Once an advertisement or campaign has ‘run’ it is impossible to measure the effects of the message separately, since the results are confounded by the frequency of the media schedule the impact of the medium used, and other market factors. To a considerable extent an after test is one concerned with total advertising effectiveness.
“After tests” may be designed in a number of ways. All, with the exception of inquiry and sales tests, are based on the respondent’s memory. This raises the question of how soon after the advertisement has run the measurement should be made. Since learning should increase with more exposure, a copy test based on a single viewing of a TV commercial will not provide an accurate measure of the ad’s performance.
Recall versus Recognition Tests: For many years, there has been considerable controversy concerning the relative merits of these two measurement methods of collecting data pertaining to the readership of print ads and, in more recent years, the viewing of television commercials.
By recall is meant the mental reproduction of some target item experienced or learned earlier. Operationally some contextual cue is provided, and the respondent must retrieve the target item from memory. Recall tests of advertisements typically cue the respondents as to the specific media vehicle involved (e.g. the June issue of Fortune) as well as the product class (e.g. mini-computers) and perhaps even the sponsor (e.g. Commodore). Given such cues, the respondent is asked to play back as much of the ad as possible. The same approach, but with modification can be used with TV and radio commercials. This type of testing yields low response rates and only a very few respondents can recall having seen an ad and especially if the lapse of time between seeing the ad and being queried is relatively long.
Recognition tests are used primarily with print ads. The critical thing that defies a recognition test is that the person is given a copy of the information he or she needs to find in memory. In the typical advertising recognition test, a respondent is shown a series of ads contained in a specific publication issue to which the respondent has been exposed. The subject is then queried as to whether or not she or he has even seen the particular ad in that issue. Thus, Starch – a widely recognized advertising rating service uses a procedure that consists of asking qualified readers of a given magazine to point out which advertisements they noted and read. If an advertisement is reported as noted the respondent is asked to indicate what parts were read.
From the very beginning recognition tests have been plagued by response bias. Many respondents falsely claim to have seen an ad that they could not have been since the ad had not run. An early study found that almost as many respondents (32.4 percent) will claim to have seen an ad which did not appear in a two to six week old issue of magazine that they read as will claim to have seen the ad that actually did appear in the same issue (33.4 percent). There are other response biases present such as tendency to guess when not certain about the response, eagerness to please the interviewer, the freedom of the respondent to claim anything (s)he wishes and the tendency of people to deny socially undesirable traits and to admit to socially desirable ones.
As a result of studying recall versus recognition measures over the years, it has been generally concluded that recognition tests generate poorer measures of memory than recall tests. Because Krugman believes that involvement affects the learning process, he argues that recall tests are appropriate for print advertising but not television, while just the opposite applies to recognition tests. His reasoning is that television advertising generally produces low involvement learning resulting from print advertising is based on words. In discussing his theory of how advertising works, he notes that:
The myth about the forgetting of advertising is based primarily on the erosion of recall scores. Yet, the inability to recall something does not mean it is forgotten or that it has been erased from memory. The acid test of complete forgetting is that you can no longer recognize the object.
Recent research indicates growing acceptance of Krugman’s theory. Zielske, for example, contends that, while recall scores may be adequate for thinking television commercials, they understate the effect of feeling commercials. He therefore recommends the use of recognition methods when measuring such commercials.