Dietary methodology: implications of errors in the measurement.

The assessment of food consumption of individuals and groups is extremely important for the science of nutrition. In the past decades several methods have been developed with various characteristics (Table 1) and adapted to serve widely different purposes of nutritional research. Although it is generally accepted that all methods have limitations and advantages, there are doubts about, for instance, the reliability of interview methods. To examine the validity of data obtained by such methods it is important to consider questions such as those phrased by Beaton, 1990 (presented at IUNS Conference Seoul, 1989): What does the estimated intake really represent? What is the nature and magnitude of the error in that estimate? What is the implication of the error for interpretation of analyses? To answer these and similar questions, methods should be validated. The criteria for the validation procedure partly depend on the purpose of the study in which the method will be used.