This research aims to contribute to a better understanding of one factor in female physical attractiveness to males– eye color. Specifically, the goal of this study was to look for a subconscious preference for blue or brown eyes in 18-49 year-old males’ mate selection. Because of previous hypothesis’ that suggest that reading the dilation of the pupil can play a key role in detecting emotions, a preference for a lighter eye, in this case blue, is expected. Twelve stock headshots of women, 6 having brown eyes, 6 having blue, were chosen and printed. The eye colors of these women were then digitally changed to the opposing color (blue to brown, brown to blue) and printed off again. These two sets of 12 photos (altered and original) were brought to a public location where they were judged by randomly selected males via survey (attached). The average score of each ‘original’ photo was then compared to that of its ‘altered’ version. The findings suggest that blue eye color is statistically preferred, but exceptions exist, pointing to outside factors.
Discovering the effect of eye color on female attractiveness is important because it can potentially further our knowledge of sub-conscious human judgment and the subsequent relational effects of that judgment. The blue-eyed females are predicted to be selected as more physically attractive than their brown-eyed counter-parts because pupil dilation can be seen more clearly in a light colored iris. Being an indication of attraction, natural selection would have undoubtedly leaned towards the preference of these lighter irises.
Methods of Research
- A. Setting
a. The surveys were distributed outside of the entrance of ‘The University Book Store’ on State Street in Madison, Wisconsin on a Saturday afternoon, a particularly busy time in the area. The distribution method was semi-random. Every third male that appeared to fit the pre-determined age range of 18-49 was asked to complete a survey for either the ‘A’ or ‘B’ sets of photographs. The following prompt was used consistently for each potential surveyor:
“Excuse me Sir. I was wondering if you could do me a favor and participate in a survey”
If the man accepted, he was given the ‘A’ or ‘B’ set of photographs, a pen, and one copy of the accompanying survey. He was not given any other direction than what was given to him on the survey itself. This was done to ensure that the men weren’t tipped off as to what they were supposed to look for specifically. The men’s opinion on the women in the photographs’ ‘overall attractiveness’ was what was most important. This is as close to a natural reaction to the woman’s face possible without physically having the woman present.
b. The time and location were chosen because of the potential high density of young males. A football game was played less than an hour before, and as the men leave the game many of them go to State Street to celebrate. This higher population of people that fit the characterization of my preferred surveyor increased the likelihood that enough data would be collected about the photos. I chose to ask every third passer-by to complete the survey to avoid asking multiple men in the same peer group. Completing the survey with others could compromise the surveyors’ honest reactions to the photographs. The outdoor environment and early afternoon time also provided consistent lighting, which is a significant factor because the eye color of each photographed woman must be evident for the study to be viable.
c. The vast majority of the men who were asked to participate in the survey declined. This lengthened the process quite a bit more than expected. But, because of the high flow of people, that wasn’t a problem in the end. A small percentage of the men who originally agreed to participate backed out upon looking at the survey itself. I suspect this had something to do with being put into what they thought to be an embarrassing situation.
d. I think my approach and request could have been improved. A physical incentive and a less formal delivery may have been more appropriate for the situation. Meaning, my target crowd, being of a younger age and leaving a football game, may have been more ‘casual’ than my pitch. Overall, however, I think the setting worked fantastically. The desired fifty surveyors were found within a relatively short time period and without much discomfort on either end.
B. Independent variable: The color of the eyes.
a. Eyes in the ‘A’ group were left untouched. Eyes in group ‘B’ photos are altered to have the opposing eye-color. This is done opposed to merely comparing the women’s natural eye color because it insures that other factors of attraction, like hair, skin tone, face shape, etc., aren’t a factor in the results. The results are calculated by measuring the average increase or decrease in scores between two pictures of the same woman.
b. While both of those markers leave room for creative interpretation, I felt as if giving each number a definition would give the men a more unified view of the scoring range. The possible differences in opinion regarding the definitions of the extremes on the scale should have been offset by the shear number of men who completed the survey, anyway.
c. Because the results show the males made a clear distinction in attraction when the eye color was changed, the operationalization was successful. The consistent change in ratings shows strong evidence for iris color being a determining factor in female attractiveness.
d. The results indicate that the procedures worked very well. I think the results could have been strengthened in respectability if my Photoshop editing abilities were improved. That said, I did average the scores of all of the “original” pictures and all of the “altered” pictures. The difference was two hundredths of a point; an incredibly miniscule difference all things considered. That would indicate that my editing abilities, while amateur, were good enough to fool the participants in the survey.
C. Dependent Variable: The ratings given by the participants
a. A 1 to 10 scale of attractiveness was used by each surveyor to indicate how physically attractive he found each woman in the photographs. The survey (attached) defines 1 as, “repulsive” and 10 as, “most beautiful woman in the world”. The survey also clearly indicates that the man can choose whether or not to use decimal points in his scoring.
b. This was chosen because each threshold, 1 and 10, could easily be defined for the participant. It was also the choice method of ratings collection because it’s a fairly common system and participants would most likely be comfortable with it, having experience with similar systems in the past.
c. Bias is only a problem if you feel as if the definitions that were placed at the near and far end of the rating’s scale were drastically different than what the participant would have assumed. Other wise, this method of data collection is left almost entirely to the participant. The question is a simple one and it’s asked in a straightforward manner, hopefully ensuring consistent reactions and results from all parties.
d. Again, the results would indicate that the operationalization worked very well. Among the scores given to each woman, there’s rarely a figure that differs by more than a number of 2 from another. This indicates a consistency in the interpretation of the scale across participants.
e. There should have been a definition given to “5” on the scale in the instructions on the survey to give the participants an idea of what could lie in between the extremes I laid out on the ends.
D. Experimental Control
a. The most important factor that was maintained under experimental control was the effect of the women’s other features on the men’s scoring. By changing the eye color and gathering information using woman ‘twin pairs’, the effects of the woman’s hair color, lips, nose, skin, etc. are eliminated. This is the key to verifying the effect of Iris color. The delivery of the original pitch to the prospective participant remained constant throughout. The survey’s format and ratings system never went through any changes.
b. The control taken over the non-iris facial features was a massive success, as a quick glance at the averages among “twin pairs” shows difference. This is all but absolutely due to the only element that was changed, the iris. The survey questions’ consistency was obviously vital, and its never-changing instructions and rating’s system pulled through.
c. The procedures were good because they helped highlight the legitimacy of the final results by removing other possible variables. By maintaining consistency in all but one part of the experiment, that one part becomes the factor we can gather real information about. Without full control, questions about the validity of your claims that this sole factor was the cause for change can be questioned.
- Unit of analysis
The findings in this experiment only serve to reinforce previous studies concerning the effect of visible pupil dilation. I do, however, think there are a lot of questions that can be asked in the future that can directly branch off of this study. One of which could be a search for commonalities in the physical characteristics of men with a preference for blue eyes. Perhaps the greatest flaw in my study was the ethnic demographic in the photos, and the ethnic makeup of the participants; the vast majority of which were Caucasian. This observation is important because