Which concept concerns whether a measure truly reflects what it intends to measure?

Which concept concerns whether a measure truly reflects what it intends to measure?

• A. Reliability
• B. Bias
• C. Validity
• D. Sampling

Answer: (C)

Validity is the idea that a measure truly captures the concept it is meant to reflect. When something is valid, its results accurately represent the attribute or construct you care about, so the inferences you draw from the scores are meaningful. For example, a test designed to assess mathematical problem-solving should yield scores that correspond to actual math ability, not to reading speed or handwriting quality. Reliability is about consistency—getting similar results across repeated measurements or different raters—without necessarily being about whether those results map onto the intended concept. A test can be reliable but not valid if it consistently measures something else. Bias involves systematic errors that skew results in a particular direction, but it doesn’t by itself address whether the measurement matches the intended construct. Sampling concerns who is included in the measurement, which affects generalizability but not whether the measurement captures the intended concept. Different forms of validity—content validity (covering all relevant aspects), construct validity (aligning with theoretical relationships), and criterion validity (predicting or correlating with relevant outcomes)—all focus on whether the measurement truly reflects the intended concept.

Validity is the idea that a measure truly captures the concept it is meant to reflect. When something is valid, its results accurately represent the attribute or construct you care about, so the inferences you draw from the scores are meaningful. For example, a test designed to assess mathematical problem-solving should yield scores that correspond to actual math ability, not to reading speed or handwriting quality.

Reliability is about consistency—getting similar results across repeated measurements or different raters—without necessarily being about whether those results map onto the intended concept. A test can be reliable but not valid if it consistently measures something else. Bias involves systematic errors that skew results in a particular direction, but it doesn’t by itself address whether the measurement matches the intended construct. Sampling concerns who is included in the measurement, which affects generalizability but not whether the measurement captures the intended concept.

Different forms of validity—content validity (covering all relevant aspects), construct validity (aligning with theoretical relationships), and criterion validity (predicting or correlating with relevant outcomes)—all focus on whether the measurement truly reflects the intended concept.