A young woman walks down a city block as night begins to fall. A middle-aged man heads her way. Before their paths cross, the woman steps aside to increase the distance as they pass.

"We all do it without thinking. In just seconds we size people up and others do it to us," says 20/20 correspondent Michael Guillen.

"How reliable are these snap judgments that our brain is always making? Are first impressions reliable."

With help from UConn professor of psychology David Kenny and some of his graduate students, 20/20 discussed these questions during a November 6 broadcast on ABC-TV.

In early September, Guillen and a 20/20 crew came to the Storrs campus to attend Kenny's social psychology class.

"When the students saw the cameras," says Kenny, a professor of psychology, "they were really excited. How often do you go to class and find network television filming."

It also happened to be the first day of class, so many students were meeting one another for the first time.

Based on their first impressions, students in the class were asked to rate six people whom they believed were also students in the class but who in fact had been recruited for the experiment.

A week before school started, Kenny's graduate students visited a number of different areas of campus, including Jonathan's, the Co-op, fraternity and sorority houses, to recruit the six individuals.

"The students explained to these people that they might not like what they heard," said Kenny.

When the six subjects of the experiment showed up in class that Thursday morning, they appeared to be just more students who had signed up for the class. Kenny told the class to judge the six "students" simply by looking at them.

Were they hard-working or lazy? Warm or cold? Serious or frivolous? Although Kenny pretended that the selection was random, in fact, his real students were judging the pre-selected subjects.

The students filled out questionnaires that were tabulated immediately. Then the six left the room, and Guillen interviewed the students about the people they had rated.

Were the first impressions accurate? To find out, Guillen then met with friends and family of the people rated.

Of the six subjects, four were "hits," meaning that four out of six judgments made by strangers corresponded with how the family and friends saw that person.

"Most of us believe that first impressions are wrong and inaccurate," Kenny says, "but this experiment and a series of others suggest that very often first impressions have a surprising degree of validity."

We should trust them, he says, even though they may not be totally accurate.

Kenny has been doing research on first impressions since the mid-1980s. So when 20/20 asked him to be a part of the story, "It was scary," he says, "because you realize they would be taking years of work and summarizing it in a minute or two."

Kenny was pleased with the results, however. He says "20/20 did a great job of portraying the conclusions that I and my graduate students have made."

Luis Mocete