CRSP Selection Testing & Validation 3 — Questions and Answers
Question 1: According to the Uniform Guidelines on Employee Selection Procedures, adverse impact is generally indicated when a protected group's selection rate is less than what fraction of the majority group's rate?
- One-half (50%)
- Two-thirds (67%)
- Four-fifths (80%) (Correct answer)
- Nine-tenths (90%)
Correct answer: Four-fifths (80%)
The 4/5ths (or 80%) rule states that adverse impact is indicated when a protected group's selection rate is less than 80% of the highest-selected group's rate.
Question 2: Which of the following is the PRIMARY advantage of a structured interview over an unstructured one for selection purposes?
- It is faster to administer
- It produces higher inter-rater reliability and validity (Correct answer)
- It is preferred by most candidates
- It requires no interviewer training
Correct answer: It produces higher inter-rater reliability and validity
Structured interviews use standardized questions and rating scales, which significantly increase inter-rater reliability and predictive validity compared to unstructured formats.
Question 3: A test developer wants to ensure that their new personality inventory measures 'conscientiousness' and not 'neuroticism.' Evaluating this concern falls under which validity category?
- Content validity
- Criterion-related validity
- Construct validity (Correct answer)
- Predictive validity
Correct answer: Construct validity
Construct validity examines whether a test truly measures the theoretical construct it is intended to measure, distinguishing it from other constructs.
Question 4: A biodata questionnaire used in selection is MOST legally defensible when:
- It includes questions about an applicant's family background
- Each item is empirically keyed and validated against job performance criteria (Correct answer)
- It focuses on educational credentials only
- It is reviewed by a panel of senior managers
Correct answer: Each item is empirically keyed and validated against job performance criteria
Biodata items must be empirically validated — each item should have a demonstrated relationship to job-relevant criteria to be legally and professionally defensible.
Question 5: Which statistical concept describes the degree to which two selection measures (e.g., a test and a rating) are measuring the same underlying construct?
- Convergent validity (Correct answer)
- Discriminant validity
- Criterion-related validity
- Test-retest reliability
Correct answer: Convergent validity
Convergent validity is demonstrated when two measures that are theoretically supposed to measure the same construct correlate highly with each other.
Question 6: An employer administers the same cognitive test twice to the same applicants one week apart. The correlation between the two sets of scores reflects which psychometric property?
- Internal consistency reliability
- Inter-rater reliability
- Test-retest reliability (Correct answer)
- Parallel forms reliability
Correct answer: Test-retest reliability
Test-retest reliability is assessed by administering the same test to the same group at two different times and computing the correlation between scores.
Question 7: In a utility analysis for a selection test, which factor would MOST directly increase the estimated dollar value gained from using the test?
- Increasing the number of test items
- Increasing the validity coefficient of the test (Correct answer)
- Reducing the time candidates spend on the test
- Lowering the minimum passing score
Correct answer: Increasing the validity coefficient of the test
In the Brogden-Cronbach-Gleser utility model, the utility of a selection test increases proportionally with its validity coefficient.
CRSP Selection Testing & Validation 3
According to the Uniform Guidelines on Employee Selection Procedures, adverse impact is generally indicated when a protected group's selection rate is less than what fraction of the majority group's rate?