.

Cognitive Domain: Comprehension

Answer Location: Using Meta-Analysis to Assess Differential Prediction
Difficulty Level: Medium

AACSB: Making sound decisions

16. A meta-analysis on the SAT and high school GPA showed that they had differential predictions such that, ______.

A. Blacks and Hispanics were overpredicted, and females were underpredicted

B. Blacks and Hispanics were underpredicted, and females were underpredicted

C. Blacks and Hispanics were overpredicted, and females were overpredicted

D. Blacks and Hispanics were underpredicted, and females were overpredicted

Learning Objective: 8-2: Design and conduct a state-of-the-science differential prediction study.

Cognitive Domain: Comprehension

Answer Location: Using Meta-Analysis to Assess Differential Prediction
Difficulty Level: Medium

AACSB: Making sound decisions

17. All of the following are suggestions given to increase power in moderated multiple regression in assessing differential prediction EXCEPT ______.

A. use a preset Type I error based on the serious of making the error, not strictly .05 level

B. plan research design to have a sufficiently large sample size

C. use a dichotomous criterion variable

D. develop and use reliable measures

Learning Objective: 8-2: Design and conduct a state-of-the-science differential prediction study.

Cognitive Domain: Analysis

Answer Location: Suggestions for Improving the Accuracy of Slope-Based Differential Prediction Assessment
Difficulty Level: Hard

AACSB: Making sound decisions

18. Which concept, that occurs when individuals believe they face the process of being evaluated on a negative typecast for their group and confirming, can lead to differential prediction?

A. adverse impact

B. cultural frame of reference

C. scale coarseness

D. stereotype threat

Learning Objective: 8-2: Design and conduct a state-of-the-science differential prediction study.

Cognitive Domain: Knowledge

Answer Location: Suggestions for Improving the Accuracy of Slope-Based Differential Prediction Assessment
Difficulty Level: Easy

AACSB: Making sound decisions

19. All of the following are strategies for reducing adverse impact EXCEPT ______.

A. improve the recruiting strategy for minorities

B. use differential weighting for the various criterion facet

C. use cognitive abilities in combination with noncognitive predictors

D. reduce face validity to prevent faking

Learning Objective: 8-4: Plan and execute strategies aimed at reducing adverse impact.

Cognitive Domain: Comprehension

Answer Location: Further Considerations Regarding Adverse Impact, Differential Validity, and Differential Prediction

Difficulty Level: Medium

AACSB: Making sound decisions

20. When you are using differential weighting to reduce adverse impact, you should give measures of general cognitive abilities.

A. zero weighting

B. more weighting

C. less weighting

D. equal weighting

Learning Objective: 8-8: Use the latest knowledge regarding test fairness to inform HR practices and public policy.

Cognitive Domain: Comprehension

Answer Location: Further Considerations Regarding Adverse Impact, Differential Validity, and Differential Prediction

Difficulty Level: Medium

AACSB: Making sound decisions

21. According to at least one study published in 2015, Black–White differences in cognitive ability tests show up as early as ______.

A. 24 months

B. 54 months

C. 7 years

D. 10 years

Learning Objective: 8-4: Plan and execute strategies aimed at reducing adverse impact.

Cognitive Domain: Knowledge

Answer Location: Further Considerations Regarding Adverse Impact, Differential Validity, and Differential Prediction

Difficulty Level: Medium

AACSB: Making sound decisions

22. Which type of banding includes the reliability for the performance measure and the predictor-performance relationship as well?

A. criterion-referenced banding model

B. sliding-band method

C. validity-referenced banding model

D. bottom-up banding method

Learning Objective: 8-5: Implement test-score banding based on technical and legal guidelines.

Cognitive Domain: Knowledge

Answer Location: Minimizing Adverse Impact Through Test-Score Banding

Difficulty Level: Easy

AACSB: Making sound decisions

23. You are selecting one new manager for your department. You run a regression with multiple predictors, but instead of just selecting the top score, you create a group within one standard error for the difference of the top score and treat them all equal. You then select from this group for your manager position. This describes the ______.

A. criterion-referenced banding model

B. sliding-band method

C. validity-referenced banding model

D. bottom-up banding method

Learning Objective: 8-5: Implement test-score banding based on technical and legal guidelines.

Cognitive Domain: Application

Answer Location: Minimizing Adverse Impact Through Test-Score Banding

Difficulty Level: Medium

AACSB: Making sound decisions

24. When using a sliding-band method of banding to select multiple people for a position, if the top scorer is chosen and more applicants need to be selected, what should be done next?

A. An applicant is picked from those remaining in the band.

B. A new band is created using the standard error of the difference.

C. The next person down on the list is added to the band.

D. The band is “slide down” to include the same number of people as the initial band.

Learning Objective: 8-5: Implement test-score banding based on technical and legal guidelines.

Cognitive Domain: Comprehension

Answer Location: Minimizing Adverse Impact Through Test-Score Banding

Difficulty Level: Hard

AACSB: Making sound decisions

25. All of the following are true of criterion-referenced bands compared to predictor-referenced bands EXCEPT ______.

A. criterion-referenced bands produce skinnier bands

B. criterion-referenced bands may not be possible because of a lack of criterion-related validity information

C. it includes the use of available criterion data, which may be imperfect

D. it includes two reliability coefficients, one for each variable in the banding model

Learning Objective: 8-5: Implement test-score banding based on technical and legal guidelines.

Cognitive Domain: Comprehension

Answer Location: Minimizing Adverse Impact Through Test-Score Banding

Difficulty Level: Medium

AACSB: Making sound decisions

26. Which court decision confirmed that banding is a legal mechanism for reducing adverse impact?

A. Chicago Firefighters Local 2 v. City of Chicago

B. Griggs v. Duke Power Company

C. Meritor Savings Bank v. Vinson

D. Watson v. Fort Worth Bank & Trust

Learning Objective: 8-5: Implement test-score banding based on technical and legal guidelines.

Cognitive Domain: Knowledge

Answer Location: Minimizing Adverse Impact Through Test-Score Banding

Difficulty Level: Medium

AACSB: Making sound decisions

27. Which type of justice says that test takers, applicants, and employees perceive the outcomes, such as hiring and promotions, to be fair?

A. social

B. interactional

C. procedural

D. distributive

Learning Objective: 8-6: Distinguish among different types of fairness perceptions and create selection systems that address each.

Cognitive Domain: Knowledge

Answer Location: Fairness and the Interpersonal Context of Employment Testing
Difficulty Level: Easy

AACSB: Making sound decisions

28. Applicants high in which personality construct are more likely to have positive perceptions of fairness?

A. openness to experience

B. agreeableness

C. neuroticism

D. conscientiousness

Learning Objective: 8-6: Distinguish among different types of fairness perceptions and create selection systems that address each.

Cognitive Domain: Knowledge

Answer Location: Fairness and the Interpersonal Context of Employment Testing
Difficulty Level: Easy

AACSB: Making sound decisions

29. According to the reading, we are currently in a state of ______ because we seek answers in a time of doubt about the quality of things.

A. naive acceptance of tests

B. intense hostility to tests

C. higher acceptance of tests

D. ambivalence toward tests

Learning Objective: 8-7: Conceptualize fairness as a technical but also a social and interpersonal concept.

Cognitive Domain: Comprehension

Answer Location: Fair Employment and Public Policy
Difficulty Level: Medium

AACSB: Making sound decisions

Short Answer

1. Define fairness as it is defined in terms of how organizations should view it.

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Comprehension

Answer Location: Chapter Opener
Difficulty Level: Medium

AACSB: Making sound decisions

2. What are the two conditions that must exist for differential validity to exist?

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Comprehension

Answer Location: Differential Validity: The Evidence
Difficulty Level: Medium

AACSB: Making sound decisions

3. Briefly describe the evidence on differential validity for men and women.

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Comprehension

Answer Location: Differential Validity: The Evidence
Difficulty Level: Medium

AACSB: Making sound decisions

4. Explain the concept of differential prediction generalization.

Learning Objective: 8-2: Design and conduct a state-of-the-science differential prediction study.

Cognitive Domain: Comprehension

Answer Location: Using Meta-Analysis to Assess Differential Prediction
Difficulty Level: Hard

AACSB: Making sound decisions

5. Explain how alternate modes of presenting test stimuli can be used to reduce adverse impact.

Learning Objective: 8-4: Plan and execute strategies aimed at reducing adverse impact.

Cognitive Domain: Comprehension

Answer Location: Further Considerations Regarding Adverse Impact, Differential Validity, and Differential Prediction

Difficulty Level: Medium

AACSB: Making sound decisions

6. Take a stand on whether banding is effective, ethical, and/or useful for organization or not and present some evidence to back your stance.

Learning Objective: 8-5: Implement test-score banding based on technical and legal guidelines. Cognitive Domain: Comprehension

Answer Location: Minimizing Adverse Impact Through Test-Score Banding

Difficulty Level: Medium

AACSB: Making sound decisions

Essay

1. Describe what a situation with equal validity between groups but unequal prediction would look like. What type of selection would this result, and what should the organization do about it?

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Comprehension

Answer Location: Differential Validity and Adverse Impact
Difficulty Level: Hard

AACSB: Making sound decisions

2. Explain some of the problems in testing for differential prediction described in the reading.

Learning Objective: 8-1: Assess the presence of differential validity, differential prediction, and adverse impact.

Cognitive Domain: Comprehension

Answer Location: Problems in Testing for Differential Prediction
Difficulty Level: Hard

AACSB: Making sound decisions

3. Describe some of the relationships and trade-offs between validity and adverse impact.

Learning Objective: 8-3: Anticipate possible trade-offs between validity and adverse impact.

Cognitive Domain: Analysis

Answer Location: Further Considerations Regarding Adverse Impact, Differential Validity, and Differential Prediction

Difficulty Level: Hard

AACSB: Making sound decisions