Skip to content
Health, Sports & Psychology

Scratch and smell experiment: The results

Updated Tuesday, 8th March 2011

During the last series of Bang Goes the Theory we invited viewers to take part in our giant, online experiment on the sense of smell. What were we up to? What were we looking for and, more importantly, what did we find?

This page was published over five years ago. Please be aware that due to the passage of time, the information provided on this page may be out of date or otherwise inaccurate, and any views or opinions expressed may no longer be relevant. Some technical elements such as audio-visual and interactive media may no longer work. For more detail, see our Archive and Deletion Policy

We decided to test four different smells; more would have been nice, but it wasn’t possible to squeeze too many onto one scratch and sniff card. The smells we chose were selected on the basis of distinctiveness and also on claims that have been made for some of them, by people such as aromatherapists.

It has to be said right away that by using cards in this way we were setting ourselves a hard task. Therapists, and also scientific researchers in laboratories, use high concentrations of essential oils; we had fairly weak synthetic odours! Nevertheless, the chance of testing people on a grand scale, and getting some idea of how we respond to smells, was too good to miss.

Which smell was which?

If you took part in our experiment, you'll be keen to know if you correctly identified the smells. We used: Peppermint (smell A), Coconut (B), Cedar Wood (C) and Rosemary (D)::

Detail from the scratch and smell card Copyrighted  image Icon Copyright: The Open University

We wanted to know how good people were at identifying smells and also how good they thought they were. Also, we looked to see whether the identification was influenced by not-too-obvious cues. We used pictures that ‘went with’ a smell, such as a picture of a tropical beach, which might be associated with coconut.

People were asked to identify a smell, by scratching the patch on the card while they looked at the picture; some people got the matching picture, but others had the picture from a different smell.

Some odours are said to make people feel relaxed, while others are supposed to keep them feeling sharp and alert. We tested this by presenting a series of problems (mental arithmetic, logic and word monitoring) under time pressure. These were to be done while sniffing one of the scents, then at the end, as well as seeing how many answers people got right, we asked how stressed and how alert they felt.

To look for any mate-finding effect we showed people a series of pictures of men and women, and for each they had to rate their attractiveness, on a scale of 1 to 10. This was done, of course, while sniffing a scent.

Lastly, we tried to find a memory effect. In a test near the start of our sequence people had been listening to words. At the end we checked to see how many they could remember. Half of you tested were scratching the same smell at the end as you were sniffing when you heard the words; the other half were using a different smell. Would repeating the smell serve as a memory jog and make people do better?

What we found

Identifying a smell

Well ladies, you rated yourselves as being 7.5% better than the men with your sense of smell. Conceited? No, actually rather modest; overall you were 20% better at identifying the smells we used. Nevertheless, in general people were not very good at putting names to smells. I think the average was pulled down by the Rosemary and Peppermint, both of which people seemed to find very difficult – fewer than 10% of you identified them. The Coconut and Cedar were doing much better at around 30%.

These differences were probably attributable to the chemicals used to make the odours, rather than people being very bad at identifying smells. Rosemary may be relatively unfamiliar, but Peppermint should have been easy enough.

Why the sex difference? My guess is that it is experience; women probably spend more time considering the qualities of perfumes, and in many families it is probably still the women who are usually engaged in food preparation. If we had to find an evolutionary explanation, perhaps we could argue that the female was more likely to be overseeing the feeding of the young, so was attentive to what was being consumed – we men just eat anything!

Adding a picture helped a lot. Cedar and forests, Peppermint and sweet-shops, Coconut and sunny beaches, Rosemary and herb gardens – they all improved the hit-rate. Just guessing? Well, some of you might have been, but imagine the smell wasn’t there at all and you just had to think of something that went with the picture; it wouldn’t be easy to guess right.

Also, I could look at the number of times people gave the ‘picture’ smell when the picture didn’t match. So, did people often say “Coconut” to the beach scene, when the smell was actually something else? No, they didn’t, so the advantage (for Coconut up from a 36% hit-rate with the wrong picture, to 63% with the right one) must have a different explanation.

Coconut beach Creative commons image Icon freshelectrons under CC-BY-NC-SA under Creative-Commons license

When we recognise things our brain makes use of all the contextual information. This is very noticeable when you see a person out of context. You may be familiar with the man who runs the corner shop, but when you see him in the street you think, “I know that face,” but you can’t for the life of you remember from where! The reason of course is that the context is all wrong. Similarly, if you sniffed Coconut while looking at a herb garden, the context was wrong and you were probably saying to yourself, “I know that smell, but just can’t place it.”

The picture I’ve painted so far makes it sound as if our sense of smell is pretty poor. Actually it’s not; we might not be good at putting names to smells, but we can assess their characteristics. We asked you to describe each smell: for example, how ‘woody’ was it or ‘nutty’, how ‘sweet’ or ‘herbal’?

To name a smell presumably means that we can make these kinds of distinctions, suggesting perhaps that people who get the names wrong get the characteristics wrong too – for example saying that Coconut smells grassy. That was not what we found. Those of you who couldn’t put a name to a smell were selecting much the same range of characteristics as those who could.

Stress and alertness

So, what did all these smells do to you? If you took part you will remember those fiendish little tests that took so much concentration, were so easy to get wrong, and must have left you feeling ‘zonked’ at the end! Were any of the smells helpful? The answer, remarkably, is “Yes!” Consistently, people who had been sniffing Coconut felt less stressed at the end of the test; it was only 3% to 5% less, but it seems to be a real effect. Sadly, enjoying the benefit of Coconut did not lead to getting more answers right!

The best scores were achieved by those sniffing Rosemary and Peppermint and, although these didn’t do anything to reduce stress, they did lead to higher alertness scores. Once again, the effect was just a few percent, but is was consistent. Where do these effects come from? It is conceivable that the chemicals have some impact in the primitive parts of the brain. Perhaps there is an effect of association, Coconut reminding us of sunscreen lotion and relaxing on a beach, or Peppermint smelling like a sweet; sucking a sweet often perks us up for a while.

Smelling attractive?

Now the bit you’ve been waiting for - finding out which scent is going to make you attractive! We didn’t have an especially attractive set of pictures for you, so you gave the photos rather middle-of-the-range scores. Overall, the female pictures were rated as more attractive than the male pictures – even by the women! Perhaps not surprisingly, women rated the male pictures higher than men rated them, and men rated pictures of women higher than women did. The question is: did the odours have any influence on this?

Well, the effect was a shift of only about 1%, but there did seem to be something going on. Women liked the men best if they viewed them while sniffing Cedar. Conversely, the men rated women as more attractive when they were sniffing Peppermint. Both sexes found Rosemary a complete turn-off!

Plausible reasons for these effects are difficult to find. Cedar might perhaps be described as a rather ‘masculine’ scent, whereas Peppermint was the closest thing we had to something floral and ‘feminine’. Also, as we have seen, Peppermint has an alerting effect, so possibly men are more likely to judge females as attractive if they are feeling more alert. That invites the question as to why alerting did not have an equivalent effect upon women judging men. One of life’s little mysteries!


Lastly, that memory test; do smells help? We know they do. A very recent study found that people who had learned material in the presence of a smell would remember it better if the same smell was there when they were tested. However, of the odours evaluated, this worked only for Rosemary. This was explained, not on the basis of some special effect of molecules on the brain, but by the distinctiveness of Rosemary and its unpleasantness (high concentrations were being used). So, it seems to require something impactful to have any noticeable effect on memory.

Rosemary Creative commons image Icon lilahpops under CC-BY-NC-ND under Creative-Commons license

What about our rather feeble little scratch and sniff scents? Well, I fear their impact was just that – feeble! There was no special effect for Rosemary, and across the whole experiment, people sniffing the same smell when trying to remember managed to recall just 0.2% more words than the people sniffing a different smell! This was almost certainly a chance difference, and doesn’t reflect any true impact on memory.

Summing up

I’m really grateful to all of you who took part. We have shown that odours really do have an impact on people. Testing that by means of scratch and sniff was never going to be easy, but on our side we had all the lovely audience of Bang Goes the Theory. Because so many of you were prepared to give up some time and take part it meant that, even if scratching scent patches wouldn’t have a great effect, we were still able to detect when something was going on.

One of the most interesting findings is that smells, like other items in our environment, are best identified when found in context. The other way round, where the smell is supposed to form the context didn’t work. In other words, trying to remember wasn’t helped by having the learning context smell back again. As explained, this is probably because the smells were just too weak, not to mention mixed in with all the other smells you had been scratching and sniffing!

If you wanted to take something practical away from this it perhaps could be as follows. If you are revising for an exam, sucking a peppermint may keep you a little more alert and better able to concentrate. You will also need to concentrate in the exam itself, so suck them there too. By doing that you will be putting something of the learning environment back again (same smell both times) so it’s just possible that will help you remember a little more. When it’s all over, you can go home and relax with some nice coconut – dare I suggest a Piña Colada?

Find out more





Related content (tags)

Copyright information

For further information, take a look at our frequently asked questions which may give you the support you need.

Have a question?