Up until the late 20th century, influential decision theorists did not consider recent events (e.g. previous choices or outcomes) relevant to risk taking, in part because risky decision-making settings are not inherently contextual or temporal in nature. For example, an outcome on a previous trial does not causally influence or directly alter the outcome on a subsequent trial. However, over the last three decades, research spanning psychology, economics and neuroscience has begun to explore the possibility that risky decision-making may instead be fundamentally contextually sensitive. Using a combination of approaches, including computational modeling of behavior, physiological arousal, and fMRI, my research examines how risky monetary decision-making is temporally context-dependent. I will present results from a series of studies that demonstrate both value-dependent and value-independent effects of temporal context on risky choice behavior at multiple timescales. These data also suggest that risky choices reflect not only temporal context, but how people compare that context to their evolving expectations (i.e., a dynamic reference point). These results are perplexing because relying on recent events at any timescale appears to be at odds with the assumed goal of risky decision-making: to maximize payoff. I will discuss potential mechanisms, including physiological arousal and the hypothesized neural mechanisms, that may support context effects in risk. Finally, I will present a new project leveraging insights from emotion regulation and cognitive control research to understand how using cognitive strategies to change goals may mitigate such temporal context effects in an effort to improve risky decision-making.
