Summary to be provided by Sean
The verbal overshadowing effect: Why descriptions impair face recognition
No TrackBacks
TrackBack URL: http://www.psychologicalscience.com/cgi-bin/mt/mt-t.cgi/184
TrackBack URL: http://www.psychologicalscience.com/cgi-bin/mt/mt-t.cgi/184
Verbal overshadowing is a concept that I didn’t even know existed before reading this article. This article did an excellent job of defining the term and studied three different experiments. When I first read through this article, it was a little difficult to keep track of all the variables each experiment was testing. After reading the discussion of each experiment and also the general discussion of the entire article, it made it clear on what the experimenters were exactly doing.
Basically, the overshadowing effect is when a person fails to distinguish between memories of visual and verbal information. In experiment one, experimenters simply replicated the verbal overshadowing effect, but was not able to replicate other tests that had differences under speeded conditions. They also were surprised to find out that the verbal overshadowing effect did not occur when they instructed the participants to be aware of their visual and verbal representations.
Experiment two tested some of the results from experiment one bringing up issues of source confusion and also holistic vs. featural shifts. This experiment’s results stated that the subjects were less accurate with describing the target’s face after they described their parent’s face. It also determined that reading a description impairs face recognition. Lastly, it replicated the first experiment that subjects were not helped when ignore instructions were given.
Experiment three tested the recognition rates for male and female faces. This test concluded that verbalization impaired recognition of the non described faces as much as it did of the described face. Also note that there was no difference between the effects of describing male and female faces.
After looking at the results to all three of experiments, many people can conclude that there needs to be more research on the topic. Overall, this article gives evidence of verbal overshadowing to be a problem in eyewitness identification. I think this article provides an excellent quote that jurors and other officials need to remember, “Remembering is a product not only of what is stored and what is activated at test, but also of how that information is evaluated (Dodson, Johnson, Schooler 1997). “ KC
The fact that two different types of memory can interfere with one another has been examined in cognitive research for quite some time. With verbal overshadowing problems occur when the verbal memory integrates into the visual memory for a face, causing different retrieval outcomes than either memory would elicit independent of one another. As such, it is important to consider this somewhat overlooked aspect of human memory when in applied settings such as legal circumstances involving eyewitness identifications.
Though it makes since that a person’s own verbal description of a suspect would interfere with their visual representation of a face in memory, I found it interesting that Dodson, Johnson, & Schooler (1997) discovered how another description of the face, given from an external source, would reduce and eliminate the verbal overshadowing effect. It seems that verbal information should still interfere with the visual representation in memory, but perhaps knowing that the source of the verbal description came from an external party allowed them to still access the visual representation of the face stored in memory.
The explanation for the verbal overshadowing stemming from the role of holistic and featural processing resonated with me, especially given my recent research pursuits with the inversion effect for faces. It makes since that when providing a verbal description, one is listing different features that they remember from the face, thus leading to a featural processing strategy of the face (known from face recognition literature to lead to different results in terms of accuracy compared to holistic processing).
I don’t remember reading about this from any of the articles we read from Sean’s discussion on verbal overshadowing, but I wonder if you were using other-race faces, what would happen with the processing strategies. If we already use a featural processing strategy for other-race faces, then the visual representation in memory should be the result of the same type of processing as the verbal based memory associated with providing a description of the face. Both processes will still probably lead to poor recognition accuracy, but of particular interest is that this would be different for same- race faces. With same-race faces the holistic strategy should dominate for the visual memory process, but then the featural processing is involved in the generation of the verbal description. Doesn’t really bring us closer to understanding why the CRE occurs, but it does suggest that the verbal overshadowing is less of a problem for other-race faces, because they are already processed in a different manner from that of same-race faces at the initial visual level. I believe Meissner and Brigham (2001) mention this in their meta-analytic review article if you are interested in learning more about the verbal overshadowing effect as it relates to the cross-race effect.
DJP