MARIA DEMATTÈ, ROBERT ÖSTERBAUER AND CHARLES SPENCE’S EXPERIMENT ON OLFACTORY CUES AND FACIAL ATTRACTIVENESS (2007)
Sometimes you meet a person and something about them really attracts you. Maybe it’s their voice, their facial symmetry (cause that’s what everyone’s looking for in a partner), or maybe it’s their laugh. There are a lot of factors that play a role in our perceived attractiveness of other individuals. One such factor is smell. We’ve all been there: you see someone cute and they smell great and they’re automatically ten times cuter. Or you meet an attractive person who smells kinda bad and suddenly their attractiveness drops just a little bit…
Demattè, Österbauer and Spence conducted their study to examine this domain further. They wanted to know if odours really had an effect on someone’s attractiveness.
The sample was made up of 16 females ranging in age from 20 to 36. They were all from the University of Oxford. Before participating, they all completed a confidential questionnaire to make sure that they had a normal sense of smell, no history of olfactory dysfunction, and normal or corrected-to-normal vision. They were also asked about their general health and their ability to perceive odours and colours.
An all-female sample was actually chosen because previous psychological research showed that women may be relatively more sensitive to the effects of olfactory cues, as compared to men.
This experiment was performed in accordance with the ethical standards agreed upon in the 1964 Declaration of Helsinki and the ethical guidelines laid down by the Department of Experimental Psychology, University of Oxford.
40 male faces were used as visual stimuli. They had been rated either as having ‘high attractiveness’ or ‘low attractiveness’. Four odours were used after being chosen from a pilot study: the pleasant ones were geranium and a male fragrance called Gravity, the unpleasant ones were rubber and body odour. Clean air was used as the neutral odour.
The experimental session was to consist of three blocks of 40 randomised trials, so every participant completed 120 trials each. Each male face was randomly presented three times during each experimental session; one time with a pleasant odour, one time with an unpleasant odour, and one time with the neutral odour. To counterbalance the presentation of each face-odour combination, the 40 faces were divides into four subgroups of 10 faces each: five highly attractive faces and five less attractive faces. Each subgroup had a different mean attractiveness and was presented with one different combination of odours:
- One group of faces alongside clean air, the geranium odour and the body odour.
- One group of faces alongside clean air, the male perfume and the rubber odour.
- One group of faces alongside clean air, the geranium odour and the rubber odour.
- One group of faces alongside clean air, the male perfume and the body odour.
The participants were seated on a chair 70cm away from a computer with their chins on a chin rest. They were told to look at a fixation cross positioned at the monitor’s centre. They had to begin exhaling through their nose when they heard a ‘quiet tone’ and breathe in through their nostrils when hearing a ‘louder tone’. One of the four odours or the clean air was delivered via an olfactometer 500 ms after the ‘louder tone’. Participants had to decide whether an odour had been presented or not by pressing a key on the keyboard. 1000 ms after the onset of an odour, the fixation cross disappeared and a face appeared for 500 ms in the centre of the screen. When the face disappeared, the odour was also terminated and clean air would be delivered in its place. The screen turned black for 2000 ms and was then followed by a nine-point rating scale. The women used this scale to rate how attractive they thought the face was.
The rating scale had one for ‘least attractive’, five for ‘neutral’ and nine for ‘most attractive’. When the participant had rated the face, the fixation cross appeared again on-screen for 10,000 ms.
Participants rested for five minutes after completing forty trials; this was to reduce chances of olfactory adaptation and fatigue. At the end of the experiment, each participant smelled the odours individually and rated each odour on different dimensions from number one to 100 on a Labelled Magnitude Scale (LMS). The odour and scale presentation was randomised, and the dimensions were:
- Odour intensity
- Odour pleasantness
- Odour familiarity
The experiment lasted for around 50 minutes altogether.
The results showed the following:
- Participants rated faces as significantly less attractive in the presence of an unpleasant odour, compared to when a face was presented with a pleasant or neutral clean air.
- There was no significant difference in ratings of attractiveness in the pleasant versus neutral condition; basically, there was no difference between facial attractiveness when a face was shown alongside a pleasant odour and neutral odour.
- Pleasant and unpleasant odours were perceived to smell more intense than air air.
- There was no significant difference between the intensity of pleasant and unpleasant odours.
- Unpleasant odours were perceived as being less pleasant than the pleasant odours but were not judged as significantly different to clean air.
- There was no significant difference between clean air and pleasant odours.
- All odours were rated to be equally familiar.
Type of research method
This was a laboratory experiment.
This was a within-participants, repeated measures design.
The IV was the odour being used, specifically: the geranium, body odour, male perfume and rubber.
The DV was the perceived attractiveness of the male faces.
- High level of control: Since most things were randomised (e.g. presentation of faces), the procedure was standardised, the clean air was used as a neutral comparison and there was a reduced possibility of external odours reaching the participant, it is likely that the results were highly accurate.
- Pilot study: A pilot study was conducted before the actual experiment which allowed the experiment to be modified to achieve high validity.
- Replicable: Scientific equipment and a standardised procedure allows us to repeat this experiment and check the reliability of the results.
- Quantitative data: Since the data was in the form of facts and numbers, it will be much easier to organise and convert into statistics. It is also less subjective compared to qualitative data and takes less time to analyse.
- Unrepresentative sample: The sample was very small and made up of people from a specific area and, most probably, a similar ethnic/social background. This means we cannot generalise the results to other countries, ethnicities or people in a wider population. It was also limited to females but this isn’t necessarily a weakness since the study was aiming to examine female reactions. Also, there is the question of sexuality: how do we know if every woman was even heterosexual?
- Lack of ecological validity: The study was conducted in a laboratory setting, an artificial environment. Looking at faces of men in this sort of context isn’t really a true-to-life activity (except maybe if you’re browsing through a magazine!).
- Subjectivity: As the saying goes, beauty is in the eye of the beholder. Rating attractiveness is not based on scientific fact or evidence, it is merely a matter of personal choice and subjective processes. While some may swoon at the sight of a beard or red hair, others may take it as something less attractive and prefer something else. So really, who says what is and isn’t attractive?
- Informed consent: Yes, this was obtained.
- Deception: There was no deception.
- Confidentiality: The personal identities of the women were not made public.
- Emotional or physical harm: There was no harm inflicted on the participants; the researchers even checked for medical issues beforehand.
- The right to withdraw: There is not much about this in the study but it is likely participants could withdraw.
- Debriefing: There was nothing to really debrief participants about and there is no mention of debriefing.
Reference: Demattè, M. L., Österbauer, R. and Spence, C. (2007). Olfactory Cues Modulate Facial Attractiveness. Chemical Senses. 32(6): 603-610.