I am currently preparing a study that uses two odors, a pleasant one and an unpleasant one, as context cues, i.e., the odors will be dispensed in the experimental chamber and ss will perform a number of tasks. We have tried to match odor intensity by means of subjective intensity ratings on visual analog scales but the results were only partly satisfactory. In particular, different samples of ss (N=10 each) rated the intensity of the same concentration of the unpleasant odor significantly different while two different concentrations of the pleasant odor were rated different. Eventually the two odors were rated as iso-intense but I'm unhappy about how this aim was achieved. Any suggestions welcome.
It sounds to me that the issue you have is not the wrong method, but a problem with the protocol you used. The method should work fine if you have appropriately trained panelists, have adequate time between samples, etc. it sounds like you either have adaptation issues, are seeing context effects, or your panel was not appropriately trained (if you used random samples of 10 untrained assesors that is far too small a sample to get consistent mean scores).
Agree, Prof. Edgar Chambers. 10 untrained ss is too small.If you used trained ss, training is crucial or how to rated VAS scale sometimes can give effect to your outcome.Make sure they really know how to handle/use the scale.
Thanks for your comments, Edgar and Ammat! I am unfamiliar with sensory evaluation training so I guess at this point we'll just have more naive participants do the rating. Apart from that, can anyone give suggestions on other procedures that work with (rather small numbers of) untrained panelists?
If you want to use trained panel, you need to really understand about the procedures and what you want as outcome. It's time consuming and need high cooperation from your panelists.However, if you managed to do it...all the procedures would be easier cause you work with small group of people (at least 10 panels). In this case, you can use untrained panel but increase the no. of panel (at least 40 persons) to get a good data distribution. About the scale, you should explain carefully to your panelist-how to use it or even having training/exercise before the real test .Just to get them familiar and use the scale correctly which can give a significant impact to your output.Good luck!!!
Peter, thanks for your comprehensive explanation. It's not what I was hoping to hear, but it has already begun to dawn on me that there is no simple solution to this problem.
Thank you all again for sharing your thoughts!
It is not trivial to match odors in intensity that vary in pleasantness. Make sure you instruct your subjects very well so that they understand the differences between odor strength and odor valence. If you have no chance to use trained panelists try to recruit colleagues who have some experience. We usually start out with a couple of intensities of each odorant and let colleagues sample without constraints; they then have to sort the samples according to their intensity. After a few iterations (and breaks) we usually get a pretty good first impression of the concentrations that match best.
Colleagues may not have enough time to devote to the experiment. You might find stereoisomers that vary in odor pleasantness but would presumably have the same chemical characteristics that would actually stimulate the receptors equally. As above the unpleasantness may translate into intensity. Some chemicals eg skatole vary in pleasantness with concentration so lowering the concentration will alter its perceived pleasantness. As above 10 ss is not enough and anosmia can be a factor.
The main issue, as I see it, is conflating concentration of the odorant in the air with perceived intensity. One could easily have a chemical with a low concentration such as a mercaptan be rated as high intensity while vanillin at the same or at a higher concentration would be rated as very pleasant. The focus on matching concentration is misplaced. What you could do is have the subjects individually match the intensity of the pleasant and unpleasant odorant. I don't know if that would work in the context of your study. One should also consider the rate of adaptation as the study progresses. The unpleasant odor may not adapt at the same rate as the pleasant. It may not even adapt at all. More details of the study would be helpful.
There are so many variables to consider when matching intensities that it is best to carry out comparisons in the same environment as that in which you will carry out the sensory test. For example you could use a series of similar rooms in which you evaporate different concentrations of odorant, each room being used for a single concentration. Use scaling or ranking to rate intensity using a panel assessing rooms in random order. Perfume houses will have dedicated booths/small rooms for this type of operation....at a pinch they'll use a large container like a barrel.
Comparisons in bottles are only of use if your subjects will assess in the same way. Remember that each material will have a different psychophysical function so the intensity of each material will change to different extents on dilution. It's even more complex for mixtures. Try to at least have equilibrium conditions so that the concentration in the room is constant (ie balance losses through leakage and adsorbtion against emanation of new material). With dynamic olfactometers remember that odour is a 'mass' sense. Intensity therefore varies with (log) flowrate as well as (log) concentration. Mixing a moving air stream from a delivery system with still ambient conditions for assessment may be very complex to characterise.
The best way to estimate the intensity of odorant is to measure the amplitude of electroolfactogram (EOG). EOGs are electrical potentials that occur in response to olfactory stimulation and are recorded at the surface of the olfactory epithelium of vertebrates. The EOG represents the sum of generator potentials of ORNs. Perceived odor intensities by humans are found to be strongly correlated with the EOG amplitude.
Alas, it is unlikely that he will be able to do the EOG study in animals to validate intensity in humans unless he uses the same odor as done in the animal studies. These odors may not be the ones under consideration. In addition, one should not neglect the considerable influence of the trigeminal nerve on the sense of smell. If the odorants under consideration have a strong trigeminal component it will influence the perceived intensity.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication