Demands are growing for information on the quality of food composition data. It is generally acknowledged
today that the traditional measures of data quality (e.g. standard deviations) do not adequately
address the complex issues affecting the quality of food composition data. Current efforts to address this
problem centre on using multiple-choice questions to create ‘‘objective” quality scores for component
values, allowing numeric scores to be set for a wide range of important factors. This paper discusses some
of the significant challenges facing us as we begin to implement the process of creating the standards for,
and assigning quality values to food composition data. Specific strategic approaches are suggested that
can reduce both the complexity and the cost of overcoming each of these challenges