Objective: The aim of this paper is to address the design and analysis of validity studies for evaluating the performance of a nutritional screening or assessment tool.
Methods: Statistical aspects of validity are discussed with the non-statistician in mind. Particular emphasis is placed on construct and criterion validity. Design issues are considered in relation to numbers of subjects and assessments. Appropriate methods of analysis are considered, as is interpretation of results. Practical suggestions regarding the implementation of validity studies in the field of nutrition risk assessment are proposed.
Conclusion: Guidelines for the standardization of the design, analysis, and publication of results are presented.