Home > Memory, Pattern Recognition > “Oh sorry, I thought you were the other Asian…”

“Oh sorry, I thought you were the other Asian…”

asian girls

Learning names is a challenge for most of us, and we all occasionally have that one person we just can’t quite remember. But is there a pattern who we remember and who we don’t? If you belong to a minority, as I do, there’s probably been at least one time when you’ve probably been frustrated and offended because people seem to always confuse you with the few other members of your race. For example, I was one of the only two Asian girls in my high school. This inevitably led to teachers calling me Jane half the time, and Cathy the other half. But should we be offended? Or is there actual legitimacy behind this unintentional racism? The cross-race effect in memory is the idea that humans are actually better at remembering faces that are the same race as them, relative to other races. Hourihan, Fraundorf, and Benjamin further analyzed this effect in their 2013 study on how cross-race effect relates to face memory.

There are some theories about why the cross race effect exists. One theory is that we have more experience in remembering faces that are the same race as us because we have more experience interacting with people of our own race. Other studies add that when we see faces that belong to our race (so for me, a fellow Asian), we focus on features that are specific to the individual (such dimples or the bright shirt they were wearing). But when we see faces that belong to any other racial group (for me, anyone who’s not an Asian), we actually only focus on features that identify the group as a whole (such as a broader nose or a different hair texture/style). If we remember people by the characteristics that are common to the entire race rather than by their unique characteristics, how can we expect to be successful in remembering those individual faces later on? The cross-race effect exists, there’s bias in which faces you remember more easily. But what would happen if someone told you that the face was the same race as you, when it actually wasn’t? Would you still remember that face better when it’s actually not the same race as you?

This is exactly what Hourihan, Fraundorf, and Benjamin asked. To find the answer, they did a study with the following circumstances in order to test this question:

1)   The faces presented to participants were not real faces, but rather faces that were made and generated by computers to be racially ambiguous. This means they are computer generated to have 50% Hispanic features and 50% African American features, with no major racially identifying features (no defining features such as very curly black hair versus brown waves, a broad nose, apparent skin color, etc.) so the face cannot be really identified as African American or Hispanic.

2)   The faces would be presented with social information that would be racially biased. “Social information” is something that offers information indirectly, so in this case the social information would be names. So the participant will see a person’s face and that person’s name at the same time. Names would be racially biased so they would be obviously Hispanic or obviously African American names such as Pedro, Juan, or Tyrone. These names were taken from “Hispanic Baby Boy Names” and “African American Baby Boy Names” lists online.

Given these two conditions, the researchers were able to control the faces as well as what information the participants had about the faces. This allowed them to observe if the cross-race effect would still happen if the faces didn’t actually look Hispanic or African American and only had a Hispanic or African American name to suggest their race.

To test this, they took their list of 80 racially ambiguous faces that were all 50% Hispanic and 50% African American, and did two experiments. In both experiments, they presented 40 of these faces in two sets, the first set of 20 labeled as Hispanic and the second set of 20 labeled as African American. These faces were presented to an equal number of Hispanic and African American participants and they were asked to remember as many faces as they could. After seeing each face, they were also asked to rate on a scale of 1-9 how well they think they will remember the face later on (1 being not at all and 9 being definitely will remember). After seeing all the faces, participants were shown all 80 faces (the 40 that were studied with 40 new faces) in random order and asked to determine if they had studied each face or not, pressing one key for “yes”, and another for “no”.

In the first experiment, they presented the faces with names when participants were learning the faces to help them “learn it better”, but didn’t have the names with the faces when the participants were being tested. They found that Hispanics learned the faces with Hispanic names better and, similarly, African Americans learned the faces with African American names better. This supports the cross-race effect and implies that it’s not just the heightened experience you have with the faces of your own race that allows you to remember them better, but that social information (in this case, racially biased names) during the learning process can also dictate how well you can remember that face later on because we are categorizing those faces as ones that belong to our own racial group.

In the second experiment, everything remained the same except for the fact that the participants were presented with the faces without names when they were learning and rather had faces with names during testing. If social information influences how participants learn the faces, this placement of the names shouldn’t interfere with how well participants remember in this second experiment at all. And they found that this was true; Hispanics remembered the same number of Hispanic named faces as they did African American named faces, and the same occurred with the African American participants.

In summary, even though the participants saw faces that were completely racially ambiguous, racially biased social information such as names was enough to influence memory. Because both experiments in this study used the same set of computer-generated racially ambiguous faces, the placement of names was really the only changing variable. This suggests that the cross-race effect phenomenon not only exists because of the heightened experience we have interacting with our own race but that, even if the faces are the same, mere presentation of racially biased information at the time of learning is also enough to trigger the cross-race effect. Given this new information, although many of us minorities may still feel offended when people can’t tell us apart, at least we can find comfort in knowing it’s unintentional and that it is based on a fundamental cognitive phenomenon. And in applying our new knowledge about the cross-race effect, maybe we all can be more successful in remembering new faces in the future.


Hourihan, K., Fraundorf, S., & Benjamin, A. (2013, April 2). Same faces, different labels: Generating the Cross-Race Effect in Face Memory With Social Category Information. Memory and Cognition.

  1. Michelle Wang
    May 19th, 2013 at 23:57 | #1

    As an Asian girl myself, I definitely feel that the cross-race effect is real in this society. There are times that I was mistakenly remembered as another Asian girl (We are friends and we talked about it before, it’s weird that people mis-remember us since according to us, we looking nothing similar!) I found it really interesting that if associated with a “race-fit” name, it would help to remember and recognize the face. I happen to came across a paper about how shifting attention can reduce the own-race bias. I did my post about faulty eyewitness testimony and it’s related to this effect since the eyewitness have this own-race bias when identifying people from another race and can cause problems like sending an innocent man to jail. This post is interesting and it got me think a lot!

  2. xzhao
    May 15th, 2013 at 15:54 | #2

    Ben: That’s a really good point. I was born in China and didn’t come to the US until I was 6, but have been in the states ever since. So this is actually really interesting to me because ever since I did this article, I’ve inevitably noticed how I actually really struggle with differentiating caucasians with certain features, like girls with blonde hair. When I’ve only met them briefly a few times, I have a lot of difficulty trying to picture their faces in my head and they actually just mix all together and I can’t really seem to tell them apart. Of course, when they’re standing side by side I’m like “oh obviously, you two look nothing alike”, but otherwise it’s still a challenge at first, which is really interesting because the majority of my development has been in the US, and in southern Maine no less, where there’s a real lack of diversity.

    Molly: That’s a really good question, I didn’t think about that. I think it would definitely be present given those circumstances as well, especially when you look at how a lot of the cross-race effect does have to do with the amount of experience you have with what you’re trying to remember and how familiar it is to you.

  3. May 14th, 2013 at 14:44 | #3

    I always thought that the cross-race effect was real (I never knew it was called that, though, so thanks!), and it is good to have that affirmed here. I guess this article supports a kind of “cross-cultural” effect too, since the names that stereotypically fit into one’s culture made the faces easier to remember. Exploring this further would be interesting. I wonder what kind of results they would get if they flipped the experiment around and tried to get people to remember the names that were attached to the faces. Do you think that maybe people remember names that are more stereotypically part of their culture, and have a harder time remembering names that are unfamiliar/uncommon in their culture? If this was true, it would support the “cross-cultural” effect that I was talking about, which would suggest that we remember aspects of people better when they are more similar to our life experiences. Interesting article – it really made me think!

  4. May 13th, 2013 at 14:33 | #4

    This was a very interesting article, Cathy. It really ties back to schemas, like stereotypes, that influence the way we assimilate new information, process it, and remember it later on. Social information that fits into one of these schemas seems to be enough to impact memory and guide assimilation based on just a name attached to the face. Most of my friends from high school are international students, and a majority of them are ethnic minorities. After spending so much time with such a diverse group of people, it now seems easier for me to identify and remember faces of individuals that are not of my race. I wonder if there is any evidence for the effect being reduced for people that have experience in multicultural environment.

You must be logged in to post a comment.