Construct Validity
The degree to which a test measures what it claims, or purports, to be measuring.
What is Construct Validity?
Construct validity links the practical measurement tool back to the underlying theoretical concept (construct) it is supposed to measure. It involves demonstrating that the measure correlates with things it theoretically should correlate with (convergent validity) and does not correlate with things it shouldn't (discriminant validity).
Why Construct Validity Matters
Many crucial variables in research (intelligence, motivation, pain) cannot be observed directly. Construct validity ensures we are actually capturing these invisible variables.
Example
A researcher creates a survey to measure 'burnout'. To prove construct validity, they show the survey scores correlate highly with clinical diagnoses of exhaustion and poorly with unrelated traits like introversion.
Common Mistakes
- Relying solely on face validity (it 'looks' like it measures the construct).
- Failing to clearly define the theoretical construct before attempting to measure it.