Study of Perceptions of Attractiveness Based on Race
Recently, many studies have been conducted to find the effects of images viewed in the media on personal attitudes and behaviors. Factors that have been analyzed in relation to amount of media consumed include acceptance of gratuitous violence and sex, fear of everyday life and so forth. One aspect that has been observed in many psychological studies involves self-esteem in those who do not physically fit the images that are constantly presented in the media. One study that demonstrates this phenomenon found that women experience a sort of self-loathing due to images presented in the media (Roberts & Waters, 2004) Another study demonstrated that being different from what appears to be the accepted way of looking can have countless negative effects (Camartin, 2003).
This discussion is important because not fitting into the tiny niche that has been carved in western media can have deleterious effects on attitudes, which may lead to eating disorders, excessive plastic surgery, and self-hatred (Perrogan, F, 2003). It has been demonstrated that different body sizes presented in the media have a direct effect on how women view their own body size (Dittmar & Halliwell, 2004). This study presented images of thin, average, and large models to a group of women and had them rate their own bodies and the effectiveness of the advertising.
It was found that women definitely were affected by the appearance of thin models but importantly, the effectiveness of each advertisement was found to be the same no matter what size model was used. This suggests that advertisers can use more representative people in their advertisements and not suffer major losses.
To further show that images in the media are important for how people perceive themselves, a study was conducted on recruitment and salience of representatives in an advertisement. The study found that minority groups were better able to relate with minority figures than people whom they did not consider themselves similar to (Avery & Hernandez, 2004). Conversely, there appeared to be no differential effect for Caucasians used in the study. A possible explanation for this is that since minorities are so underrepresented in the media, it is difficult to relate to what is shown and so a minority representative will be more effective than a non-minority.
For the sake of this discussion, a euro-centric media refers to the fact that an overwhelming majority of figures presented in the media are of European descent. This issue is of special interest since the population rates of minority groups is steadily increasing and western media is becoming dominant in countries outside of the United States (Lydia, 2003).
A study conducted outside of the United States by Darling (2004) looked at the effects of western media on opinions of female attractiveness in Japanese women. The experiment was set up to determine what sort of effect western media would have on Japanese women, in particular, those with a strong sense of culture and who would strongly critique what they saw in the media.
The experimenter sat and talked with each of the Japanese women while they looked at Japanese fashion magazines. Caucasian models were presented the most in these magazines and the women who looked at these images all said that Caucasians are simply more attractive than the Japanese. Ergo, 29 Japanese women were studied in depth for eight months and it was determined that they possessed the same ideals of attractiveness that are found in western society. Also, when the women rated themselves, it was discovered that they compared themselves to the western ideal. Most of the women said they wish they were taller and had larger eyes, for example (Darling, 2004).
This study highlights the fact that a euro-centric media can have invasive effects on cultures that are geographically and culturally distant from the United States. If western media can have such a strong effect overseas, it is of even further importance to study how people rate attractiveness in relation to race here in the United States. A study by Makkar & Strube (1996) involved the ratings of one’s own attractiveness after viewing three images of white or black models.
In support of the hypothesis contained in this proposal, it was found that both white and black participants that viewed the white models found themselves to be less attractive in comparison and those who viewed the black models did not have as much difficulty perceiving themselves as attractive.
The main shortcoming in this study is that participants were not asked to rate the models, only themselves. This does not provide information on whether or not people view whites as more attractive than non-whites, which is the purpose of this proposal. Although, it does signify that people engage in a comparison process with what they see in the media and since whites are depicted much more often, that standard of beauty may be more influential as also found in the Dittmar & Halliwell study (2004).
An important aspect to consider in a discussion on physical attractiveness is whether or not there is a biological basis for it. Past studies on this topic have shown that there is a biological basis for facial symmetry and averageness. To test this assertion, Clark et al. (2001) tested Chinese and Japanese subjects and had them rate computer generated faces for attractiveness. They found that, similar to westerners, Japanese and Chinese people preferred faces that were symmetrical and average. Furthermore, the study cited the fact that there was no racial gap in the ratings.
The basis of this study is good for determining what is considered attractive but it did not do enough to find if there are racial differences in attractiveness. There needs to be more variation in the races of faces presented in the study. Basically, the study established the fact that there is a biological basis for attractiveness so it follows that equally symmetrical and average faces of all races should be judged equally attractive. One study that came close to testing this assertion is mentioned below. Cooper et al. (2004) tested many women to see what differing characteristics they would use to describe black and white men’s faces. They looked at traits such as gender identity, dominance and pertinent to this study, facial attractiveness. One hypothesis of the study was that white faces would be perceived as more attractive and this hypothesis was supported. Therefore, this study was able to find a discrepancy in ratings of attractiveness according to race.
The Cooper et al. study is a beginning for what this study concerns. It succeeded in testing differences of opinion on attractiveness between two races but this factor needs to be studied in greater depth to come up with the conclusion that a euro-centric media is having a strong effect on people’s ratings of physical attractiveness. Essentially, the study needed to be conducted on a wider range of people. Most people involved in the study were Caucasian so that does not fully test if whites are being perceived as more attractive than non-whites because of the media.
The purpose of this study is to determine if there are racial gaps in what is considered attractive amongst different races. Past studies have determined that media may play a role in perceived attractiveness and that different traits are more likely to be found in some races than in others. However, past studies have been lacking in testing if many people view whites as more attractive than non-whites. Therefore, this study hypothesizes that different races of people will perceive Caucasian faces as more attractive than any other race.
Method
Research Design
The experiment used four different racial groups and measured and compared how each racial group rated members of other racial groups. A computer based program measured each group’s ratings of attractiveness.
Participants
The study consisted of 60 male and female college students at the University of Minnesota between the ages of 18 and 25. Signs were posted informing students of the study in various buildings throughout the University of Minnesota campus. Most people who responded to the postings were selected but this was dependent upon their race since we needed equal groups. In order to select students appropriately for this study, a brief phone interview was conducted. Questions were basic biographical questions asking for information such as major, year in school, etc. The only pertinent question for the purposes of this experiment in the survey was “What is your racial background?”. Participants were told they would receive either ten dollars or three research experience points, which count for extra credit in their courses.
Procedure
Potential participants for this study were interviewed over the phone in order to determine their racial background. We were looking to attract a total of 60 participants with 15 being described as “White”, 15 as “Black”, 15 as “Hispanic”, and 15 as “Asian”. After interviewing each candidate, if they fit into the category we needed they were scheduled for a time to come into the lab. We scheduled 60 candidates at first and then an additional two from each race to account for absenteeism.
The actual study began taking place two days after the phone interviews. We scheduled five subjects for each time slot and we had 15 different hour-long time slots over two days. Subjects were not assigned to time slots by race but how they chose in order to accommodate their schedule. Once participants entered the lab, they were seated and then were told to complete a brief written survey asking for information on amount of television they viewed, what they watched, where they have lived in the past, and who primarily raised them.
After the five participants completed the survey, they were placed into a second room that contained five computers. Students were told to sit at the computers and begin reading the on-screen instructions. The instructions guided the participants to begin the study. The actual study consisted of random faces being shown one at a time on the computer screen. The program that was used for this was developed by the experimenters and is detailed below.
To create the computer program for the experiment, a panel of 100 college students at the University of Minnesota was used in order to standardize ratings of attractiveness for photographs of faces. Foreseeing the purpose of this study, the panel was composed of 25 “Whites, 25 “Blacks”, 25 “Hispanics”, and 25 “Asians”. This panel viewed 300 faces, 200 were actual people and 100 were computer-generated mixtures of several different faces. Only four racial groups were used for the faces as defined by this study. The 100 computer-generated faces contained equal numbers from the four racial groups and the 200 actual faces also contained equal numbers of faces from each racial group.
The panel viewed all 300 faces as randomly presented on a computer screen one at a time and rated the faces for attractiveness on a scale of one to five. A score of one was “very unattractive”, two was “somewhat unattractive”, three was “neither attractive nor unattractive’, four was “somewhat attractive” and five was “very attractive.” After scores were calculated for each of the 300 faces, 100 of the faces that received the most consistent scores were selected for use in the study. 25 from each racial group were taken with each group containing the same distribution of scores on ratings of attractiveness. For example, the “Hispanic” group of faces had five faces that got scores near five, five faces that got scores near four and so on. The overall means for the 25 faces within each racial group varied insignificantly from 3.67 for “Asians” to 3.69 for “Whites” which showed we had succeeded in standardizing the faces.
The program in the actual study contained 100 faces standardized for ratings of attractiveness. 25 faces from the four relevant racial backgrounds were used and they were presented randomly on the computer screen. Each face was presented for 1.5 seconds and participants had to type a number between one and five on their keyboard within that time. If a participant failed to respond within the allotted time, no score was recorded and the face was redrawn randomly a second time. This was done to gauge immediate reactions to the attractiveness of a face and disallow time for thinking and reconsideration. After each participant came to the conclusion of their computer program they were given a briefing on the computer screen and told to leave quietly. Participants then selected what form of compensation they desired. When all participants had finished their ratings, the experimenters retrieved the scores from each computer and recorded which of the four racial groups the participant belonged to.
Results
This experiment was conducted in order to possibly suggest that the disproportionate euro-centric images presented in the media may lead people to view those of Western European descent as more attractive than people from other ethnic backgrounds. Males (n=22) and females (n=38) were divided into four racial groups: Asian (n=15), White (n=15, Black (n=15), and Hispanic (n=15) were used to arrive at the following results.
After a two-day process of running participants through the experiment and calculating the data, a one-way analysis of variance was used to compare how members of each racial group rated faces from the four racial groups. The mean ratings of each racial group, for each racial group are presented in figure 1.
Discussion
The results showed that there are significant differences in ratings of attractiveness by different racial groups. We discovered that those who are represented in the media the most were the ones who received the highest attractiveness ratings from each of the four racial groups. In this case, whites received the highest ratings from each of the four racial groups tested and Asians received the lowest scores from each of the racial groups except themselves. These findings fall in line with hypotheses put forth in past research (Makkar & Strube, 1996) and go against evolutionary explanations for attractiveness as illustrated by Clark et al. (2001).
The implications of these findings are that there seems to be a cultural bias present in how people are perceiving and judging others. Research in the field of evolutionary psychology has demonstrated that humans have a preference for symmetrical and average faces so it can be assumed that a symmetrical and averaged Japanese face should be perceived by virtually all people as more attractive than an asymmetrical Hispanic face, for example, but this and prior research has demonstrated otherwise (Donnely & Valentine, 2004) (Clark et al., 2001).
This study aimed to conclusively determine if different races have differing perceptions of attractiveness for the different races present in the world. It was discovered that there is a bias present, therefore the underlying reasons for this need to be discovered so that current problems such as low self-esteem and suicidal ideation can be eradicated from the minds of those who do not perceive themselves as attractive due to a narrow and unrealistic definition of beauty. This study was effective in determining that there are unjustifiable differences in how races are perceived in relation to levels of attractiveness but there were also some shortcomings of this study.
One major limitation of this study was that it did not create separate conditions to fully illustrate that the media is a cause for the significant differences of ratings of attractiveness for the different racial groups presented in the study. This study involved a brief questionnaire that assessed the amount of time spent watching television as well as the content but not enough attention was put on this part of the experiment. If we had perhaps assigned participants into groups based on amount of television watched this would better account for the effects of media on ratings of what is considered attractive.
Another limitation is the potential inability for males to rate other males as attractive. In this society, it is generally considered a homosexual activity for males to rate other males and so, as evidenced in our study, males may not fairly rate the faces of other males. In the future, it might be a consideration to test only females since they seem to be more comfortable with rating members of the same sex.
A third problem with this study was the sample size. Only 15 members of each of our defined racial groups were used in this experiment. This may not be enough to eliminate the inability to extrapolate our findings to the general population. Furthermore, the “Asian” group had two participants that were adopted and raised by Caucasian families. The fact that these participants grew up mostly around non-Asian people may aid in explaining why they rated racial groups other than their own as more attractive. This interesting finding for adopted participants is something that should be studied in greater depth in another study. While the study contained a few limitations, some contributions to the field of research were made.
Past research on this topic failed to adequately test more than one racial group when determining if a racial group was generally considered to be more attractive than another (Cooper, et al. 2004). This study was the first to take four different racial groups and have them rate the attractiveness of the same four racial groups. This design allowed us to determine how much of a difference existed between one’s racial background and their perceived attractiveness of both their own race and of different races.
Also, this study demonstrated that one race is typically perceived as more attractive than other races, by other races. This evidence contradicts the evolutionary explanation for attraction as well as the mere exposure effect, which states that simply being around someone will contribute to attraction (Williams, 2004). With the results of this study, it is hoped that future research will conclusively determine that the media and perhaps other underlying causes are behind the distorted perceptions of attractiveness found in cultures that are exposed to westernized media.
As a follow-up to this study, we suggest that people exposed to westernized media be compared on their ratings of attractiveness with people who have never been exposed to westernized media. By doing this, it can be determined that it is indeed the images presented in the media that are changing how humans view the world in which we live.
It would also be fruitful for future research in this area to duplicate what was done in this study but create two conditions, one that is exposed to many images in the media and one that is not. From there, it can be determined if images recently viewed in the media will have an effect on how people are rating different faces categorized by race. Furthermore, research could present afro-centric media to one group and euro-centric media to another group and see how this affects their ratings of attractiveness.
