Recall and Recognition: A Very Close Relationship
Recall VERSUS Recognition? No, all recall comes from the much larger groups of people who recognise the advertising efforts. Average proved recall of print ads occurs among one out of every three recognizers. Recognition data indicate how many saw the ad. Recall figures tell us how many of them did become sufficiently interested to actually look and often read what the advertiser tries to tell them. They prove it by playing back main content later on.
IN HIS JAR ARTICLE titled 'Recognition and Recall,' du Plessis (1994) notes 'that there exists virtually no empirical evidence ...' on the title topic (or on several derivative topics). In that article, du Plessis then proceeds to present empirical data he has gathered in South Africa; he also includes data on likability. We will join him here in that endeavor.
We at NIPO have only recently begun to gather similar data for television advertising, so a replication of du Plessis' work must wait. However, we have long been conducting similar research for magazine and newspaper advertising and can add data for the print media to the empirical data that du Plessis has offered for television.
These post tests of print ads started in 1949 and right from the beginning combined the procedures of both Gallup (proven recall) and Starch (recognition when going through the magazine or newspaper later on).
Both recall (at the start of the interview) and recognition (at the end), therefore, are available from the same people at the same time. Both effects consequently derive from the same source. That eliminates the problems of comparing recall data from one source (like G&R) with those from other surveys (like Starch).
The average Recognition and Recall figures of both magazine and newspaper advertisements prove to be highly correlated. (See Figure 1.) In magazines, average recall equals 0.33 times average recognition (minus 6 percent) and r = 0.96. In newspapers, the equation reads average recall equals 0.32 times average recognition (minus 5 percent) and r = 0.95.
Why should we compare averages rather than the effects of individual ads? Because all post tests prove these measurements of the perception of advertising stimuli vary tremendously as a result of two groups of causes: One of them is what money can buy - the size of the ad and black-and-white or color. The other is of a purely qualitative nature: the attractiveness of the...