Midwest Cognitive Science Conference Snack Food Poster
I have been working on the snack food project since its creation. I have been a part of data analysis and all other aspects of the study. I was able to create a poster and present it during the Midwest Cognitive Science Conference. This is my first research conference outside of Miami University and I am honored that I got invited.
Abstract
A healthy diet is an important part of living, so understanding how healthy choices are made is of pivotal. There are many important food attributes and food choices that face decision makers. We investigated college students preferences for 16 common snack foods with a variety of attributes including health as well as taste, convenience, familiarity, and satiation. We looked at choices among pairs of single options, as well as between two 50/50 grab bag gambles composed of two snacks each. Decision-making has historically been focused on the result of the choice or preference. In our study, we included process tracing methods such as mouse movement, eye fixation location, and self-reported assessments of snack foods and their attributes, in addition to the final choice. We present results illustrating attention, valuation, and weighting of options and attributes during snack food choice; the effects of similarity on choices, attention, and response conflict; distinctions between pairwise choices among individual options and those with probabilistic outcomes; and the effect of variability among outcomes (or risk-aversion) in the latter. Our results both validate the assumptions of process-tracing methods and extend our understanding of the cognitive processes underlying decision making. These results can be used to predict the outcomes of individual decision makers when faced with familiar snack choices, by informing both utilitybased and computational models of decision making. Ideally, this understanding could potentially target interventions to increase healthy eating habits, and our methodological approach can be applied to many broad choice domains.
