Abstract
Validation considers the context within which measures are being made (national, household, individual) and helps to describe the extent to which observed values differ from the truth. This chapter outlines the general principles that should be applied in the design of validation studies of dietary measurements, and considers some of the statistical techniques that have been developed to overcome the problems which arise from the absence of a reference measure of ‘true’ dietary intakes. The concepts of validity and reproducibility are reviewed. The context of validation is considered (current or past intake; food consumption or nutrient intake; absolute or relative intakes; group versus individual intakes). Validation techniques, including the identification of relevant reference measures, validation procedures, and factors affecting the design of validation studies are considered. Statistical techniques for interpreting findings from validation studies are described in detail.