Although the phrase “they all look alike to me” is widely known and used as a joke

(Goldstein & Chance, 1985) the implications that the phrase yields may not be as widely

understood as they should be. In the criminal justice system, eyewitness identifications can play a

major role in the decisions made by jurors even though it has been shown that jurors have little

awareness of factors that affect the reliability of eyewitness accounts (Cutler, Penrod, & Dexter,

1990). One such factor is identifications made by people identifying someone of a different race,

(when describing race as Caucasian, of African descent, of Asian descent, or Latino),

which can be referred to as other-race identification or other-race recognition (Lavrakas, Buri, &

Mayzner, 1976). If the phrase “they all look alike to me”, which states that people of another

group appear homogenous, is true for most people then it is possible that other-race

identifications should be considered less reliable than within-race identifications. It has in fact

been shown that people are more accurate at identifying others of their own race than at

identifying others of a different race (Penrod, Shapiro 1986). Since it has been

established that the other-race recognition is not always as accurate as within-race recognition,

finding the cause for this effect is important. One possible cause is the quantity of experience a

person has with another race. People are less accurate at identifying faces of a different racial

group than at identifying faces of their own racial group because they have less experience with

the other race. We would expect that as the amount of experience a person has with a different

racial group decreases the less accurate they will be at identifying faces of that race. It would also

be expected that if you raise someone’s level of experience with a different racial group, possibly

through training, the ability to identify faces in that racial group should increase.

The amount of experience a person has with another racial group should determine how

accurate they will be at identifying people of that race. Those who report having large amounts

of contact with another race should show better accuracy in identifying people of that race than

those reporting small amounts of contact with the other race. In a study done by John Brigham

and Anne Maass, 64 white and 9 black 17-60 year old clerks were asked to identify from photo

lineups customers who had been in their store earlier. Although the clerks showed little own-race

bias and were correct less than half the time, the white clerk’s ability to identify the black

customers was substantially related to the amount of self-reported cross-racial experience they

had. But the black clerks didn’t show the same relation. It could be possible that the previous

experience that the white store clerks had with black people may have been positive or negative

experiences. These experiences may have caused them to pay greater attention to black

customers. For example, if the store clerk was robbed by a black customer, he may have a

prejudice against them and now watches them more closely. Here, it is not just the quantity of

experience someone has with another race that can affect accuracy of identification, it is also the

quality. To judge whether the quality of the experience may have affected the accuracy, a

structured survey can be given to the store clerks which asks them specific questions on what type

of contact they have had with the other race, not just how much. In another study of the effects

of experience on the accuracy of face recognition, Paul Lavrakas, John Buri and Mark Mayzner

tested the ability of white subjects to identify black faces. Both quantity and quality of experience

are considered. To measure these, subjects were asked for their familiarity with blacks
(quantitative experience) and their current number of black friends (qualitative experience). They

found that the quality of the white subjects experience with black people was more important

than the quantity. It was also found that having black friends was more positively related to the

recognition of black faces than having grown up in a integrated neighborhood. But, it is possible

that the whites who currently have black friends are more likely to have positive attitudes about

blacks. This positive attitude could either lead to them coming into greater amounts of contact

with other black people, or it could be the positive attitude itself that aids them in the testing

process. By having black friends, these white subjects may have come into as much contact with

blacks as those who were raised in integrated neighborhoods. Also, a large proportion of the

subjects with black friends also may have had contact with blacks while growing up. While the

white subjects, who were raised in integrated neighborhoods and currently have either

zero or few black friends, may be more likely to have negative attitudes towards blacks which

could cause them to perform poorly even though they have a large amount of experience with

blacks. It is also possible that if after growing up in an integrated neighborhood, the reason that

certain white subjects don’t have black friends is because they chose to segregate themselves from

blacks and therefore never actually gained the experience they would be expected to have. If the

subjects are going to fill out a questionnaire asking about their experience with blacks why not

also ask about their attitudes towards blacks to see if it is affecting the results.

It may be possible that it is a third factor affecting the other-race recognition testing.

Perhaps it is a increase in experience that decreases the belief in out-group homogeneity. The

more experience the white store clerks had with blacks the less they believed they were all the

same, which increased their ability to identify them. One way to test this would be to have
subjects place faces in either a white, black or neither category, with a given prototype for both

races. Whites should show a inclination to group all black variations together if they are affected

by out group homogeneity. Another possible third mechanism is that it is different facial features

that different races have that are unfamiliar and hard to distinguish. But research shows that no

racial group can be characterized as more or less physically homogeneous than others (Goldstein

& Chance, 1976). Nevertheless, a possible way to test this would be to test for other-race

recognition patterns using pictures of faces that have been tinted or lightened, or altered pictures

that give, for example, a white face eyes that are more typically associated with Asian faces.

If experience increases the ability to recognize other race faces, then people from an

integrated neighborhood should be more accurate in other-race recognition than people from a

segregated neighborhood. One study found that white children from a segregated neighborhood

showed a higher differential race recognition than children from an integrated neighborhood

(Cross, Cross, & Daly, 1971). Another study of children from integrated schools showed small

own-race bias, but the findings were not entirely consistent. It could be that the children were not

fully developed in their ability to identify any particular race better than another. More years of

experience might yield a greater result. Also, tests could be done on the ability of people to

recognize other race faces over a period of time as an area becomes more and more integrated. It

should show that as integration increases the ability to recognize other-race faces also increases.

If it were truly the case that experience with a particular race was the cause of the differential

recognition, than a white child raised in a black home should recognize black faces better than

white faces and vice versa. Unfortunately I was unable to find these studies.

To increase the level of experience a subject has with a racial group, training can be done,
which should increase the ability to identify faces in that racial group. A study done by Alvin

Goldstein and June Chance took subjects who initially showed poor ability to remember Japanese

faces. Some were intensively trained and a control group received no training. The trained

subjects showed marked improvement on tests with sets of new Japanese faces. Here, after

gaining experience in Japanese faces, the subjects increased their ability to identify faces in that

racial group. Though the results are mostly straightforward in this study, it should be noted that

only the ability to identify Japanese faces was tested. It cannot be concluded that due to

experience with an unfamiliar race, the subject is now familiar with that race. The subjects may

actually have just been trained to become better at the test, or simply better at identifying faces

of any race based on an increased awareness of facial features. I would suggest also testing the

subjects on a different racial group, other than their own, but giving them no training in that

group. If their ability to identify that group also rose then it is the task itself that the subjects

improved upon.

Although not all of the evidence points to experience as the cause of the other-race effect

on face recognition, it is clear that experience at least some type of role in the ability of people to

identify others of a different race. Also, future experiments would benefit from a more clear and

concise measure of experience, whether it is quantity or quality, to increase the validity of the

tests. Future work should also examine whether white children raised in black homes are better at

identifying black faces, as we would expect them to if experience was the cause of the other-race

effect. Continued research in the subject would benefit the criminal justice system in its

understanding of the reliability of eyewitnesses because although much of the data supports the

thesis the presence of possible alternative explanations prevents the drawing of a firm conclusion.
REFERENCES

Brigham, John C., & Malpass Roy S., (1985) The role of Experience and Contact in the Recognition of Faces Of Own- and Other-Race Persons. Journal of Social Issues, 41, 139-155.

Lavrackas, Paul J., Buri John R., & Mayzner Mark S., (1976) A Perspective on the Recognition of Other-Race Faces. Perception & Psychophysics, 20, 475-481.
Shepherd, John, Deregowski, Jan B., & Ellis, Hadyn D., (1974) A Cross-Cultural Study of Recognition Memory For Faces. International Journal of Psychology, 9, 205-211.
Goldstein, Alvin G., & Chance, June, (1985) Effects of Training on Japanese Face Recognition: Reduction of the Other-Race Effect. Bulletin of the Psychonomic Society, 23, 211-214.
Anthony, Tara, Copper, Carolyn, & Mullen, Brian, (1992) Cross-Racial Facial Identification: A Social Cognitive Integration. PSPB, 18, 296-301.
Shapiro, Peter N., Penrod, Steven D., (1986) Meta-Analysis of Facial Identification Studies. Psychological Bulletin, 100, 139-156.
Cutler, B. L., Penrod, Steven D., (1990) Juror Sensitivity to to Eyewitness Identification Evidence. Law and Human Behavior, 14, 185-192.