Jingles in Advertisements: Can They Improve Recall?

Wanda T. Wallace, Duke University
[ to cite ]:
Wanda T. Wallace (1991) ,"Jingles in Advertisements: Can They Improve Recall?", in NA - Advances in Consumer Research Volume 18, eds. Rebecca H. Holman and Michael R. Solomon, Provo, UT : Association for Consumer Research, Pages: 239-242.

Advances in Consumer Research Volume 18, 1991      Pages 239-242


Wanda T. Wallace, Duke University

The traditional approach to studying music in advertising has focused on the effect of music on attitudes towards the advertised product (Gorn 1982; Kellaris and Cox 1989; MacInnis and Park 1990), the consumer's mood (Alpert and Alpert 1990), and the consumer perception of an ad containing music of any form such as the perception of an ad being upbeat, informative, etc. (Stout and Leckenby 1988). Music has also been shown to influence the consumer's pace while shopping and eating (Milliman 1982, 1986). All of these papers address more of the affective side of the consumer response as opposed to the cognitive side. Within a more cognitive perspective, music has been viewed as a distraction (Park and Young 1986).

In contrast to the above approaches, the current paper wakes a strong cognitive approach and considers how and when music might serve as a recall aid. Some experiments supporting this view are presented.

Music in this paper will be primarily lyrical music rather than background or nonvocal music. Although some of the ideas can apply to nonvocal music, that is not the focus of the investigations. Also, the paper considers unique, novel lyrics written for a particular ad rather than old songs imported to or adapted for an ad. Although again, some of the ideas have a direct application to the use of well-known music.

Why might music aid or improve memory for an ad? First, consider how difficult it can be to get a jingle out of your head when you have heard it several times. Whether or not you like the jingle is irrelevant; you just can't seem to stop humming it. One well worn example is the Oscar Mayer song. Just saying the brand name is usually enough to start part of the jingle playing through your mind. Second, jingles seem to be easily recalled even if they have not been heard for years. For example, the Mounds/Almond Joy "sometimes you feel like a nut" jingle was just recently revived for a Mounds ad campaign because consumers consistently reported remembering those lines even though they had not been aired for years (Dagnoli 1989).

There is some experimental evidence to support the notion that music can improve recall (Wallace 1990). This experiment involves ballads rather than advertisements. The advantage of using ballads is that they are novel melodies which are mostly unknown and the texts describe simple events and ideas. Since these stimuli are purely auditory any potential interaction with or interference from a visual display is eliminated.

In this experiment, subjects either heard three verses of a ballad spoken or the same three verses sung and then recalled in writing the text they had heard. Subjects knew in advance that they would be asked to recall the text. In addition, subjects were instructed to recall the text as close as possible to the exact, original wording. Subjects listened to one of the two ballads five times and recalled the words of that ballad in writing after the first, second, and fifth repetitions. After learning an additional song, subjects were asked to recall the song they had learned once more in a delayed recall task. The time between the last recall and the delayed recall is about 15 minutes; however, learning the additional song is somewhat disruptive.

The subjects in this experiment are undergraduates with no particular expertise although most subjects had a few years of choral experience. According to pretest measures, no participants were familiar with the ballads to be learned.

Two different ballads were used in the experiment; however, since there were no differences between the ballads this factor will be ignored here. The set of three verses from each ballad- contained 80 or 85 words. Both the sung and spoken versions of each ballad were equally understandable and were performed by the same person.

The percentage of words recalled verbatim was calculated for each subject. The basic results are presented in Figure 1. A repeated measures analysis of variance (ANOVA) with condition and ballad as between subjects factors and trial as a within subjects factor compared verbatim recall for sung and spoken conditions. As expected, performance improves across trials 4(2,120) = 598.96, ;2 < .0001). More importantly, both overall (E(1,60) = 19.95, g < .0001) and at every trial, verbatim recall is significantly better for the sung condition than for the spoken condition. After five trials, recall is very good, averaging about 86% of the words recalled verbatim in the sung condition and 76% in the spoken condition. Even with such high recalls on the last trial, there is still a significant advantage for the sung condition in the delayed recall task 4(1,63) = 12.49, ;2 < .001). Mean verbatim recall for the sung and spoken conditions at the delayed recall task are 81% and 67% words correctly recalled, respectively.

Music however does not always result in improved recall. One such limitation involves the repetition of the melody and how easy it is to learn the melody. In a follow-up experiment (Wallace 1990) to the one described above, the first verse from each three-verse ballad segment was excerpted from the original tape. Subjects received the same instructions and the same procedure as in the three-verse case. That is, subjects heard either the sung or spoken version for one of the two ballad segments and were asked to recall the verse as close to verbatim as possible after the first, second, and fifth repetitions. After learning additional material, subjects were asked to recall the verse again in a delayed recall task. However, in this experiment the results differ dramatically from those in the prior experiment.

The advantage of hearing the ballad sung that is found in the three-verse case reverses when subjects heard only one verse. The spoken condition results in better verbatim recall than the sung condition 4(1,35) = 4.29, p < .05). At the first and fifth trials, the difference between the sung and spoken conditions is only marginally significant (F(1,35) = 3. '9, E. < .06 and F(1,35) = 3.86,; < .06, respectively). On the second trial, the difference between conditions is significant (E(1,35) = 5.26, p < .03). Finally, in the delayed recall task, no effects or interactions are significant. Thus, once the verse is well learned, there appears to be very little long term difference between verbal recall of material that was sung and material that was spoken. Perhaps over a very long period between learning and recall, such as months or years, a difference could emerge.



Even though the differences between conditions are borderline on the first and last trials and nonexistent in the delayed task, the results clearly indicate that music is not a facilitator as in the prior experiment and that music can serve as a distractor. It is important to keep in mind that the verse heard in this experiment is identical to the first of the three verses heard in the prior experiment. The only difference between the experiments is that the first group of subjects heard and recalled two additional verses which the second group of subjects did not hear.

Part of the difference between the one-verse and the three verse experiments could result from a decreased cognitive load. For recalls of the common verse between the two experiments, subjects in the one-verse spoken condition have better recall than those in the three-verse spoken condition because of the reduced cognitive load. Nevertheless, that same advantage of reduced cognitive load is not found in the sung condition. Here, subjects recall the one common verse between the two experiments equally as well, regardless whether they must learn two additional verses or no additional verses. Therefore, those subjects hearing three verses sung have a definite advantage at recall even if they do have more material to remember. In addition, the music appears to be a distractor when subjects hear only one verse. Perhaps this occurs because subjects have not yet had sufficient rehearsals of the melody to make it salient or clear. Even though subjects hear the melody five times in the one-verse experiment, they should not learn the melody as well as when they hear it in a multi-verse context. In the three-verse context, encoding variability should facilitate learning (see Hintzman 1976 for a review). Since the music is not sufficiently learned to facilitate recall, it becomes a distractor. In the three-verse experiment, subjects must learn the melody better and/or be better able to use that melody as a retrieval aid.

Music provides a very powerful retrieval cue. Music is more than just an additional piece of information, it is an integrated cue that provides information about the nature of the text. The music defines the length of lines, chunks words and phrases, identifies the number of syllables, sets the pattern of stressed and unstressed syllables within the text. Thus, the music acts as a frame within which the text is tightly fit. That frame can connect words at encoding, limit retrieval search, as well as constrain guessing or recreation at retrieval.

For example, the melody could assist subjects in distinguishing one verse from the other by making each verse a coherent unit. In addition, the melody could cue subjects to search for lines to fit points within the melody thus reducing the chances of a subject omitting a line. Furthermore, the structure of the music, that is the rises and falls in the melody, the accents, et cetera; accentuate particular components of each verse and thus may make it more memorable. At this stage in the experimental work, these mechanisms are not mutually exclusive and cannot be distinguished.



There are some limitations to interpreting these results for advertising. For one, subjects in this experiment were attempting to learn the song whereas consumers watching an ad are not likely to give as much attention or effort to the ad nor are they likely to try to recall the ad, although, on occasion, consumers may make an effort to recall the brand name or perhaps a new brand claim. Second, these recalls are reported as verbatim recalls which again is an unlikely scenario for advertising. However, gist recall and verbatim recall are correlated. Given the emphasis on learning the material, a verbatim recall measure is a reasonably stringent measure. In conditions for which consumers are not intentionally learning material, recognition or gist measures should be comparable in difficulty to the verbatim measure used here. However, if you want consumers to sing a jingle over and over to themselves, you expect them to learn the words accurately and not alter the words. One of the points of the above experiments is that it is easier for subjects to learn words verbatim when they are sung.

Obviously, music should only aid recall when the lyrics are as clearly understood when they are sung as when they are spoken. In addition, it is important that the text match the music in terms of rhythmical structure, stress patterns, phrasing and points of emphasis. In order to be memorable, the i music should also have a simple form with a basic pattern of ascents and descents, and a clear rhythmical pattern. These musical factors should facilitate learning of the musical score as a frame or retrieval cue. English and Scottish ballads that are still sung in North Carolina contain these musical properties (Wallace forthcoming). Since this is a tradition that has been passed from generation to generation by word-of mouth for two hundred years, it is reasonable to assume that they are in a good form to facilitate memory. Thus they provide one example of memorable melodic structures.

By considering music's effect on recall, I certainly do not intend to imply that all music does is improve recall. Certainly there are emotional reactions to music which may spread to attitudes about the product. Music also probably serves many additional functions as well.

In conclusion, there is evidence that music can improve recall of a text and there is reason to expect that this will hold for jingles as they occur within an advertisement.


Alpert, Judy I. and Mark I. Alpert (1990), "Music Influences on Mood and Purchase Intentions," Psychology and Marketing, 7 (Summer) 109-133.

Dagnoli, Judann (1989), "Best-loved Themes Get Sweet Reprise," Advertising Age, (September), 32.

Gorn, Gerald J. (1982), "The Effects of Music in Advertising on Choice Behavior: A Classical Conditioning Approach," Journal of Marketing, 46 (Winter) 94-101.

Hintzman, Douglas L. (1976), "Repetition and Memory," in The Psychology of Learning and Motivation, Vol. 10, e 1. Gordon H. Bower, New York: Academic Press.

Kellaris, James J. and Anthony D. Cox (1989), "The Effects of Background Music in Advertising: A Reassessment," Journal of Consumer Research, 16 (June) 113-118.

MacInnis, Deborah J. and C. Whan Park (1990), "The Differential Role of Music on Consumers' Processing of and Reactions to Ads," Working Paper No. 5, Karl Eller Graduate School of Management, Tucson, AZ.

Milliman, Ronald E. (1986), 'The Influence of Background Music on the Behavior of Restaurant Patrons," Journal of Consumer Research, 13 (September) 286-289.

Milliman, Ronald E. (1982), "Using Background Music to Affect the Behavior of Supermarket Shoppers," Journal of Marketing, 46 (Summer) 86-91.

Park, C. Whan and S. Mark Young (1986), "Consumer Response to Television Commercials: The Impact of Involvement and Background Music on Brand Attitude Formation," Journal of Marketing Research, XXIII (February) 11-24.

Stout, Patricia A. and John D. Leckenby (1988), "Let the Music Play: Music as a Nonverbal Element in Television Commercials," in Nonverbal Communication in Advertising, eds. Sidney Hecker, and David W. Stewart, Lexington, Massachusetts, Toronto, Canada: D.C. Heath and Company, 207-223.

Wallace, Wanda T. (1990), "Memory for Melodies: The Effect of Learning Music and Text Together," Working Paper.

Wallace, Wanda T. (forthcoming), "Characteristics and Constraints in Ballads and Their Effects on Memory," Discourse Processes: A Multi-disciplinary Journal.