Say you saw something unusual—such as a blue strawberry or a purple cat. You’d engage with it more, hoping to make sense of it. Psychologists have recognized this tendency for many years. Even infants stare longer at an object they find surprising. We have found that people will also use more words to describe something that defies their expectations for others rather than conforming to those expectations. We call this phenomenon the “surprised elaboration effect.”

This effect can reveal hidden or subconscious biases about people from different racial or ethnic backgrounds, according to research by my colleague Lauren Eskreis-Winkler of Northwestern University and me. In the U.S., society often links Black and Latino communities with people living in poverty or in a dangerous neighborhood. These associations become the seeds for implicitly biased thinking: without bad intentions, people start to expect the worst for individuals from minority backgrounds. In a series of studies, we found that when people are asked to write about situations that run counter to expectations linked to race or ethnicity—a good thing happening to a member of a racial or ethnic minority group, for instance—they use a lot more words. The length of their responses points to stereotypes that they may not even think they carry but that nonetheless influence their thinking.

To understand how stereotypes spur surprised elaboration, we studied public records to compare reports related to individuals of different races and ethnicities. For example, we reviewed 1,051 missing-child posters created by law enforcement agencies in California, Texas, Florida and New York State. The state records identified these children as white, Black or Hispanic.

When we analyzed the reports, we found that posters about white children were 30 percent longer than the others. This difference was not a function of relative rarity: no one group of children was statistically more likely to have gone missing. What was going on? We suspect that surprised elaboration was at play. In the U.S., many people associate white people with better life outcomes and Black and Hispanic people with more negative experiences. So the reports may reflect the writers’ conscious or unconscious stereotype that white children are less likely to go missing than Black or Hispanic kids. Their sense that a missing white child was an unusual and surprising event led them to write more about it.

Another set of public records revealed a similar pattern. When medical examiners wrote reports of unidentified bodies, these reports were 20 percent longer for white individuals, compared with Black or Hispanic people. Once again, we suspect that people were more surprised by unidentified white individuals than they were by Black or Hispanic ones, prompting them to write more.

To explore this idea further, we designed a few experiments. Working with more than 1,200 people and several different scenarios, we presented participants with a few basic details—such as a photograph and brief description of a situation—and then asked them to write up some form of report. The scenarios varied: for example, participants might have learned that the individual photographed was a teacher who had received an award or a professor who was recently caught taking drugs. In some cases, we also asked people how surprised they felt about the particular situation.

What we found repeatedly was that participants wrote more—indicating their surprise—in situations where negative events were associated with white individuals and positive events were associated with Black individuals. For instance, people wrote 25 percent more about a white teacher, as opposed to a Black teacher, who was fired for sexual harassment. Another group wrote 30 percent more about a Black teacher, as opposed to a white one, who won an award. Participants were more surprised when white teachers were at fault and when Black teachers won a prize. The number of words they wrote revealed the societal stereotype that white people have more positive life outcomes. We discovered the same effect of surprise on elaboration in a sample of all-Black participants, who wrote more about a white college professor accused of coming to class while on drugs than a Black professor in the same scenario.

Putting this evidence together, we start to see how surprised elaboration can become a tool for identifying and documenting society’s stereotypes. People don’t usually admit their negative expectations for social groups, even to themselves. We do not believe that the bulk of writers involved were consciously expecting less of Black or Hispanic people. Still, we would argue that the length of their writing reflected their recognition of counterstereotypical events such as a good outcome to a member of a minority group or a bad outcome to a member of a majority group.

This form of surprised elaboration can have significant consequences. In another set of experiments, we presented more than 400 people with reports of various lengths—such as unidentified body or missing child reports—and asked them to assign each a priority level that would be used to help determine government spending to address that case. We found that longer reports got higher-priority scores from people, and they wanted the government to spend more to solve those cases.

We even created two missing children and two unidentified body reports that were nearly identical in details but that varied in length. We found that 64 percent of our respondents preferred to funnel resources toward the case that featured a wordier report. Lengthy descriptions signaled greater importance to the reader. Because our earlier work hints that longer missing child and unidentified body write-ups generally involve white victims, our finding raises the troubling possibility that people may prioritize those cases as a function of their longer reports.

Writers and speakers, whether journalists or law enforcement or public officials, should be mindful when discussing a negative event that has happened to a member of a minority group. When we say little about these incidents, we should ask ourselves whether we are less surprised and hence assume these events are somehow less interesting or lower priority.

As readers and listeners, we should recognize this potential bias in communication so that we can think critically about how stereotypes might in turn enforce structural barriers. We can remember that what others say is just one part of their message: how much they say can be every bit as revealing.