Evaluating bias in validity studies of developmental/behavioral screening tests.

This commentary describes six potential problems and sources of bias that are often overlooked and under-reported in validity studies of developmental/behavioral screening tests, namely, 1) problems with small samples, 2) verification bias, 3) inappropriate/non-equivalent test bias, 4) procedural bias, 5) incomplete reporting of results, and 6) spectrum bias. These issues are then applied to review an unusually complete published report of a screening test validity study along with a discussion of how results may affect application in different settings.