Our initial sensory representations are fleeting and forgotten in a matter of seconds, at most. Our work has demonstrated that even the memories we intend to maintain are lost in a similar fashion unless some time is given for attention to dwell on the internal sensory representation. This attentional dwell leads to a more robust trace that is resistant to future memory decay. We call this process short-term consolidation. At present we are investigating the nature of short-term consolidation's contribution to the creation of a memory trace within working memory, the relationship between short-term consolidation and a phenomenon known as the attentional blink, and several other questions related to these processes.
What are the mechanisms that lead to forgetting during multitasking?
There has been a debate about the causes of forgetting in short-term/working memory since the emergence of cognitive psychology as a discipline. We investigate several aspects of this debate in relation to multitasking while actively maintaining memories in the working memory system. One line of work examines whether interference or distraction is the primary driver of forgetting. While it could be that processing the secondary task actively disrupts memory traces, it could also be that memory traces are forgotten because they are not actively attended during secondary task processing. We also examine issues related to the nature of the memory traces themselves. For example, do high-detail memories and more gist-based categorical memories suffer from similar multi-task related impairment? What are the effects of memory items and concurrent processing of another task sharing modalities or features? How does the nature of the memory set impact performance on a secondary task?
Using eye tracking & pupillometry to understand attention allocation during working memory tasks
Gaze location can tell us what people are thinking about in real time. The location of an individual's gaze is strongly related to what they are thinking about under normal conditions. Pupil size can be used to understand task difficult during task execution. We use these measures to understand the temporal dynamics of attentional allocation during memory creation and multitasking to answer questions that are difficult to address with standard behavioral methods.
How can we improve mathematical models of visual working memory?
Traditionally, memory recall has only been possible with verbal material where participants can simply speak or write the correct response. In a visual memory task, recall would have involved participants drawing a shape or some similar reconstruction method. This response method introduces a set of confounds that make theoretical conclusions difficult under most circumstances. More recently researchers have begun to use computerized recall methods with visual working memory tasks that allow a participant to reproduce a color, orientation, or shape by selecting a value within a continuous circular space. The use of this response method allows us to create mathematical models of recall that give much more information about what participants are able to remember than does simple accuracy measures.
The original visual production models were very basic, with few parameters and minimal theoretical assumptions, characterizing guessing patterns and memory precision. For example older models assume that participants are always remembering the exact shade of red presented to them, with some noise. We seek to improve these models of memory by examining how categorical memory states play a role in addition to memories for the exact values of color or orientations presented. We assume that sometimes people ignore the exact color-shade or orientation of a memory item and simply remember that an item was red, blue, etc. Understanding this gist-based memory has important implications for human behavior.