Date of Award


Degree Name

MS in Agriculture - Food Science and Nutrition


Food Science and Nutrition


Amy Lammert


The objectives of this study were: (1) to evaluate consumer acceptance of cheeses varying in fat and sodium levels, (2) to determine if sensitivity to reward and body mass index has an effect on product liking based on fat or salt content, (3) to evaluate the use of FaceReader technology during consumer evaluation and, (4) to determine if consumer’s self-selected, conscious emotions matched with the expressed, subconscious emotions acquired by FaceReader.

Consumer acceptance testing (n=108) was conducted on two medium cheddar cheeses with varying fat levels and two low-moisture part-skim mozzarella cheeses varying in sodium levels. Attributes were measured using a 9-point hedonic scale. In order to measure reward sensitivity, participants completed the BIS/BAS questionnaire and the SPSRQ prior to consumer acceptance testing. SIMS sensory software was used for data collection. The complete consumption experience was video recorded (n=83). A choose-all-that-apply format was used so participants could indicate all emotional states before and after consumption. A total of 332 pairs of videos (83 subjects, four samples, before and after consumption) were used for FaceReader analysis.

Regular cheddar cheese scored significantly higher than the reduced fat cheddar cheese for mean overall liking, flavor, texture, creaminess, saltiness and aftertaste. The higher sodium mozzarella scored significantly higher than the lower sodium mozzarella for mean flavor, saltiness and aftertaste (p

FaceReader Results indicated:

  1. Neutral was the most accurately matched self-selected emotion (100%) before and after consumption, followed by happy (82% and 63% respectively). FaceReader was unable to correctly match surprised/angry before consumption and angry/sad after consumption.
  2. FaceReader acquired 420 and 495 additional non-self-selected emotions before and after consumption, respectively. Neutral and angry were most commonly expressed when not self-selected. Disgusted and scared were rarely expressed when not self-selected.
  3. FaceReader was not as successful matching the self-selected emotions after consumption. Surprised and happy were commonly missed both before and after consumption. Disgusted was missed primarily after consumption.
  4. "Happy" is self-selected and expressed more times for regular cheddar than the reduced fat cheddar. The mean overall liking score was also significantly higher for the regular cheddar than reduced fat cheddar. Similar results were found with mozzarella.

Although low fat and low sodium cheeses represent a healthier option, consumer acceptance indicated that the higher fat and higher sodium samples scored higher; changes in flavor and texture need to be made in order to produce a more liked product. There is a complex relationship between product liking, body mass index, gender and sensitivity to reward but further research needs to be conducted to investigate how the variables interact.

FaceReader technology did match some of the self-selected emotions identified by the subject. However, one question remains: which emotions, self-selected/conscious emotions or subconscious/expressed emotions, are a better predictor of liking?

Included in

Food Science Commons