Advertise - Print Edition

Brandeis University's Community Newspaper — Waltham, Mass.

Brandeis researchers find older adults still recognize subtle facial expressions

Published: November 8, 2013
Section: News, Top Stories

Professor Leslie Zebrowitz (PSYC) and Robert Franklin of Zebrowitz Face Perception Lab conducted research comparing the level of emotion overgeneralization effects in older adults and younger adults. They published their findings in the paper “Older Adults’ Trait Impressions of Faces are Sensitive to Subtle Resemblance to Emotions” in April 2013.

It has been previously found that first impressions are affected by emotion overgeneralization, which refers to people’s tendencies to attribute emotion-related traits to people whose neutral faces resemble a certain emotion. For example, if a person’s neutral face resembles an angry face, people are more likely to perceive this person less likable and hostile even though this assumption might be completely invalid. It is theorized that this overgeneralization is a byproduct of our ability to appropriately perceive and respond to emotional expressions, which is an important survival skill.

For this study, the Zebrowitz lab compared their subjects’ evaluation of a series of photographs of neutral faces with the evaluation of a computer program that was trained to analyze which emotion the neutral face actually resembled by measuring and calculating the face through its characteristics. When they compared the young adults’ answers with the computer’s answers, they found that young adults could effectively recognize the subtle resemblances of emotions in neutral faces. For example, the more hostile or unfriendly they thought a neutral face looked, the more resemblance to expression of anger the computer calculated the face to be.

Given the knowledge that older adults show deficit in accurately labeling strong emotional expressions, Zebrowitz thought that perhaps older adults wouldn’t show this bias as much as young adults. Amazingly, their study showed that older adults are just as effective as young adults in recognizing the subtle emotional resemblances in neutral faces.

“There’s no difference. Older adults, just like younger adults, see faces that the computer thinks look more angry, and think they’re more hostile, and less trustworthy; they think the faces that the computer believes show more surprise, actually look more naïve,” Professor Zebrowitz said. This means that although older adults show deficit in identifying real emotional faces accurately, they are very sensitive to subtle variations among neutral faces that resemble emotions.

There are a number of different explanations that they have been pursuing, Zebrowitz explained. One may be that multiple choices invoke a different kind of thinking that older adults perform worse on. Another explanation is that they use a different kind of processing, automatic or controlled, and because the parts of the brain that control automatic processing are more preserved with age, older adults show no deficit in emotional overgeneralization.

Older adults usually do worse on tests that measure speed of processing and executive control; they usually also have poorer visual cue, and perhaps less contrast sensitivity. There is a lot of evidence that older adults function very well in interpersonal relationships, and because of these test results that show lower cognitive function, this social ability has been a puzzle to scientists.

Zebrowitz expressed that this study is important because it that shows social cognition functions are intact in older adults. “Maybe we shouldn’t generalize too much from what happens in cognitive tests that don’t have meaningful social content or the cognitive tests that doesn’t really represent what goes on in everyday life,” she said. This study shows that when older adults face a social cognition task, they perform just as well as young adults do.

The Zebrowitz lab has since furthered their research and published another paper that studied the accuracy of people’s first expressions in regards to aggression. They found that people did show accuracy greater than chance in correctly judging how aggressive a person is. Professor Zebrowitz further mentioned that the lab is currently studying how accurate people are in predicting a person’s health and competence based on appearances.