Describir: Uses and limitations of statistical accounting for random error correlations, in the validation of dietary questionnaire assessments.