Videos footage from RuCCS Colloquium Talks can be found on the RuCCS YouTube Channel. For all other events, please check the sponsor's website for more detail.
To filter by event category, click on the event category link in the table below or use the menu on the right.
List of Past Events
Reconstruction from Memory in Naturalistic Environments"
Dr. Pernille Hemmer
Thursday, September 27, 2012, 12:00pm - 07:00pm
Rutgers University, Department of Psychology
Many aspects of our experiences do not have to be explicitly remembered, but can be inferred based on our knowledge of the regularities in our environment. Suppose you witnessed a crime and are asked to recall the height of the perpetrator. You might not only utilize the episodic information related to the witnessed incident. Your recall for the height of this particular person might also be influenced by general knowledge about heights of people, as well as specific knowledge about the height of men and women.
In recalling a variety of episodic events (e.g., the height of people, objects in scenes and the order of events) people appear to use a strategy of incorporating prior knowledge to improve average recall performance. This behavior can be explained by a rational model of memory that assumes that people seek to optimize recall performance by combining the available episodic and semantic information. In a series of experiments I quantify people’s a priori expectations and memory for the heights of men and women. In the context of this data I introduce a simple rational model to explain the effect of prior knowledge on memory.
While rational models are useful to describe behavior, it is however, unclear whether people always perform rationally within a given recall task. In the work presented here I seek to extend the rational model of memory to allow for individual differences. I present a framework using Bayesian estimation procedures within the rational model, such that Bayesian Data Analysis is applied directly to the observed data as a method to learn about the underlying psychological variables of the rational model. In this way, Bayesian inference is used both as a model of the mind of the participant and for analyzing the data from the perspective of the experimenter. I will present two different ways to evaluate individual differences in the rational model using this recursive approach, a mixture model allowing for two different groups of participants, and a continuous model to infer the subjective prior for each individual subject.