Abstract:

Profits on repeat-purchase consumer goods can be increased by replacing raw materials with cheaper substitutes while maintaining sales. In these cases, it is obviously essential that the performance of the finished product does not suffer. But it is also desirable to pre-test that consumers cannot tell that the product has been changed at all. In the course of several testing programmes of this kind on medicinal products, we have been faced with some interesting problems of research design, some practical difficulties, some nice questions of interpretation and some anomalous results. In describing and discussing some of these tests, more-or-less as they happened and in their essential detail, it is hoped to show how these interesting aspects of discrimination testing grew naturally out of the questions being asked. Firm solutions are not offered in all cases and, indeed, several question marks are left for debate or for guidance.