Discriminant (divergent) validity indicates which of the following?

Discriminant (divergent) validity indicates which of the following?

• A. The test correlates strongly with measures of different constructs
• B. The test does not correlate too strongly with measures of different constructs
• C. The test correlates with all measures of every construct
• D. The test has poor reliability

Answer: (B)

Discriminant (divergent) validity checks that a test is measuring a distinct construct and not something else. The idea is that a test designed to measure a particular concept should not show a strong relationship with measures of different, unrelated constructs. If it did correlate strongly with these other measures, it would suggest the test isn’t capturing a unique construct after all. That’s why the best statement is that the test does not correlate too strongly with measures of different constructs. It reflects the idea that separate constructs should be distinct enough to show low correlations with each other, supporting the test’s specificity. Context helps: tests that measure related ideas should align (convergent validity), showing high correlations with measures of the same construct. But tests should not align closely with measures of different constructs, which would blur the lines between distinct traits. Why the other thoughts don’t fit: a strong correlation with measures of different constructs would undermine discriminant validity, indicating overlap with something not meant to be the same construct. Correlating with all measures of every construct would be nonsensical and would imply a lack of construct specificity. Poor reliability is about consistency over time or items, not about whether the test relates to other constructs, so it’s a separate issue.

Discriminant (divergent) validity checks that a test is measuring a distinct construct and not something else. The idea is that a test designed to measure a particular concept should not show a strong relationship with measures of different, unrelated constructs. If it did correlate strongly with these other measures, it would suggest the test isn’t capturing a unique construct after all.

That’s why the best statement is that the test does not correlate too strongly with measures of different constructs. It reflects the idea that separate constructs should be distinct enough to show low correlations with each other, supporting the test’s specificity.

Context helps: tests that measure related ideas should align (convergent validity), showing high correlations with measures of the same construct. But tests should not align closely with measures of different constructs, which would blur the lines between distinct traits.

Why the other thoughts don’t fit: a strong correlation with measures of different constructs would undermine discriminant validity, indicating overlap with something not meant to be the same construct. Correlating with all measures of every construct would be nonsensical and would imply a lack of construct specificity. Poor reliability is about consistency over time or items, not about whether the test relates to other constructs, so it’s a separate issue.