Deck 9: Construct Validation

Question 1

________ validation is the most comprehensive umbrella approach to validation.

A) Predictive criterion-related
B) Concurrent criterion-related
C) Content
D) Construct

Answer 1

D

Question 2

Cronbach and Meehl (1955) argued that establishing a measure's validity included developing a theory about the construct of interest. The specification of that theory refers to establishing a ?________.

A) table of specifications
B) nomological network
C) multi-trait multi-method matrix
D) generalizability theory

Answer 2

B

Question 3

Convergent validity could be examined by ________.

A) having subject matter experts rate the degree to which items assess the construct of excitability
B) correlating applicant interview scores with job-performance scores
C) correlating scores on a newly developed measure of cheerfulness with scores on the Beck Depression Inventory
D) correlating scores on a newly developed measure of intelligence with scores on conscientiousness

Answer 3

C

Question 4

A clear case of discriminant validity would be demonstrated when two measures are found to correlate near ________.

A) -1.0
B) 0.0
C) )50
D) 1.0

Question 5

Assessing multiple measures using Likert-type self-report response items may inflate observed correlations due to ________ .

A) convergent validity
B) construct underrepresentation
C) common method variance
D) discriminant validity

Question 6

When examining a multitrait-multimethod matrix (MTMM), a correlation between theoretically distinct measures that are assessed using the same method of measurement would provide evidence for ________.

A) convergent validity
B) construct underrepresentation
C) common method variance
D) discriminant validity

Question 7

The reliability of a measure could be a useful element in the ________ validation process.

A) content
B) predictive criterion-related
C) concurrent criterion-related
D) construct

Question 8

Today we recognize that the focus of validation must be on the ________ .

A) test as a whole
B) items that compose a test
C) correct choice of type of validity study
D) use of test scores

Question 9

If a measure of intelligence was impacted by one's cultural understanding, the measure would suffer from ________.

A) construct underrepresentation
B) construct irrelevant variance
C) unification
D) unreliability

Question 10

Criterion-related validation evidence would be most consistent with Messick's (1995a) ________ aspect of construct validity.

A) external
B) Substantive
C) Structural
D) consequential

Question 11

Cronbach and Meehl (1955) defined a nomological network of a construct. Identify and briefly explain each major component of a nomological network.

Question 12

Discuss how convergent and discriminant validity can provide evidence for construct validation.

Question 13

(a) Define common method variance (CMV). (b) Explain why CMV can be a challenge for demonstrating construct validity. (c) Briefly explain how CMV can be examined using a multi-trait multimethod matrix (MTMM).

Question 14

Numerous research studies may be conducted to provide construct validation evidence. Explain how studies of group differences and studies of process can provide construct validation evidence.

Question 15

It is sometimes said that a test "is valid" or "has validity". Provide a critique of such an assertion using the modern conception of validation.