Silverman, (1971) conducted a field experiment which, like Murstein’s investigation, provides indirect support for the Matching Hypothesis. Couples were covertly observed in naturalistic dating settings (bars, clubs, theatres) by two males and two females; unmarried and between the ages of 18 and 22. Each observer rated the opposite sex partner of the couples on a 5-point scale. It was found that couples’ attractiveness ratings were often very similar, and it was noted that the more similar the ratings given to a couple, the happier (as measured by level of intimacy displayed) they were together.
Formulation of Aims
The aim of this study is to replicate Murstein’s aforementioned investigation into the Matching Hypothesis. In doing so the purpose is to find further evidence for the Matching Hypothesis. My modifications of Murstein’s experiment include: participants will be rating strangers in photographs rather than themselves/partners, the ratings will be delivered using a 10 point scale, all of the people in the photos are either engaged or married. It is predicted that couples will be rated very similarly as they have chosen eachother and assuming the Matching Hypothesis is correct, they have selected eachother on the premise that they are of a similar level of attractiveness, either to avoid rejection and/or maintain balance. Due to a fair amount of supporting research, I have chosen to employ a directional hypothesis, in favour of the Matching Hypothesis being true.
Statement of Hypotheses
Alternative Hypothesis: There is a positive correlation between attractiveness ratings of separated photographs of married or engaged heterosexual couples.
Null Hypothesis: There is no correlation between attractiveness ratings of separated photographs of married or engaged couples.
In this investigation, I will be testing the alternative hypothesis as much of the research referred to in the introduction, including the study on which my investigation is based support the validity of the Matching Hypothesis.
Methodology
Design and Participants
Data was collected using a ‘visual questionnaire’ whereby participants had to rate photographs using to a scale (1-10). Correlational analysis was employed.
The researcher was the author of this report. I conducted this research individually.
Participants used were AS Level psychology students whom were all aged 16 to 17. Participants were recruited using opportunity sampling. In total, they numbered 20; comprising of 13 female and 7 male individuals. Participants completed the visual questionnaires individually, in a different room to the remainder of the sample.
Confounding variables
Participant reactivity bias was minimised by placing the researcher at a sufficient distance from the participant during scoring. Possible researcher bias, e.g. researcher commenting on the research/stimulus materials, was reduced by the deliverance of a standardised set of instructions read in a neutral tone of voice. The instructions are detailed below.
Control Procedures
Participants remained in a ‘holding room’ before and after taking part in the study. This was to prevent discussion of the true aims of the investigation being held before all participants have had their responses recorded; which would have negated the internal validity of the study. All participants were debriefed together in the holding room immediately after the study.
Ethical Issues
To ensure participant confidentiality, no names were recorded during the questionnaire and to prevent pressure being put on the participants (re: scrutiny of their responses to the stimulus materials) the researcher did not look at the response sheets as the participants were completing them. Participants were asked to record their gender on the sheet simply to give numbers of each gender partaking. All participants were debriefed to deal with the minimal deception incurred in this study. Since the people in the stimulus photographs consented to having their pictures printed in a nationwide publication, it was assumed that they give their consent for their image to be in the public domain.
Materials
Stimulus material: 2 a4-sized cards each with 10 slightly larger than passport-sized photographs taken from women’s weekly magazines. The articles from which the photos were taken clearly stated that the couples were married or engaged. One card contained the female photos, arranged in rows and labelled A to J. The other card had male photos, labelled in the same way. These 20 photos were originally 10 photos of couples and these had been separated so that the participants could rate each half of the couples hopefully without guessing the aim of the experiment. The face and neck of each person had been cut out of each picture so that clothes and background were not included, as subliminally, colour preferences could have influenced the participants’ judgements. The 20 separated photos were then mixed up (after having recorded who the couples were, of course) and arranged in rows onto each card, glued on, labelled and laminated. [Included in the appendices.]
A double-sided response sheet for each participant to record their ratings of the photographs. [Also in the appendices.]
Standardised Procedure
1 Standardised instructions are read aloud to the participant. (see appendices for a specimen)
2) If the participant does not want to continue, thank them for their time, if they do, proceed.
3) Ask if they have any questions regarding the procedure.
4) If they do, respond to their questions, but do not disclose the true aim of the experiment. If they enquire further as to the aim of the experiment simply reiterate what was stated in the standardised instructions: that you are investigating levels of attractiveness and remind them that they will be debriefed later. Any deception now will be dealt with later in the debriefing session.
5) Give the participant a blank response sheet and a pen and also stimulus sheet 1 of the female photographs. Ask them to rate each photograph using the point scale as described in the standardised instructions. Ensure that you, the researcher sits far away from the participant such that it is obvious that you cannot see the participants’ responses as they are writing them.
6) Once they have finished, take away sheet 1 and give them stimulus sheet 2 containing the male photographs. Repeat instruction 5).
7) When the participant has completed their responses for both male and female photographs, give them a square of paper with a number (the number represents the order the participant came in only) and also write this number on the response sheet. In the eventuality of a contributor wanting to withdraw their results, their sheet can be taken out of the data set without having to have recorded their names on the sheets.
8) Repeat the whole procedure exactly with the remainder of the participants, one by one, directing each participant back to the holding room. Once everyone has taken part, carry out the debrief within the holding room. Hand each participant a debrief sheet (included in appendices) and read a copy aloud also. Ask if there are any questions and if anyone would like their results to be withdrawn.
9) Finally, thank everyone for their time in helping with the research.
Results
Table to show Averaged Raw Data
Line Graph to show Female Ratings against Male Ratings
The above graph shows how the averaged ratings for each respective partner in each couple correspond with eachother. It shows that, overall, females were rated more highly than males, yet the two jointed lines share a similar fluctuation pattern, suggesting that there may exist a relationship between the two sets of data.
Scattergraph to Illustrate the Correlation between Averaged Ratings of Male and Female Photographs
Whilst the trendline on the scattergraph above indicates a slight negative correlation due to the slight slope up from right to left, the data points of this graph are not uniformly scattered about the trendline. This indicates that the very small negative correlation given by the line (negative gradient of equation of trendline) may not be significant. To establish significance a statistical test will have to be carried out.
Inferential Statistics- Spearman’s Rank Correlation [Spearman’s rho]
This method of inferential statistics was used as it is most appropriate for the data type. i.e. correlatory data, related pair of scores which are scored on a scale with equal intervals, i.e. ordinal data. The value for rho was calculated using the graphing package, Autograph. This gave the value for rho =-0.078 (to 2 significant figures.)
The statistical level of significance for a two-tailed test at p<0.05 and N=20 is 0.447. As the observed value of rho is less than the critical value, the experimental hypothesis must be rejected and the null hypothesis accepted; i.e. the apparent correlation shown is not significant, equivalent to stating that ‘there is no correlation between the variables’.
Discussion
Explanation of Findings
My findings do not support the Matching Hypothesis, as the scattergraph indicated a very weak negative correlation between the two sets of ratings, which following statistical analysis was found to be insignificant, leading me to accept the null hypothesis: i.e. that ‘There is no correlation between attractiveness ratings of separated photographs of married or engaged couples.’ From the table of averaged results it can also be seen that the male of each pair was consistently rated as being of a lower level of attractiveness than his partner. A reason for this could be due to the sample: which was composed of a female majority. (there are differences between male and female perceptions of attractiveness)
Relationship to Background Research
My findings are contradictory to much of the research described in the introduction, suggesting that the Matching Hypothesis may not be true in certain situations, and there are therefore other factors which determine whether two people will be attracted to eachother or not. Reasons why Murstein’s and Silverman’s findings are refuted by my results could be due to social norms shifts that have occurred within the last 30 + years, e.g. differing emphasis placed on attractiveness to other qualities when choosing a partner-likely catalysed by having greater social freedoms. As it was found that males were universally rated as less attractive than their partners, (this could either be significant or due to methodological issues) this may reflect a sociobiological difference in mate selection, perhaps a greater emphasis on physical attractiveness in females and more stress on a man’s ability to provide resources. There is also evidence for alternate variants of the Matching Hypothesis which may operate alongside the attractiveness variant, i.e. matching high attractiveness for low intelligence. For example, Alvarez et al. (2004) propose that humans ‘mate assortatively’ ; that facial resemblance is often an important factor in mate selection, however there always remains a trade-off between other factors; as in those ‘reward factors’ as described in the introduction. “Evidence for assortative mating among humans seems well established. Human's mate assortatively regarding age, IQ, height, weight, nationality, educational and occupational level, physical and personality characters” (Alvarez et al. ,2004) It therefore seems inappropriate to reduce attributions of attractiveness to just one reward factor when many others are also important in mate selection in such a complex organism.
Limitations and Modifications
The limitations of this investigation may be a crucial factor in explaining why, despite a replication of a study which gained positive results, failed to lend support to the Matching Hypothesis.
Firstly, the sample size was small (N=20) and thus impossible to generalise findings to the population. Also, the sample consisted solely of 16-17 year old students which is significant in that such a limited sample is extremely likely to skew the results to fit the opinions of this segment of the population rather than being representative of the whole population, which was the aim. The sample was also skewed in favour of females, the number of females participating being almost twice as many as the number of males. This obviously is a confounding factor because the two genders will have different ideas of what constitutes attractiveness. Relating to the subject of this study, younger people may find older people (the people in the photos being 5-30 years the participants’ senior) less attractive generally. To rectify these problems, a larger study would have to be carried out, using a larger sample, within which, if possible, participants would be selected randomly. Further faults of the investigation lie in the methodology. The platform by which the people were rated (2D-photographs) is unreliable, firstly because the sizes, contrast and facial expressions of each photograph was different, but also because, in reality a dynamic 3D picture is available when we are internally determining someone else’s attractiveness. This could only be completely resolved by conducting a field experiment, as in Silverman’s field study, so that participants could gain a more externally valid picture of that person’s attractiveness.
Implications and Suggestions for Future Research
My findings indicate that many factors are in operation as regards to the perceived levels of attractiveness in potential partners. This could have real-life implications re: social networks e.g. speed dating, online dating sites, whereby agencies profit by matching people together. Such agencies could review their methods to see whether displaying specific personality details as well as pictures leads to better matching success.
Regarding further research into this topic, I would investigate how knowing more about people in stimulus photographs affects participants’ perceptions of their attractiveness. If I were to carry out another fait-accompli matching study, I would aim to do so in a naturalistic setting, as a field experiment.
Resources and References Utilised
Resources (not quoted in the text)
Gross, R., Rolls, G. Essential AS & A2 Psychology for AQA A (2006) Hodder Arnold
References
(relating to specific rewards factors quoted in Introduction)
Alvarez, L. & Jaffe, K. Narcissism guides mate selection: Humans mate assortatively, as revealed by facial resemblance, following an algorithm of “self seeking like” human-nature.com/ep – 2004. 2: 177-194
Clore,G.L & Byrne, D.S (1974) A reinforcement-affect model of attraction. Foundations of Interpersonal attraction. New York: Academic Press.
Dion, K.K., Berscheid, E. & Walster, E. (1972) What is beautiful is good. Journal of Personality and Social Psychology, 24, 285-290.
Thibaut, J.W. & Kelley, H.H. (1959) The Social Psychology of Groups New York: Wiley.
Appendices
Appendix 1: Materials and Information Sheets
The following information sheets were physically handed to all participants, and the information contained was read aloud to assure comprehension.
Standardised Instructions:
Request to Participate in Study into Levels of Attractiveness
Hello, my name is Lucy and I am a psychology student conducting an investigation into levels of attractiveness. In my experiment, if you decide to participate, you will be shown 20 photographs of human faces; 10 female and 10 male. You will be given a response sheet on which to record, using a 10 point scale how attractive you think the people shown in the photographs are. (1= highly unattractive, 10= highly attractive) On the response sheet you must indicate whether you are male or female, but your name or any other information is not required. Please be assured that all responses will be kept strictly confidential. Do you wish to participate?
~
Do you have any questions regarding the procedure?
~
Debrief Sheet
At the start of this investigation I told you that the purpose of the experiment was to ‘investigate levels of attractiveness.’ This statement was deliberately formulated to sound vague so that you would hopefully not guess the true aim of the experiment. In fact, the 10 male and 10 female photographs each correspond with each other in that they are 10 married or engaged couples. The experiment was a replication of an earlier study by Murstein into the ‘Matching Hypothesis’ which states that people are attracted to people of the same or similar level of attractiveness. The true aim of my experiment was therefore to see whether the couples in the photographs would be rated by you as having similar levels of attractiveness. Please understand that the somewhat deceiving instructions given to you at the start were necessary in order for the results of this investigation to be valid and it is hoped that you take no offence from this. If at all you do not approve of this experiment given the true nature of the study you have a right to have your responses withdrawn from the investigation. Firstly, do you have any further questions about the experimental procedure?
~
Do you still wish your results to be included in my research?
~
Finally, thank you for your time in partaking in this research.
Stimulus Material: on green cards.
Participant Response Sheet (side A)
Participant Response Sheet
Gender: .
Participant Response Sheet: (side B)
Appendix 2: Raw Data
Ratings of Female Photographs
Participant Number
Ratings of Male Photographs