Across three papers and seven experiments, I test predictions based in fuzzy-trace theory that individuals' decisions will be affected by a combination of reward sensitivity (a neurobiological construct related to increased activity in the dopaminergic reward pathways in the brain), inhibitory mechanisms which have been associated with prefrontal areas of the brain, and mental representations of the decision problems. In the first chapter, we report that subjects showed standard framing effects (i.e., risk taking for losses and risk aversion for gains) regardless of Cognitive Reflection Test (CRT) score if expected values are equal. However, verbatim calculation trumped the gist in those with high CRT scores when expected values were unequal. The second chapter reports tests of the prediction that qualitative representation of options or health state will be preferentially relied on compared to quantitative representations, which is tested through manipulation of quantitative risk information and of qualitative health status quo to examine the effect on likelihood of starting a medication. Controlling for side effect, quantitative risk, age, sex, ethnicity, and numeracy in a multivariate model, the status quo and all three qualitative (i.e., ordinal) risk dimensions (worry, risk perception, and overall gist of options) predicted likelihood of taking the medication, while objective risk and numeracy did not. Finally, I test a model that predicts that global assessments such as life satisfaction will endure as gist representations, whereas tasks requiring precise judgments, such as numeric ratings of happiness, will activate verbatim representations that are subject to greater interference over time, resulting in less consistency between ratings in the moment and recollections of happiness judgments. We support these predictions by introducing a new measure of instability that captures the level of verbatim detail incorporated into judgments of well-being, and corroborate a growing body of research that demonstrates that focusing attention on verbatim details can have detrimental effects on judgment in real-world outcomes. These results are consistent with a theoretical mechanism in which the perception of the gist of choices, as well as individual differences in reward salience and neurobiological responsiveness, each account for unique variance in predicting risk-taking.