rff HE objectlve of this study can be stated as follows: What is the efI fect of varying frequency and eontinuity on the results of the advertising program? Is it possible to develop a technique which will provide a quantitative answer to two practical questions confronting every advertiser: (i) How often should I advertise? (2) Should I advertise at regular or irregular intervals? These questions cannot be answered until two more fundamental problems have been settled. The first is an appropriate definition of advertising results, a step highly dependent on the relevant point of view. The sociologist, for example, would be interested in the impact of advertising on social behavior; the psychiatrist in its effect on personality; and the billboard reformer in its effect on landscape. These alternatives are suggested by way of emphasizing the necessity for clarifying which aspect of advertising is to be researched. From the point of view of the advert'iser, the objective of advertising is to influence buying behavior, or sales; advertising results are the sales which are caused by the advertising for which the results are to be measured. It is in this context alone that advertising research is considered in this study. Nevertheless, it should be clear that other aspects of advertising, even other economic aspects, are important to the advertiser. For example, there is the question of control or position in the industry, a long-run problem excluded from consideration here. The second fundamental problem is the technique of measuring the sales results of advertising. The methodological difficulties involved in this process are almost overwhelming. The very magnitude of these difficulties explains the diversion of much research energy into activities (such as magazine readership) which have no necessary connection with the measurement of the sales attributable to advertising. Such substitute activities may be completely justifiable, but their vindication can come only from a successful assault on the basic objective. Certainly not until sales results are measured can the extravagant claims both {for and against advertising be subjected to disinterested analysis. While this study represents a partially successful assault, no effort has been made to conceal the difficulties still to be overcome. Of the conclusions which can be drawn, none is more important than the tentative answers to the questions about frequency and continuity posed in the first paragraph of this introduction.