CONSUMER PRODUCT TESTING 773 incalculable. It bridges the gap between the technical analysis and the consumer responses, so that the test operators will be able to relate con- sumers' descriptive terms to the use profile and finally to the technical or flavor analysis. To report consumer findings to the product devel- oper, one must be able to speak in his technical terms. In addition to use profiles, the panel will also produce abuse profiles. Such information enables one to characterize the inherent latitude of the product and also to anticipate untoward responses and to guard against them. If, for example, the directions for preparing a soup call for 10 minutes' simmering, the effects of under- and over-simmering should be known. Undercooking could cause the noodles to be tough and the flavor underdeveloped. Overcooking could concentrate the soup, making it strong and salty. The Testing Situation During this study period, the product elements that consumers are likely to observe are sorted out, and decisions are then made regarding the test method. As previously mentioned, if the product is not a finished one but is to be tested in order to obtain guidance for its future direction, one might decide to test it in comparison to another unfinished product. If it has a counterpart on the market, then its performance could be tested in comparison to the "blind" marketed product. And similarly, if it has been designed as an improved product, the assessment of its improvement features could be made through a comparison with the unchanged product. A new-concept product would of course be tested by itself in fact, single-product testing may also be applied to any product if the test can be designed to meet its objectives. The product's properties and its intended use will determine some of the instructions for the consumer panel as well as duration of the use period and, therefore, the supplies needed. A side-by-side comparison would be requested in a statistically designed difference test aimed at defining flavor attributes of puddings differing only in sweetener content. This technique provides for a minimum time lapse and thus a direct comparison, which is a stringent test. Stronger tasting products, such as mouthwashes, would not be amenable to such immediate comparisons. Instructions would therefore request alternate use of one mouthwash on one day and the other on the next day, both to be used according to the panelists' normal patterns. Alternating-day use gives a closer compari- son than alternating-week use, for example.

774 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Obtaining Responses The test objective, the product's use properties, and its intended use will determine how to obtain information from the consumer panel. If, as in foods, eating quality is the primary product property to be tested, responses can be obtained by questionnaire. If, as in hard liquors, other unknown attributes may supersede flavor, face-to-face interviews are in order. Interviews are worth their higher costs for the advantages al- ready cited, i.e., they are time saving and provide more definitive infor- matioI1. The interviewers are personable technical people, who are made to be fully cognizant of the purpose of the test, the client, the product's Flavor Profile, and its use and abuse profiles. These technical people are experienced flavorists, members of the A.D.L. Food & Flavor stall who train in on the particular test. During their training period, they use the product as the consumers will and, under supervision by their peers, practice first among themselves and then with consumers. In their practice sessions they learn to establish rapport, how to conduct an un- structured interview on this product, and to train their memories so as to be able to write up the interview or fill in their data sheets away from the scene of the interview. In other words, they learn to have a conversa- tion about the product, allowing the consumer to describe her own im- pressions without channeling her responses by checklist type questions. They rely mainly on open-end questions and are allowed to pursue what- ever the consumer considers important, probing on the spot for clarity and definition of descriptive and vague terms. Similar precepts guide the development of questionnaires to be filled in by the consumer panelists. The note transmitting the questionnaire tells the consumer in general terms why the test has been conducted. After asking about the use of the product, the questionnaire itself asks mainly open-end questions about the product. If the panelists have been stimulated to be communicative, they will give their appraisal, sort out the favorable features and those that, in their opinion, need to be improved, and indicate the relative importance of the features they have discussed. This result, however, is not left to chance. A pretest of the instructions, use period, and questionnaire is almost invariably con- ducted, using about five families. Other things not left to chance concern the test samples themselves. Before they are placed for home use, they are sampled at random and checked out to make certain that they do represent the product to be