The Neuroscience Department welcomes Patrick Simen from Oberlin College. He will be giving a talk titled "A Random Walk Model of How We Track Time and Make Decisions."
Abstract: Can we really infer anything about how the brain is organized, just by examining how long it takes animals — including humans — to respond to stimuli? Researchers have made a surprising amount of progress doing just this in recent decades. One result of their efforts is a class of mathematical models that explain the accuracy and speed of animal decision making, including its well-known variability, using processes called "random walks". A stock market share price that fluctuates over time is an example of a random walk. You could make decisions to buy or sell a stock based on this fluctuating price using the following rule: if the price ever descends lower than some level, then sell; if it ascends higher than some higher level, then buy. That might be a terrible financial strategy, but as a basic decision process — a process to decide, for example, whether a colorful stimulus is closer to red than to green — this sort of process is often the best way to decide. Brain imaging and neurophysiology have supported this notion of random walks plus decision thresholds as the basis for simple perceptual decisions in the brain. My colleagues and I have further extended the range of phenomena this model explains, to the problem of when to act. I will present this model of decisions, and of time tracking, and show the results of some experiments my students and I have conducted to test its predictions.
No recent activity