Michael E. Young - US grants

People face numerous decisions throughout the day, but nowhere are these decisions as critical as when they occur under conditions of stress, complexity, and time pressure. Choosing among a range of possible causes of explosions, weird noises, malfunctioning machinery, or unusual readings on instruments is difficult, especially when there are multiple competing candidates with varying time signatures, likelihoods, and intensities. Previous research on causal judgment has focused on controlled experiments, but everyday decision making occurs under less pristine conditions. To examine the extent to which prior approaches apply in complex dynamic situations, the proposed studies examine behavior within the context of a video game that involves multiple causal candidates, many visual perspectives, and varying degrees of complexity. Participants in the study will be faced with battlefield decisions regarding the source of enemy fire or other threats.

The theoretical basis for these studies is that causal judgments rely on the subjective certainty regarding whether an effect will occur and when it will occur. The experiments are designed to (a) evaluate the relative effectiveness of various methods of bridging delays between causes and their effects (given that delays are known to be detrimental to effective source identification), (b) assess theoretical predictions relating to the integration of information about contingency, endogenous temporal variability, and exogenous temporal variability, and (c) determine whether a proposed theoretical model extends to situations in which a cause and its effect are continuously related (e.g., the proximity of a cause determines the magnitude of its effect). A critical contribution of this project is the creation of a video game environment for studying decision making. This platform will be shared with other scientists to encourage the development of theories of dynamic decision making.

DESCRIPTION (provided by applicant): The proposed research extends prior work on impulsive behavior and delay discounting. When choosing among options with different delays to their outcomes, people and other animals devalue later outcomes relative to sooner ones. Delay discounting is thought to underlie much deleterious behavior. People eat highly caloric desserts, charge trivial goods and services, take illicit drugs, smoke, or drink excessively without regard for the long term consequences of those momentary pleasures. People presumably make these decisions because the value of the delayed consequences is discounted. Improving our understanding of the basic principles of behavior underlying the socially significant decisions may improve our ability to address those problems. The present application extends prior work by assessing delay discounting in a dynamic environment where reward amount grows continuously with wait time. Historically, delay discounting has been most frequently studied using discrete trial choice procedures. Subjects typically make a series of choices between smaller sooner and larger later rewards (e.g., "Would you prefer $100 now or $150 in a month?" for people or 2 pellets immediately vs. 4 pellets at a 10 second delay for rats). Although the utility of these procedures is unquestioned, they represent but one type of relationship between behavior and proximal and distal outcomes. Thus, we hope to extend the delay discounting account of impulsive behavior to an environment where the relationship between action and outcome is more continuous and dynamic in real time. Humans will perform within the context of a 3-D immersive video game in which firing a weapon sooner will produce less damage than firing it later, whereas rats will be trained to hold down a lever to increase available food amounts upon its release. That is, we will assess when the subject will "cash in" and take the immediate reward rather than continuing to wait for a larger reward. In addition, we will also assess how reward probability affects delay discounting in this task. The proposed methodology produces significant challenges to quantitative descriptions of behavior and a new framework is offered to predict behavior. Unlike discrete trials procedures, reward rate will vary with performance (fire rapidly with little damage per shot vs. fire slowly with a lot of damage per shot). We hope to expand models of delay discounting to describe real- time interaction with the environment. PUBLIC HEALTH RELEVANCE: The proposed research will help us to understand when people will behave impulsively and when they will not. A general tendency toward impulsive decision-making has been correlated with smoking, substance abuse, pathological gambling, and engaging in HIV risk-related behavior. Understanding the reasons why people make impulsive choices may inform therapeutic and educational interventions to address those socially significant behavior problems.