Validation of a self-reported shelf inventory to measure food purchase behavior.

A mailed, self-reported shelf inventory was validated for use as a tool in assessing the impact of a community nutrition intervention that included a point-of-purchase component. The self-reported inventory was evaluated for overall accuracy as well as for the effects of gender, age, and shopping responsibility on accuracy. In addition, the food-shelf inventory was compared with a specific food frequency questionnaire. Specificity and sensitivity were calculated for self-reported inventories using an interviewer-completed, same-day inventory as the gold standard. Overall sensitivity was 86% and 87%, and overall specificity was 92% and 90% in two validation studies. Results show that the mailed, self-reported shelf inventory is a valid measure of the presence or absence of particular foods in households. As such, it may be a useful tool for assessing the impact of point-of-purchase nutrition interventions.