Comment on Silk and Geiger

Abstract

No one can fail to agree with Silk and Geiger [3] when they conclude that further investigation of the selectivity-ad size relationship with an improved and larger data base is warranted. The evidence they present appears to be at variance with the hypothesis (based on the principle of selective perception) that people who pay attention to small newspaper ads include a disproportionately high number of prospects for the advertised product [1]. Silk and Geiger cite findings that appear to contradict this hypothesis drawn from readership studies made by Behavior Systems, Inc. for the Philadelphia Inquirer in the mid-1960's. They also reanalyze a recent study by Copland [2] of 100 female readers of 99 ads in the British weekly magazine, Woman, and conclude that the smaller ads had fewer brand users among those who noted and read them than bigger ads did. Both the smallest ads and the full page ads got less noting and reading among users of the product than ads between / and 1/2 page in size. The Copland findings are puzzling, but they must be looked at with the following observations in mind: (1) The ads measured include products with a broad market (food and soap) and others with a thin market (proprietary medicines). There is no way of determining from Silk and Geiger's published data what the distribution of these various types of products was for ads of different sizes. In general, widely used products (like groceries) characteristically use larger space units than infrequently used products (like analgesics). The small ads may have represented products with different patterns of usage, different levels of interest or involvement, and different degrees of purchase frequency than the bigger ads. (For these reasons, ad recognition norms differ markedly by product category.) Unless these factors were controlled, the data would not be interpretable in relation to the hypothesis. In psychological terms, status surely involves a different motivational set and hence a different kind of response in the case of an ad for a widely advertised and frequently used consumer product than in the case of a specialty product that is bought on rare occasions. (2) Fourteen of the ads studied were in color, and these presumably included the larger ads, thereby introducing another uncontrolled element. (Splitrun studies have shown that one color adds about 50% to ad readership in newspapers, and three colors about 75%.) (3) The complex and heterogeneous assortment of advertising and editorial matter on a newspaper spread is quite different from the comparatively simple array of elements in a typical magazine spread. It is precisely the newspaper's visual complexity that prompts the need for perceptual selectivity on the reader's part. The same response pattern would not necessarily apply in an environment where fewer messages compete for attention and where exposure opportunity (as measured by page opening) is not as predictably high. (4) Past usage of a product or brand is not a uniform criterion of prospect status, since for different products it may represent a different degree of propensity for future purchase. The attention paid to advertisements reflects, among other things, the salience of what they promote. One would expect a young mother to respond selectively at any time to an advertisement dealing with a baby product, but to respond selectively to an ad for a staple product (like sugar or flour) only when her stock has run low, when she is worrying about her diet, or before a major shop-