Card sorting tasks have been commonly used to measure issues of importance in market research. We sought to compare results that would be obtained when using a card sorting task with results using a more traditional rating scale in an online survey. Respondents were asked to indicate how often they consumed a series of 20 different foods or beverages. This served as criteria to determine the validity of the sorting task or rating scale. Respondents were then either asked to sort the 20 different foods into 1 of 5 categories which were labeled from 'Very much like' to 'Very much dislike' or were asked to indicate how much they liked each of the 20 different foods using a grid format. We found some distributional differences due to task and, though the card sorting task took longer, we found little difference in the average correlation with the key criteria.