Date of Completion


Thesis Advisor(s)

Valerie B. Duffy

Honors Major

Allied Health Sciences


Dietetics and Clinical Nutrition | Medical Nutrition | Medicine and Health Sciences


Feasible ways are needed to screen for dietary behaviors and taste preferences in clinical settings. We examined the internal validity and reliability of a brief survey to assess food preferences and dietary behaviors among morbidly obese individuals considering bariatric surgery. Survey-reported liking is a proxy for dietary intake, correlating with reported food intake and biomarkers of nutritional status (Sharafi et al, 2015) and linking taste genetics with diet and health (Pallister et al, 2015). Methods –Enrolled were 138 morbidly obese patients awaiting bariatric surgery, who completed a 100-item liking survey of foods/beverages, physical/sedentary activities, and pleasurable/unpleasurable experiences. They were oriented to the survey with examples of activities that could be liked (winning the lottery, succeeding), neutral (doing a routine chore), and disliked (running out of money, paper cut) on a bi-directional, horizontal scale labeled at either end with strongest disliking/liking of any kid and mid-point of neither like/dislike. The survey took <10 minutes to complete. Survey items were averaged into nutritional, sensory (bitter, sour, spicy), and activity groups. The nutritional groups were formed into a dietary quality index (Sharafi et al, 2015) and, with activities, into a behavior index. The indexes had internal reliability (alpha>0.65) and were normally distributed. The most liked items were fruit, pleasurable activities, high fat proteins, and sweet foods (listed from highest); least liked were unpleasurable activities, alcoholic beverages, adventurous foods, and physical activities (listed from disliked to barely liked). In exploratory principal component analyses, >50% of variability in either index was explained by 2 factors—less healthy (sweets, fats, salty, television) and more healthy (vegetables, fruits, fiber, physical activities) behaviors. Patients who reported greater liking for bitter beverages and spicy items had significantly higher dietary quality. Summary: A simple liking survey is a feasible and relatively valid/reliable tool for assessing dietary and taste-related behaviors in a clinical setting.