Doctor of Philosophy (PhD)


Nutrition and Food Sciences

Document Type



Visual and cognitive cues can affect overall liking (OL) and consumer perceptions, emotions, and behavior. The first study explored the effect of product color difference on the liking, perception, and purchase intent (PI) of cheese-flavored-tortilla chips (CFTC) formulations (A and B) on serving plates (plastic, foam, and paper). Color differences between formulations influenced crunchiness and saltiness liking and perception, which together with overall flavor liking and formulation, mainly determined CFTC OL. Although having similar fracturability (N) and sodium content, formulation A had higher crunchiness and saltiness likings. PI was influenced by crunchiness, saltiness, and OF liking with 37, 49, and 60% increases in PI odds per liking-unit increase, respectively. Plate effect on product liking was minimal. The brighter and less-yellow color of CFTC possibly influenced crunchiness and saltiness liking, which significantly contributed to OL and PI. Sustainable and nutritious edible insects are unfamiliar to Westerners and often associated with negative sentiments. The second and third studies evaluated the effects of disclosing edible-cricket protein (ECP) presence and benefits on chocolate brownies (CB) expected and actual sensory acceptability, consumption intent (CI), PI, sentiments, and variables importance for PI prediction. ECP added to chocolate brownies [0% ECP=CBWO (without) vs 6% ECP=CBW (with) w/w], and disclosed information [no ECP added=(-) vs ECP with benefits=(+), ECP- and ECP+, respectively] yielded CB treatments (CBWO-, CBWO+, CBW-, and CBWO+). Subjects (N=112 female and N=98 male) rated expected and actual likings, selected emotions before- and after-tasting, and determined CI and PI after tasting. Likings were analyzed with mixed-effects ANOVA and post-hoc Tukey’s HSD test. Emotions were evaluated with Cochran’s-Q test and correspondence analysis. Emotions driving/inhibiting mean overall liking (OL) were assessed with penalty-lift analyses using two-sample T-tests. A random forest algorithm predicted PI and estimated variables' importance. Female’s and male’s expected OL were higher for CBWO- than for CBWO+. Females’ actual OL was higher for CBWO than for CBW regardless of the disclosed information but males’ actual OL was identical across treatments. Females exhibited negative-liking disconfirmation for CBW-. In both tasting conditions, the disclosed information affected treatments’ emotional profiles more than formulation. After-tasting “happy” and “satisfied” were critical PI predictors.



Committee Chair

Prinyawiwatkul, Witoon