Hill et al. (1976) found that couples who stayed together were not only similar in age and intelligence, but also physical attractiveness than were couples whose relationships broke down. We prefer people similar to us for several reasons: they are likely to share our interests and enjoy the same activities; they validate our attitudes and general view of the world, thereby enhancing our self-esteem; they are easy and relaxing to communicate with.
According to a study by Murstein (1972), people tend to choose a partner of similar levels of attractiveness. He asked a group of people to judge the attractiveness of 99 engaged or courting couples. Another 99 photos of men were placed at random with 99 photos of women. Each photo was rated on their level of attractiveness. The results were clear that the couples who were going out together were rated as more similar in levels of attractiveness than were the couples matched at random.
Silverman (1971) in a field study, observed couples and rated their attractiveness. He confirmed Murstein's findings and noted that, the greater the degree of physical attractiveness; the more physical intimacy was displayed. He carried out a study of 'fait accompli' matching (i.e. matching which has already occurred). Couples were observed in naturalistic dating settings - bars, social events, theatre lobbies. Two males and two females aged 18-22, unmarried, formed the observer team. Each observer independently rated the dating partner of the opposite sex on a five-point scale. There was an extremely high degree of similarity between the attractiveness of the couple members. Also, the more similar their attractiveness, the happier they seemed to be with each other (as reflected by the degree of intimacy).
Berscheid (1985) proposed that all attraction ultimately depends on our evaluation of the other person. If we decide that they are 'good', we like them; but if we see them as 'bad', then we do not. The different factors in liking and attraction, Berscheid argued, have their effect because they make us more likely to evaluate someone positively: it is the evaluation that provides the key.
Berscheid argued that our basic feelings of attraction and repulsion are rooted very deeply in our biological needs. According to Berscheid, we interpret our approach-avoidance tendencies in terms of evaluations: we see things which are safe to approach, as 'good', whereas we see things which we need to avoid as 'bad'.
Many studies, for example Dion (1972), show that people who are physically unattractive are treated and perceived differently by others, than attractive people. This is especially true for suspected criminals. The 'matching hypothesis' predicts that people select partners of comparable attractiveness. This may be to maintain balance (the equity theory), or a fear of rejection, or because of the 'halo effect'. This is when the level of attractiveness rated for their partner, if they have been seen together, may influence the rated level of attractiveness of a person. This is a problem of Silverman's (1971) investigation.
The following investigation will show how substantial the matching hypothesis is. It will see whether people of a particular level of physical attractiveness are married to someone of the opposite sex with the same, or similar, levels of attractiveness.
The aim of this investigation was to determine whether the similarity in the level of physical attraction between males and females was a determinant of marriage. This should be evident if examples of married couples are perceived to share similar levels of attractiveness. The expectation is that there will be a similarity between the level of attractiveness of the husband and the level of attractiveness of the wife, as judged by independent raters. The prediction of this investigation is that two individuals who share a similar level of physical attractiveness are more likely to form closer intimate relationships.
ALTERNATIVE HYPOTHESIS:
There will be a positive correlation between the perceived level of attractiveness judged of husbands from photographs on their wedding day, and the perceived level of attractiveness judged of their wives.
NULL HYPOTHESIS:
There will be no significant correlation between the perceived level of attractiveness judged of husbands from photographs on their wedding day, and the perceived rated level of attractiveness of their wives. Any relationship will be due to chance factors.
THE DESIGN:
- The design used for this investigation is a correlation. This is because the scores that independent judges rated for the husband were plotted against the rates of attractiveness for the wives.
- The data collected from the investigation was numerical and can be placed in rank order. Therefore, the level of measurement is ordinal.
- The inferential statistical test used in the investigation was Spearman's Rho. This is because it corresponds with the level of measurement used and the design of the study. The Spearman's Rho is used when the level of measurement is ordinal and the design is a correlation.
-
The level of significance used was p ≤ 0.05, which is a 5% probability that chance factors were responsible for the results. This significance level is used because it is the conventional level to use in coursework as it strikes a reasonable balance between making a type 1 and type 2 error.
APPARATUS:
Before starting the investigation there were certain materials that were needed and the photographs of the couples needed to be prepared. They were done in the following procedure:
- Ten black-and-white photographs of newly wed married couples from the local newspaper were chosen.
- Each couple were separated into bride and groom and then stuck on two separate pieces of paper.
- Each person on the two sheets was then labelled in order for them to be identified with their original partners again. Both groups were labelled from A to J.
SUBJECTS:
The subjects were then needed to be chosen for the investigation. It is important that they all have similar characteristics in order to ensure that the differences of the participants do not influence the results and become confounding variables, therefore the following groups of subjects were chosen:
- 20 subjects between the ages of 16 and 18 were selected to take part in the investigation through opportunity sampling.
-
The participants were chosen from the 6th form students in a North London School.
- In the investigation a gender balance was achieved as 10 males and 10 females were used.
THE METHOD:
- Each participant was asked to rate each of the brides and grooms for their level of physical attractiveness from a scale of 1 to 10. In this investigation:
1 = extremely unattractive
10 = extremely attractive
- Each participant was taken into a room by themselves and given a score sheet. As this was done the following instructions were given each time:
" Could you please rate each of these individuals for physical attractiveness on a scale of between 1-10. In this case 1 is the lowest level of physical attractiveness that an individual can be rated, and 10 is the highest." These instructions were also written on the score sheet.
- Each subject was left in the room on their own with only the investigator in order to reduce the chance of them being influenced.
- The participants were then given instructions as to what to do:
" Please circle the number that you think represents the level of attractiveness of each person. Try to be as accurate and as objective as possible.
These instructions were also written on the score sheet. (This can be seen in the Appendix).
- The subjects were then handed a score sheet and asked if they understood the
instructions. If so, then they were asked to start on part (i) of the investigation.
- They were then given the following instructions once they had finished this:
"In part (ii) please match up the husbands to the wife that you think they married. There is a ten minute time limit for this part."
- Each participant was then asked if they had understood these instructions and then
asked to proceed.
- While each subject did this part they were timed to see how long they took. This
was then recorded.
- Once they had completed the score sheet, each subject was thanked for taking part
in the experiment.
THE TOTAL PERCEIVED LEVEL OF ATTRACTIVENESS OF MALES:
THE TOTAL PERCEIVED LEVEL OF ATTRACTIVENESS OF FEMALES:
THE PERCEIVED LEVEL OF ATTRACTIVENESS OF MALES AND FEMALES IN EACH COUPLE:
CALCULATING THE STATISTICAL TEST - SPEARMAN'S RHO:
Formula: Rs = 1 - 6 ε d2
N (N2 - 1)
Therefore: d2 = 83.5, N = 10
Rs = 1 - 6 x 83.5 Rs = 1 - 501
10 (100 - 1) 990
Rs = 0.506
The box above shows that after calculating the Spearman's Rho statistical test,
R = 0.5.06. This has then got to be compared to the critical value of a one tailed hypothesis with a significance level of 0.05:
C.V. = 0.564
The rules of calculating Spearman's Rho states that R must be equal to, or exceed the critical value for significance at the level shown.
I am then able to conclude that the coefficient is not significant. This means that I will be accepting my null hypothesis. That there will be no significant correlation between the perceived level of attractiveness judged of husbands from photographs on their wedding day, and the perceived rated level of attractiveness of their wives. Any relationship will be due to chance factors.
The results from the Spearman's Rho formula show that the correlation coefficient of 0.506 was not significant. The calculation was carried out using a one-tailed hypothesis, a significance level of 0.05 (by convention) and when N = 10. This gave a critical value of 0.564. In order for the results to have been significant the value of Rs needed to be more than, or equal to, the critical value.
Due to the results of the formula indicating no significant correlation, the alternative hypothesis was rejected. Therefore, the null hypothesis was accepted. This stated that there would be no significant correlation between the perceived level of attractiveness judged of husbands from photographs on their wedding day, and the perceived level of attractiveness of their wives. Any relationship will be due to chance factors.
From the investigation I determined that the perceived level of physical attractiveness of females is far greater than the perceived level of attractiveness of males. There are many studies that give reasons for this. For example, women usually chose partners that are less attractive than they are. Huston (1973) suggested that people were afraid of being rejected by their prospective partners. They deliberately choose someone who is similar to them, not because they find them most attractive, but because they don't want to be rejected.
Another possibility for the result is that even though an equal number of male and female subjects were used, females are often more open to rating people of the same sex. They find it easier to say exactly how they feel and so may have rated the couples honestly as to how they feel. However, males are often reluctant to rate other males so highly. This may be due to them being oblivious to what women find attractive in men, or they may feel that if they rate other men a high score then they would be perceived as holding homosexual tendencies.
There are a number of additional factors that might have caused the females to be perceived as more attractive than the males. For example, the photographs of the couples were of their wedding day. This would make both the bride and the groom, however especially the bride, look more attractive. This is because on wedding days the bride is especially groomed to look her best - and so usually looks much better than that in everyday life. This is, therefore, a confounding variable, as she is perceived to look much better than she actually does.
There were several limitations of the investigation that I carried out. For example, the photographs that the 20 participants were rating were actually black and white and had been photocopied several times. Therefore, the photographs were not of as good a quality as they could have been. To get the best results it would have been better if the photographs were in colour, or for even better results, the individuals could have been rated in person.
Another drawback was that the photographs were all of couples on their wedding days. This would have distorted the results as most usually, the bride and groom look much better on this day, than in everyday life. This is because they would have been groomed to look their best. Therefore, the photographs were not a true representation of what the couples looked like, and were misleading. This would, inevitably, lead to the results being higher than they would have been if it had not been photographs of couples on their wedding day.
Also, a factor that could have had an adverse effect on the results could have been the fact that there was no inter-rater reliability. This is because the scores that were given to each individual were all carried out in a subjective manner. There were no criteria under which the judges followed in order to rate the couples. Therefore, in order for the results to be completely accurate a rating scale should have been produced - in the attempt to make the investigation more reliable.
Another reason why the results may have been distorted, as the inaccuracy of the results might have come from a type 2 error. That occurs when the null hypothesis is accepted, but the results may really be due to the independent variable. A way of reducing the chance of a type 2 error from occurring is to decrease the level of significance from 0.05 to a more stringent level of 0.01.
When carrying out the investigation there had to be many controls. This is because the participants were using demand characteristics. They were trying to give me the results they thought I wanted by trying to match up the backgrounds of each picture to identify the couples. After carrying out a pilot test - where there was an unlimited time limit - the participant took 20 minutes to fill out the score sheet. In this time she was able to identify each couple by determining which of the backgrounds of the females matched up to that of a particular male. This factor led me to believe that there should be a ten minute time limit when filling out the score sheet for each participant. This did not stop the demand characteristics, but made it harder for the participants to pick the right couples according the their backgrounds. The only way it would have completely stopped was if I had tried to remove the faces of each individual from the background of the photographs.
When taking into account ethical considerations, it can be argued that the couples, whose photographs were used, did not consent to taking part in the investigation. Also, as the photographs of the couples were taken from a local newspaper, the couples' next of kin may have been part of the investigation, which means that they would rate the couples on more than physical attractiveness. Also, the couples themselves may not appreciate being rated for attractiveness by people of the local community. In order to avoid embarrassment, photographs of couples from a different community or borough should have been taken.
To extend the research on this investigation, further work could be carried out in order to emphasise or contradict the results found. For example, the subjects could be taken to a common place where couples would be seen in order to rate the individuals without using photographic representation. Also, instead of rating individuals who are all part of a couple, some people could have been rated who were not married. This would test the similarity factor to a greater extent.
Another factor that could be taken into account if further research was carried out would be, like Murnstein's studies (1972), the sample of the individuals to be rated should have been larger. This is because it gave Murnstein's results more credibility and could, therefore, be more safely generalised to the whole population.