Marshall Roedel and Dr. C Brock Kirwan, Pschology and Neuroscience

Current memory theory suggests that two separate processes contribute to recognition memory. Recollection is when recognition is accompanied by contextual information regarding the encoding event. Familiarity is when recognition is only a companied by a vague sense of having encountered something previously.

Memory researchers theorize that certain stimulus features, like word concreteness, may affect the recollection process without interacting with familiarity. Word concreteness refers to how material the object that the word describes is. For example, desk is a very concrete word as opposed a highly abstract word such as mercy. My purpose was to determine if word concreteness interacted with familiarity as well as recollection. To do this I conducted two experiments: one behavioral and one using EEG data.

For the behavioral data we tested the subjects’ ability to identify previously studied words embedded in a list of novel words. Some of these words were highly concrete while others were highly abstract. Using various analyses we were able to obtain quantified estimates for the contributions of recollection and familiarity. We saw that both of these estimates were greater for concrete words than for abstract words. These finding contradict past conclusions that recollection is sensitive to word concreteness while familiarity is not. We wrote a paper on these findings and sent them to the journal Learning and Memory. It was rejected so we sent it to Experimental Psychology: Learning, Memory & Cognition. We are waiting to hear back from reviewers.

An electroencephalogram (EEG) can detect electrical changes in an active brain. Stereotypical patterns or components are associated with certain process. For example, a frontal negative going trend that occurs 400 milliseconds after stimulus presentation known as the FN400 is theorized to index familiarity. We gave subjects a similar recognition memory test to the one in the paragraph above while taking their EEGs and observed the FN400 when they encountered studied words. The FN400 for abstract words was significantly less negative than it was for highly concrete words. If the FN400 really does index familiarity, this would suggest that concreteness interacts with familiarity just as the behavioral results did. However, as we were analyzing the EEG data I came across a study that suggests that the FN400 does not index familiarity but semantic processing. We are now conducting an EEG study to investigate this claim.

I presented the results of both the behavioral and the EEG data at the Snowbird Neuroscience symposium in Salt Lake on October 21st. I will also be presenting them and the Society for Neuroscience conference in Washington DC in November