A Controlled Recognition Technique for Measuring Magazine Advertising Audiences
Darrell Blaine Lucas · Journal of Marketing · 1942
THE OBJECTIVE of this different recognition technique of copy testing is to determine the percentage of readers of a particular magazine issue who see and get some measurable impression of a specific advertisement. The feature of the method, known as the controlled recognition test, is the elimination-complete eliminationwe believe-of all sources of inflation of scores. An incidental feature is elimination of those inequalities of interview which give unfair advantage to one advertisement or one magazine over another. Beginning in 1937, the controlled method has been developed, tested and compared with other copy measures. From the start the test has followed accepted routines of scientific procedure. One ideal scientific method of measurement is to set up two test situations in which only one element is permitted to vary. The variable we want to measure is the effect of circulating a specific advertisement in a particular magazine issue. The first of the two tested situations is the period just before the advertisement is put into circulation-at which time the effect of circulating it is obviously zero. Current readers of a magazine are called upon and shown an advertising scrapbook containing current advertisements together with advance copy from the forthcoming issue. The readers are asked to identify the advertisements which they remember having seen. This provides advance scores on some of the test copy. Next week these advance advertisements become current and are inserted in another scrapbook. This is the second test situation for the pre-tested copy. The only significant variable is the fact that the advertisements have come into circulation in between the interview periods. Interviews are made with a new sample similar to the first. In that way current scores are obtained on the same advertisements on which advance ratings were obtained earlier. Psychological experience has consistently shown the inaccuracy of responses in recognition tests. No matter how sure a respondent may say he is of his identifications, there are likely to be some spurious responses resulting from confusion of similar things, loose guessing, careless exaggeration and outright lying. False identification is false only when a person has not seen the advertisement. It is apparent that this source of inflation operates equally in both the pre-test and the current test of the same advertisement. For that reason the