Evaluation of Diet Quality Measurement Instruments Informed by Item Response Theory

Item Response Theory (IRT) is one of measurement validity methods and used for evaluating item difficulty and item discrimination of measurement scales. The purpose of this secondary analysis was to evaluate Canadian Healthy Eating Index (HEI-C) and Mediterranean Diet Score (MDS) informed by IRT methods, which were used in the CHANGE study. A total of five HEI-C components showed ceiling or floor effects indicating item difficulty. Total fruits and vegetables and milk and alternatives showed relatively clear discriminations regarding the level of relative nutrient intakes and those changes over three months. MDS provided weak information regarding nutrient intakes except for fruits and vegetables. Modified version of MDS had weak associations with relative HEI-C components and nutrient intakes. In conclusion, HEI-C and MDS indicated weak discriminations against nutrient intakes except for some components, however this research provides some ideas for developing and evaluating a new assessment tool informed by IRT.