32 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS area of non-verbal responses. If people cannot express themselves in words it may be possible for them to do so in some other way. In addition, techniques are being developed to measure responses which are inde- pendent of people's own volition. A psychogalvanometer is an instrument which involves the fitting of electrodes to the wrists to measure the variation of rate of perspiration which is quite involuntary and responds to outside stimuli. The difficulty is that we cannot always ascertain what is a positive and what is a negative response we only know there is a response. Another technique involves the measurement of pupil dilation. The pupil of the eye dilates when exposed to strong light but until recently it was not fully appreciated that it also responds to stimuli such as advertise- ments, sounds or smells, all of which can cause minute changes in pupil dilation. DISCUSSION MR. S. V. BRASCH: To what extent are psychologists used in the interpretation of consumer research tests? THE LF•CTURER: Psychologists and people of other disciplines have been employed for instance, in the use of the pupil dilation technique which is derived from medicine rather than from psychology. Psychologists are not used a great deal for run-of-the- mill product testing, but they have been used in determining attitudes to perfumes, how these can be categorised and how people express differences between them. In comparative product testing they are rarely used. In concept testing, on the other hand, they are often used. MR. R. CLARK: Mrs. Ludford mentioned the effect of claims on acceptance of products. What are your views? THE LECTURER: We must first establish the objectives of product testing. On the one hand, in blind product testing, we are usually trying to determine whether a product is thought better, and in what respects it is seen to be better, than either a competitor or the standard product. On the other hand, you can give two different samples of people exactly the same product, but say something different about it, either in the form of a concept statement or an advertisement. One may then en- counter a different playback from each sample which can only be due to the secondary influence and not to the product itself. This is not so much product testing as testing the additional stimulus. MR. C. PUGH: What are your views on the use of staff shops or local shops as a way of combining the concept with product development and seeing if you can get some idea of its saleability, without going to the expense of really selling it? TH• LECTURER: I do not favour using staff shops, mini markets, and so on. The difficulty is that staff shops almost certainly have customers who are very well acquainted with your products and are certainly more aware of them than the population at large, and are thus not representative of the population. Local shops are better but one must be very precise in establishing the objectives of a test, as a local shop is unlikely to be representative of all shops it is better to use a number of shops throughout the country rather than use a local one merely because it is handy and because the manager agrees to cooperate. A difficulty arises because you cannot put a product into a shop without giving it proper packaging and a brand name, supporting it with advertising and possibly other forms of promotion. The cost of

PRODUCT TESTING IN THE SPHERE OF PRODUCT DEVELOPMENT 33 producing a few hundred packs of a new product for such a test would normally be prohibitive. Regarding extended usage tests one should distinguish between new products and existing products. If one tests a completely new product, it is advisable to employ some form of extended usage test, ahnost in the same way as a test to destruction with an airframe or a submarine. One wants to see where the breaking point comes, possibly by making people use the product more often than they would ever do in real life. You may then force them to the point where they are antagonised. Such reactions do not necessarily mean that the product would not sell, because in reality they would rarely use it exclusively for months they may only use it once every three months, possibly buying some other brand as well, so that they never get to the point where they dislike it. Chocolate manufacturers, for example, allow employees when they first join, to eat as much chocolate as they like, free of charge. After a time they grow to dislike the chocolate, not because the chocolate is not a good product because they have become sated with it. For this reason one must use extreme care in interpreting extended usage tests. Also, in testing new products a measure of novelty appeal is often found. The degree of this is difficult to measure precisely, but it nearly always exists. MR. F. D. GRAINGER: What do you consider the value of a study of consumer com- plaints in product development? Do you consider this an extension of a testing panel, and do you think that technical evaluation of letters of commendation for a product should also be carried out? In my experience the technical people see all the complaints, and the marketing people see only the letters o[ commendation. THE LECTURER: Strictly speaking, this question is not really within the sphere of product testing. In consumer testing, complaints or commendations as such are not sought, merely reasons for preference. These may in fact danm the product, but I would not regard them as complaints. A complaint normally comes from someone who has bought the product. When the product is on sale great attention should be paid to any complaints. Even if there is some real deficiency in the product, only a small number of complaints may arise, but these may represent only the tip of the iceberg. Complaints should be referred to marketing, production or product research as appropriate for thorough investigation. MR. F. D. GRAINGER: I was really referring to the actual complaint against the pro- duct as such, not the presentation or anything which concerns marketing. I am thinking about complaints on flavour which recur on most flavoured products. A certain set of people dislike that flavour but the technical people whose job it is to develop the product, only hear about the complaints and they do not know, unless they make it their task to try and find out, how many people have commended the taste of the product or whatever facet it is they are considering. Without this co-ordination of knowledge about customer reaction once it is on the market, it is very difficult to assess whether the complaints are outweighed by commendations or how they fit in. THE L•CTURE•: This is a question for management. If a research report is correctly written, and it lists the reasons that people have given in fayour of the product, as well as their dislikes, indicating the proportions of people involved in each, and if that report is seen by everyone concerned, the technical people may well then realise the true situation. Regarding flavours with a minority appeal - if, for instance, you were dealing with passion fruit - a lot of people may say they dislike it, but if your