Send letter to journal:
Re: Recall - recognition

Thank you for your well done study.

I had a question about your measure of recall, which in effect requires the ability to think abstractly and verbalize to in fact 'prove' to the interviewer that the ad and its message were seen, heard, and 'digested'.

Our organization in NY, the Advertising Research Foundation, which may not be familiar to you, has embarked on a series of studies about the role that emotion plays in advertising effectiveness. As such, we are growing in confidence regarding the use of 'forced recognition' of ads to tap into a non-verbal and emotional layer of impact that is, in essence, stripped away by the more cognitive-based recall methods. In short, we feel these two techniques are measuring two different things - one, a perhaps mostly unconscious recognition that remains in memory, and does affect behavior - and, of course, the more common recall method, which is definately precise if one wants to 'force' the respondent to prove they have seen an ad, and even understood the message.

This forced recognition is fairy easy to do now with the use of online surveys - to show the ad in as close as possible to its natural setting.

Our concern is, particularly at younger ages (12-18)where teens are not necessarily fully cognitively developed, that there may be two things happening to descrease the accuracy of findings, and subsequent linkage to ad costs:

1. respondents, particularly younger ones, who cannot verbalize an ad which may have affected them deeply from an image perspective, will be terminated from the surveys. Or, if forced by an interviewer to elucidate, they may drop out. This results in the stickly problem of completely missing, non-randomly, as you know a horrible state we try and avoid. And nearly impossible to back track and figure out.

2. For all respondents, the fact that the early level of processing - which we call recognition - is in effect over ruled with the recall line of questioning, we may be missing a substantial opportunity to capture the effect of the image, emotional, and unconscious on the effect of the ad on a respondent. If respondents are merely presented the ad in, for example, an MPEG file, and asked if they have seen it, you recieve a purer and broader measure of the potential effect of the ad. In short, with recongition you will get accuracy, with recall, precision.

The ARF has a concern with using media weight (here called GRPs) as a measure for predicition purposes. In our opinion, media weight represents what was bought - it is better, in our opinion, to focuse on what you 'got' in the form of psychological GRPs - frequency and reach. Then the link to and justification for relating the cost of media to the 'results' of the advertising can be calculated with more confidence.

I am very curious to hear your response to this, for my own edification, and to share with my colleagues here. As you could guess, we all operate in 'silos of knowledge' and it would be terrifcally fascinating to see how we can cross -educate each other.