92 JOURNAL OF COSMETIC SCIENCE WHO SAID SENSORY MEASUREMENT WAS SIMPLE. 9 Michael O'Mahony, Ph.D Department of Food Science and Technology, University of California, Davis, CA 95616 maomaho n y @ uc da v is. e du This talk will demonstrate some of the tricks that the senses might play on unsuspecting persons required to make judgements about the sensations they are experiencing. It will also examine the ways in which people can make mistakes in reporting what they think they are perceiving. Marketing, advertising and packaging are vital for the sale of cosmetics, personal and household products. Yet, there is a further important aspect: the sensory properties. These are the responsibility of the manufacturer and this means that the manufacturer must be able to measure them. Such measurement is necessary for the development of new products, for the measurement of consumer concepts, for quality control, ingredient specification, storage studies, product optimization etc. Although a company will have laboratories full of instrumentation for making such measurements, it tums out that the human senses are often more sensitive to variations in sensory characteristics than many laboratory instruments. So it is necessary to use the human senses as measuring instruments. The problem with humans is that the information processor for their senses is the brain and this was not designed or programmed to mn instrumentation. It was programmed to be the information processor for cavemen and cavewomen. It has the unfortunate habit of enhancing the perceptions that would be useful to a caveman or cavewoman but which can seriously interfere with sensory testing. So, to use the human senses effectively, we have to understand the workings of the nervous system and brain. Then we can avoid the biases that it will introduce into our perceptions and so enable us to make accurate assessments of the sensory properties of the products we are testing. The same argument applies to consumer product testing and marketing measures, as well as more medically related research. There are something like a trillion connections in the brain which makes it the most powerful computer on the planet. Yet, despite this, it cannot process into consciousness all the sensory information that reaches it from the body. It is basically overloaded with information. So it has a set of strategies for avoiding 'information overload'. The first strategy is simple. The brain does not arrange that you pay attention to all the input from the senses. Humans tend to focus on only a small part of the information arriving at its 'door': visual and sound input. Animals tend to focus on smell. The second strategy has been given the technical name of 'sensory adaptation'. This refers to the mechanism in the brain that allows it to disregard redundant information. If a stimulus is completely unchanging, the brain does not want to pay attention to the senses as they keep repeating the same message over and over again. It is too busy processing other information for it to be bothered with the same redundant message being repeated again and again. So it simply turns down the volume switch on those repeated messages for the sensation to vanish or at least be greatly reduced. But if there is even a slight change in that message, the sensation reappears loud and strong. This is a common everyday experience. Imagine that you walk into a room and you immediately notice that there is a slight smell there. As you walk around the room your nose will be sniffing the same concentration of volatile chemicals because they have reached some equilibrium concentration throughout the room. So the stimulation at the nose is constant. The input reaching the brain is constant. So the brain, which is too busy processing other information to bother with repeated stimulation, simply tums down the volume switch and the smell slowly vanishes. The smell system has been desensitized or adapted to that particular odor. Yet walk out of the room and come back in again and the smell will reappear. There will be a change in the volatile chemicals surrounding the nose as you leave and re-enter the room, so the brain tums up the volume switch once more and the smell re-appears.. It is said that you can only smell a rose once. This is not completely true but the smell of a rose is reduced as you repeatedly sniff its aroma. Everyone will be desensitized to perfumes or colognes with repeated sniffing. This can be inconvenient for any form of sensory assessment. Thus, the sensory analyst must work out ways of getting round this problem my making sure that the assessement of a scent is completed rapidly before it vanishes. It is also important that the nose is cleared by removing it from the area, before assessing the next scent, so as to defeat this desensitization effect. This desensitization does not only apply to smell it applies to all the senses, even vision.

2002 ANNUAL SCIENTIFIC MEETING 93 The third strategy is for the senses not to give the brain all the information that they have gathered. They only give a fraction of the information: just enough for the brain to be able to piece together and recreate the whole picture. The senses extract the essential features of what it is perceiving and the brain has to examine them and decide what is being perceived. It is a truly remarkable piece of computing. Working in this way shields the brain from 'information overload' because less information arrives at the brain. Yet, the brain can sometimes make the wrong decision then we will experience illusions and misperceptions., which can be easily demonstrated. It is not only the input side that can play tricks. The output side can also be misleading. There are many examples. Firsfly, when a person estimates the strength of a sensation or how much they like a product, their estimations are not very numerical. The spacing between their numbers tums out to be pretty uneven. It is not equal. This means that the basic assumptions of mathematics are broken. After all, we assume that the distance between '1' and '2' is the same as the distance between '101' and '102' or that '10' is twice as big as '5'. When humans make numerical estimates this is just not true. Humans are basically uncalibrated instruments. So we need a few tricks in our statistical analysis to cope with this. Secondly, merely changing a few words in the instructions for a test can switch on a different program in the brain and alter a persons discrimination skill. Furthermore, a consumer can swear that he prefers one product over another yet the products can be identical. This is normal brain function. But understanding these mechanisms allows us to make valid and sensitive measurements. If you know how your instrument works, you can use it well. You can even make solid numerical measurement for such abstract concepts as the sophistication of scents, the appearance of refreshingness of toothpastes or any sort of abstraction that the marketing department comes up with. In summary, if you understand how the human senses work you can upgrade your methods of measurement, making them more efficient and accurate, whether it be with highly trained experts or naYve consumers