Concurrent Validity

Concurrent validity is a type of criterion validity that evaluates how well a new measurement instrument agrees with an established gold standard when both are administered at the same point in time. If the new measure produces scores that correlate strongly with the reference standard, researchers can be more confident that it captures the same underlying construct. Concurrent validity is distinct from predictive validity, which assesses agreement between a measure and a criterion assessed at a future time point.

Researchers frequently need shorter, cheaper, or less invasive instruments that perform as well as existing gold standards. A screening questionnaire that can substitute for a lengthy diagnostic interview saves both time and resources. Concurrent validity evidence is what justifies that substitution. Without it, there is no empirical basis for claiming that the new tool can stand in for the reference standard.

Researchers administer the new instrument and the reference standard to the same group of participants during a single testing session or within a narrow time window. They then quantify how closely the two measures correspond, typically using Pearson's r for continuous scores or Cohen's kappa for agreement between categorical classifications. A strong positive correlation (often r greater than 0.70) suggests acceptable concurrent validity. Researchers may also examine sensitivity and specificity if the new measure produces binary classifications. A strong correlation does not by itself show that the two measures agree: a new measure can correlate highly with the reference standard while reading consistently higher or lower. When both measures use the same units, Bland Altman plots show the level of agreement across the range of scores, revealing any systematic difference and whether discrepancies are larger at certain values.

Use concurrent validity testing when you are developing or adapting a measurement instrument and a credible reference standard exists. It is especially relevant when the new measure offers practical advantages: lower cost, reduced participant burden, faster administration, or applicability in settings where the gold standard is impractical. Concurrent validity is also appropriate when translating an instrument into a new language or adapting it for a different cultural context and you want to compare the adapted version with the original.

One frequent error is confusing concurrent validity with convergent validity. Convergent validity is evidence for construct validity: the new measure correlates as theory predicts with other measures of the same or related constructs, none of which needs to be a gold standard. Concurrent validity is a form of criterion validity: it compares a new measure with a reference standard for the same construct, measured at the same time. Another mistake is using an inadequate reference standard: if the gold standard itself has questionable validity, the comparison tells you little. Researchers also sometimes conclude that high concurrent validity means the new instrument is "valid" in all respects, ignoring other dimensions such as content validity or construct validity. Finally, sample size matters; a small sample can produce misleadingly high or low correlation coefficients with wide confidence intervals.

A team of psychologists develops a ten item screener for generalized anxiety disorder and wants to validate it against the Structured Clinical Interview for DSM 5 (SCID 5). They recruit 200 adults from a primary care clinic. Each participant completes the screener and then undergoes the SCID 5 interview on the same day. When a cutoff score is applied, the screener identifies SCID 5 diagnoses of generalized anxiety disorder with sensitivity of 0.88 and specificity of 0.79. These results support concurrent validity and suggest the screener can serve as a practical first stage assessment in busy clinical settings.