J. Soc. Cosmetic Chemists, 19, 381-393 (May 27, 1968) Comparison of Local Vs. National Consumer Panels in a Paired Preference Test. II. Evaluation of Responses JEAN F. CAUL, Ph.D.* and SHIRLEY A. RAYMOND, B.S. ** Presented on May 17, 1967, at the 27th Annual Meeting of the Institute of Food Technologists, Minneapolis, Minn. $ynopsis--A local panel of 89 fa•nilies known to discriminate, i.e., accurately distinguish product properties, was compared to a national sample of 249 families with respect to their evaluation of two toothpastes. Previously reported results had shown virtually complete parallels between the two panels for over-all preference, hedonic scoring, and correlation of flavor preference with over-all preference. Results reported in this paper showed that the same information about the flavor properties of the test toothpastes was obtained from both panels. Further, it was found that the over-all preference trend for one toothpaste had been based on dislike of the other, whereas the losing product was liked by the minority who pre- ferred it. INTRODUCTION The testing of newly developed products through consumer panels is a tool used both by marketing and research groups. In using con- sumer product testing, both groups have the same ultimate goal: to have a winning product. But their immediate goals are different. Marketing people consider the consumer panelists as representative of a predefined market. Research people consider the consumers as a means * Dept. of Foods and Nutrition, College of Home Economics, Kansas State University, Manhattan, Kan. 66502. ** Food and Flavor Section, Authur D. Little, Inc., Cambridge, Mass. 02140 381

ß 382 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS of checking out the product's properties in terms of its specifications or blueprint. They use consumers as pilot testers to tell them what is right and what is wrong with the product if the product passes the pilot test it can be released to the marketing department if it fails, the research group should have obtained useful information so that the product can be purposefully modified. This paper considers the application of consumer panels in pilot product testing. For a number of years the authors have been pilot- testing products with relatively small consumer panels of so-called articulate discriminators (1). Discriminating consumers are persons who can accurately distinguish a product's properties. To decide if they are flavor discriminators, one evaluates their responses against laboratory flavor analyses of the test product. Understandably, if a consumer panelist is unable to communicate his observations, it is not possible to judge if he is a discriminator. Therefore, he must be articu- late. Our source of articulate discriminators is the employees of our company's local operations. Since size and type of panel are open to contention and scrutiny, research was undertaken to compare the per- formance of a small, local panel of articulate discriminators with that of a more conventional consumer panel: a larger, national quota-sample of consumers. The local panel consisted of 89 ADL families (403 persons) sdected because they were known to discriminate. The national panel consisted of 249 families (832 persons): 16% West, 26% Northeast, 29% Central, and 29% South. They all tested the same two coded toothpastes ("R" and "W") during the same two-week period. Responses were obtained through questionnaires and hedonic scores. Questionnaires for the local panel (except for hedonic score cards) were entirely free re- sponse. Those for the national panel were both structured and open- ended. The local panel was instructed to use the toothpastes alter- nately the national panel was instructed to use each of the pair for one week (order of sample use was balanced). The first step in comparing the performance of these two panels showed virtually complete paralleling in over-all preference favoring Toothpaste R over Toothpaste W, distribution of hedonic scores, and correlation of flavor preference with over-all preference (2). From this horizontal analysis it was concluded that the general reactions to a nationally used product--in this case, toothpaste--were similar from consumer panelists whether they were selected to represent a national quota-sample or to represent known product discriminators.