Verbal Labeling Effects in Short-Term Memory For Character/Product Pairings

M. Carole Macklin, University of Cincinnati
ABSTRACT - An experiment was conducted to test whether verbal labeling would assist children in the learning of character/product pairings. It was hypothesized that cued processors (ages 6, 7, 8) would benefit, but strategic processors (ages 10, 11) would experience little or no effect. Results from the experiment were as expected and, thus, were in accord with the Predictions from information-Processing theory.
[ to cite ]:
M. Carole Macklin (1984) ,"Verbal Labeling Effects in Short-Term Memory For Character/Product Pairings", in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 343-347.

Advances in Consumer Research Volume 11, 1984      Pages 343-347

VERBAL LABELING EFFECTS IN SHORT-TERM MEMORY FOR CHARACTER/PRODUCT PAIRINGS

M. Carole Macklin, University of Cincinnati

ABSTRACT -

An experiment was conducted to test whether verbal labeling would assist children in the learning of character/product pairings. It was hypothesized that cued processors (ages 6, 7, 8) would benefit, but strategic processors (ages 10, 11) would experience little or no effect. Results from the experiment were as expected and, thus, were in accord with the Predictions from information-Processing theory.

INTRODUCTION

One promising approach to understanding age differences in children's responses to advertising is information-processing theory. Children's reactions to the object of advertising are based on the following: 1) the incoming advertising information, and 2) the information represented in short-term memory that is retrieved (Bettman 1979).

According to information-processing theory, information from a commercial is represented in short-term memory. Through rehearsal and retrieval, information is transferred to long-term memory. Permanent storage occurs with active rehearsal and elaboration of the advertising information in short-term memory.

One widely recognized mnemonic strategy affecting children's storage is verbal labeling (Reese 1962). Labeling refers to the verbalization of information to be stored when it is presented. A child can label covertly or overtly .

The ability of children to use mnemonic strategies, such as verbal labeling, is viewed as a major reason for age differences in the learning of information. Children are grouped roughly into the following processing categories: 1) limited processors, below age 6, who cannot use mnemonic strategies, 2) cued processors, 6 to 9 or 10 years, who use strategies irregularly, and 3) strategic processors, 10 or 11 years and above. who spontaneously use them.

Roedder (1981) provided a good review of these groups noting that the information-processing view interprets age differences in children's reactions to marketing stimuli in terms of cognitive abilities to store and retrieve information. The current research will highlight a storage strategy employed by the cued and strategic processors.

To be reviewed in the next section, key studies in developmental psychology have provided evidence that instructions to label have improved the developing child's success in memory tasks. Most of the developmental inquiries have involved recall tasks, however. In contrast, scant evidence has established whether similar support is indicated for tasks requiring a child to recognize, rather than to recall, previously presented information.

This former type of task, recognition, is often of focal interest to the marketer. The advertiser hopes that the child will at least distinguish, through recognition, the advertised product from the other product alternatives at the store. Rossiter (1976) provided research evidence that children had strong visual representations of cereal boxes that were not apparent through traditional measures of verbal memory. In fact, Rossiter argued that the visual representations were so vivid that children could recognize and then point to the cereal in a supermarket without verbal assistance. Thus, aids to increase recognition, such as verbal labeling to enhance visual memory, may result in a greater number of in-store influences to purchase. Given the importance of product recognition, research by the author will be presented exploring the effects of labeling on the short-term memory for character/product pairs.

THEORETICAL BACKGROUND

The bulk of evidence on memory development has emphasized developmental changes in the recall of information but the lack of changes in recognition performance. These two research thrusts will be reviewed in brief.

Most efforts at establishing the effects of instructions to label have involved serial recall tasks. A classic study was by Hagen and Kingsley (1968). They administered a serial recall task composed of sixteen trials in which eight picture cards were shown to children aged 4, 6, 7, 8, and 10 years. The experimenter sequentially showed the eight cards, then a cue card and asked the children to point to the correct matching card in the face-down position. Across trials. the order of the pictures varied.

According to theory, Hagen and Kingsley (1968) hypothesized that verbal labeling should facilitate the memory of the 6-, 7-, and 8-year-olds, but that the older children should show less or no facilitation because they implicitly verbalize. The researchers also hypothesized improvement with chronological age. In addition, they expected the serial-position curves of the 6-, 7-, and 8-year-olds who verbalized to approach those of the 10-year-olds.

In the overt label condition, the subjects were asked to say aloud the names of the pictures upon presentation. The pictures were of easily-recognized animals. In the no-label condition, children were not required to label overtly the pictures.

Hagen and Kingsley (1968) reported three main effects: age, condition, and serial position. The 6-, 7-, and 8-year-olds who labeled recalled more pictures than those in the control condition. Further, labeling was found to enhance learning primarily in the recency position but not in the primacy position. Recall by the 4- and 10-year-olds was not affected overall. The results from the 10-year-olds were not so straightforward, however.

The 10-year-olds benefitted from labeling on the recency positions of the serial-recall curve, and the interaction between age and label-versus-nonlabel did not reach significance. However, t tests showed significant differences between the label and nonlabel-control at ages 6 and 8 but not at 10. Despite the 10-year-old's benefit on the recency position, their overall performance was negated by poor recall on the primacy portion. The researchers believed the labeling by the older children interfered with other spontaneous strategies on the early part of the curve.

These data were consistent with information-processing predictions about age differences suggesting cued processors profit from labeling prompts. Moreover, in view of the primacy-recency findings, verbal labeling can be interpreted as affecting short-term storage. The learning of primacy material involves long-term storage. Other studies have concurred with these theoretical findings (Hagen and Meacham 1967, Flavell 1970, Cole et al. 1971, Wheeler and Dusek 1973).

As to the second research thrust, the preponderance of evidence suggests that recognition tasks do not require deliberate mnemonics. Impressive results with preschool children have been reported suggesting recognition does not depend on the use of encoding or retrieval strategies (Brown and Scott, 1971, Brown and Campione 1972, Brown 1972, Hagen, Jongeward and Kail 1975). None of these studies supporting the lack of developmental progress have involved verbal labeling as an encoding strategy.

Several studies have challenged the assumption that success on recognition tasks is not improved by the use of remembering strategies, however. Several studies have reported significant age increases in picture recognition (Kagan et al. 1973, Mandler and Stein 1974, Perlmutter and Myers 1974, Rogoff, Newcombe and Kagan 1974, Nelson and Kosslyn 1976, Newcombe, Rogoff and Kagan 1977).

One study (Nelson and Rosslyn 1976) specifically examined the role of verbal labels in picture recognition by 5-year-olds and college-age adults. Nelson and Kosslyn (1975) hypothesized that both the children and the adults would use labeling to maximize recognition of information. Their view assumed that recognition involved similar processes as recall and that the performance of adults would not be at a ceiling level as information-processing would suggest.

In the experimental condition, the experimenter read descriptive labels for realistic and abstract pictures. In the control condition, the same pictures were shown without labeling. All subjects were tested with a two-choice recognition paradigm (and "old" picture and a "not previously seen" one).

The researchers found main effects of age, conditions, and type of picture (realistic or abstract). Adults made almost no errors on realistic pictures, and the children in the labeled conditions performed nearly as well as the adults (an insignificant difference). The abstract pictures resulted in somewhat different findings. Labels facilitated recognition of the abstract pictures of both the children and the adults, but the children performed less well than the adults on the abstract stimuli.

Therefore, in contrast to those studies indicating no developmental changes, Nelson and Kosslyn (1976) found that adults made fewer recognition errors on abstract pictures than the children. However, the performance on the less difficult realistic pictures was about the same for the labeled condition for children and the labeled and control conditions for adults. The difficulty of the stimuli was important, therefore. Very importantly, the researchers argued that, like recall, recognition involves active encoding and retrieval processes. Additionally, adults, as well as children, can benefit from labeling prompts. Therefore, Nelson and Kosslyn (1976) have extended the information-processing predictions to recognition tasks. Yet, they challenged the notion that people who are capable to labeling spontaneously do not benefit from prompts.

RESEARCH HYPOTHESIS

An experiment was designed to assess whether labeling of endorsed-product information would affect children's success on correctly recognizing the character/product pairings. The central question of the study was the following: Do cued and strategic processors use verbal labels to increase their recognition accuracy of character/product pairings? It was hypothesized that verbal labeling would increase the learning of the cued processors (age 6, 7, and 8) but would have little or no affect on the learning of the strategic processors (ages 10 and 11). This hypothesis was consistent with information-processing predictions but the following points should be stressed from the work by Hagen and Kingsley (1968) and Nelson and Kosslyn (1976).

Hagen and Kingsley (1968) found results supportive of information-processing theory. However, they found 10-year-olds benefitted from verbal labeling but not at a statistically significant level. No three-way interaction among age, condition, and serial position occurred as theory would strictly suggest. Additionally, Hagen and Kingsley (1976) used a serial-recall task.

On the other hand, Nelson and Kosslyn (1976) used a recognition method to test the effects of labeling. With abstract pictures, they found improved performance with labeling for both 5-year-olds and adults even though the children performed less well. Information-processing would not predict significantly improved performance by the adults who could spontaneously label. For realistic pictures, the adults' performance was consistent with theory, however.

The current research stated the major hypothesis as information-processing would suggest. The performance of the younger children was expected to improve. Although theory would suggest no improvement for the strategic processors, Nelson and Kosslyn (1976) and Hagen and Kingsley (1962) found improvement (but the latter not at a statistically significant level).

A recognition task was selected as Nelson and Kosslyn (1976) elected to do. For the current study, a recognition task was deemed more important for two reasons. First, written recognition rather than written recall was deemed more appropriate in minimizing differences in retrieval due to the younger children's less sophisticated verbal skills. Second, and as briefly discussed in the introduction, recognition is important to the advertiser. The marketer at least hopes for distinction, through recognition, from alternative offerings at the point of purchase.

METHOD

A 2 x 2 experiment was designed to test the hypothesis that verbal labeling would assist the learning of character/ product pairings by cued processors but would have little or no effect on the learning of strategic processors. One factor consisted of two levels of the experimental manipulation: l) overt labeling of the stimuli, and 2) no labeling. The other factor, based on age, reflected the two processing groups of interest: 1) cued processors, children from the first and second grades; and 2) strategic processors, children from the fourth and fifth grades.

Subjects

Thirty subjects in each age level were studied: cued processors (mean age 7 years 6 months, range 6 years 10 months to 8 years 10 months) and strategic processors (mean age 10 years 7 months, range 10 years no months to 11 years 11 months). The children attended a public elementary school in suburban, upper-middle income neighborhood of Cincinnati, Ohio. There were 28 boys and 32 girls in the study.

Stimuli

Stimuli were eight 5 1/2" x 9" color cards depicting a character and a product. One set contained labels written in a sentence structure. The other set had no labels. The images were transferred to the cards by a Kodak-color copying process.

The pictures of the characters were obtained from playing-card materials available through a firm specializing in school supplies. Four of them were animated animals (2 male/2 female) and four were human characters (2 male/2 female).

The pictures of the products were real and intentionally familiar to minimize the difficulty of the task and to increase the external validity of the study. The eight products were for: candy (2), cereal (2), chewing gum (2), and snack items (2).

For the label condition, one example of a sentence was: "DEBBIE DANCER LIKES BUBBLE YUM." The left side of the card portrayed a young woman dancing while the right side showed a picture of the actual product, Bubble Yum. The pictures on the cards were identical for the no-label condition, but the labels were absent. The pictures of the characters and products were randomly assigned to the cards. The order of the carts was also randomized but held constant across conditions and subjects.

Pretest

A pretest of the characters and products confirmed their easy labeling. Nine boys and nine girls aged 6 to 10 individually rated the objects singly in randomized order with a five-point question, "How much do you think you would like this character/product?" With the use of a small poster board, the children pointed to a smiling or to a frowning face ranging from "like a lot" to "don't like at all." The children attended an after-school program at a suburban day-care center and were of a similar socio-economic background as the children in the experiment.

While not all the characters and products received the same ratings all were judged to be appealing, recognizable, and easily labeled. (At a statistically significant level of p < .05, one character was rated higher than the mean character rating and two products were lower than the mean liking of products). To minimize any unintended effects of character/product differences, the following decisions were made: 1) to randomly assign the objects to the pairs, 2) to randomly assign the order of the pairs for both exposure and testing, and 3) to hold the order of the pairs constant across subjects and conditions.

Procedure

Small groups of children (3 to 5 in number) were randomly selected from classrooms to accompany the experimenter to a large testing room. The children were seated spaciously (at least ten feet apart) so that materials were not within another's view. Random assignment was made of the treatments to the subject groups.

The subjects were told that the task involved memory for some cards showing pictures of characters and products. The children were told that they would look at some cards one at a time. The experimenter paced the exposure so that each card resulted in equal times for both conditions.

In the label condition, the children were instructed to read aloud with the experimenter the sentence on the cart. In the no-label control condition, the children quietly studied the carts which had no labels. The experimental treatments were not combined in any of the administrations Verbal interaction was kept to a minimum. All carts were collected before the recognition task.

The recognition test consisted of eight, separate answer sheets illustrating a picture of the character on the left side and four of the eight products on the right site. The three incorrect product choices were randomly selected and held constant across subjects and conditions.

Children were asked to circle the correct product that was paired with the character on the cart. No labels were provided on the test instruments, but the word "likes" was typed next to the picture of the character. The children experienced no difficulty in completing the task which took approximately ten minutes.

RESULTS

Preliminary analysis showed no effect due to sex of the subject, and this variable was omitted from subsequent analysis. Table 1 gives the means correct and the percentages correct for each condition for each age group.

TABLE 1

MEANS AND PERCENTAGES CORRECT FOR EACH AGE AND CONDITION  (N = 15 PER GROUP)

A two-way analysis of variance revealed significant main effects for the two variables: age level (F = 4.01, df = 1, p = .049) and label-versus-no label condition (F = 5.78, df = 1, p = .02). The interaction between age and condition did not reach significance (F = 2.23, df = 1, p = .141). However, t tests showed significant differences between the label and control groups at the younger age level (t = 2.88, p = .008) but not at older age level (t = .62, p = .542). The ANOVA results are illustrated in Table 2

TABLE 2

ANOVA RESULTS

DISCUSSION

The major hypothesis was supported by the research findings. Verbal labeling encouraging the use of a storage strategy was particularly effective in increasing the short-term memory performance of cued processors. The older children, the strategic processors, benefitted very little from the labeling instructions. Indeed, as theory suggests, the older children spontaneously use one or more mnemonic strategies.

The experimental findings were very similar to those of Hagen and Kingsley (1968). Both studies provided for main effects due to age level and to label-versus-no label condition. Additionally, the two-way interaction was insignificant for the current study as was true for the failure of Hagen and Kingsley (1968) to establish a three-way interaction among age level, condition and serial recall. In the recency position, those researchers found improvements in remembering by the 10-year-olds. Such a small, but insignificant, improvement was likewise indicated by the current short-term memory research for character/product pairings.

An important distinction exists between the current research and the classic study by Hagen and Kingsley (1968), however. The latter researchers used a recall task involving sixteen trials. The current project used a recognition test as an indicator of short-term remembering. In this regard, the current research was more similar to that of Nelson and KosslYn (1976).

Like Nelson and Kosslyn (19763, the current research suggested that children can use labeling to improve the recognition and the recall of information involve more similar than distinct processes. In terms of product-related information, it is not wise to assume that recognition is an automatic process less dependent than recall on active encoding strategies.

For the advertiser, the current findings suggest the use of verbal labeling to improve short-term remembering. The pairing of a tiger espousing, "They're Gr-r-reat!," with a cereal makes good intuitive sense in terms of these theoretical findings. Such a pairing should assist cued processors in recognizing the product on the shelf. Moreover, the research findings to not rule out help for the strategic processors.

As Rossiter (1976) pointed out, children's visual memory may be a critical factor in the product selection process in the supermarket. Consumers may use information stored uniquely in visual memory (Rossiter 1976, p. 527). Therefore, verbal labeling may assist the young child in storing important visual information about the product had will result in in-store recognition ant, then, purchase. The research results imply that when verbal labeling of a character with a product occurs through the advertiser's use of sing-along jingles or catchy phrases, hen the likelihood of product recognition in the store probably increases. Recognition may well result in the child influencing the purchase of the product by the adult shopper.

Future research is needed to investigate further the effectiveness of encoding strategies in a marketing context. or example, other encoding strategies, such as rehearsal, may also affect recognition performance.

Additionally, future efforts need to pursue the effectiveness of encoding strategies in ensuring the permanent storage of information in long-term memory. While labeling provides for the cued processor's retention of information for further processing, rehearsal may be needed to permit the permanent storage of information (Ornstein, Naus and Stone 1977). This second storage strategy involves the repetition or elaboration of material after its presentation as well as during. The child may need to visualize the tiger, to say his name, and to repeat, They're Gr-r-reat!," long after the viewing of the ad to ensure permanent storage.

In terms of children's responses to advertising messages, further inquiry can increase our current knowledge of processing differences among children of different age levels.

REFERENCES

Bettman, James R. (1979), "An Information Processing Theory ," Consumer Choice, Reading, MA: Addison-Wesley Publishing Company.

Bray, Norman W., Justice, Elaine M., Ferguson, Robert P. and Simon, Deborah L. (1977), "Developmental Changes in the Effects of Instructions on Production-Deficient Children," Child Development, 48, 1019-26.

Brown, Ann L. (1972), "Context and Recency Cues in the Recognition Memory of Children," American Journal of Mental Deficiency, 77, 54-58.

Brown, Ann L. and Campione, Joseph C. (1972), "Recognition Memory for Perceptually Similar Pictures in Preschool Children," Journal of Experimental Psychology, 95, 55-62.

Brown, Ann L. and Scott, Marcia S. (1971), "Recognition Memory for Pictures in Preschool Children," Journal of Experimental Psychology, 11, 401-12.

Cole, Michael, Gay, John, Glick, Joseph and Sharp, Donald (1971), The Cultural Context of Learning and Thinking, New York: Basin Books.

Flavell, John H. (1970), "Developmental Studies of Mediated Memory," in Advances in Child Development and Behavior, eds. H. W. Reese and L. P. Lipsitt, New York: Academic Press.

Hagen, John W., Hargrave, Susan, and Ross, William (1978), "Prompting and Rehearsal in Short-Term Memory," Child Development, 44, 201-4.

Hagen, John W., Jongeward, Robert H. and Kail, Robert V., Jr. (1975), "Cognitive Perspectives on the Development of Memory," Advances in Child Development, 10, edited by Hayne W. Reese, New York: Academic Press.

Hagen, John W. and Kingsley, Phillip R. (1968), "Labeling Effects in Short-Term Memory," Child Development, 39, 113-21.

Hagen, John W. and Meacham, John A. (1967), "The Effect of Task Strategy Upon Intentional and Incidental Recall," Report No. 17, University of Michigan, Center for Human Growth and Development.

Kagan, Jerome, Klein, Robert E., Haith, Marshall M. and Morrison, Frederick J. (1973), "Memory and Meaning in Two Cultures," Child Development, 44, 221-23.

Mandler, Jean M. and Stein, Nancy L. (1974), "Recall and Recognition of Pictures by Children as a Function of Organization and Distractor Similarity," Journal of Experimental Psychology, 102, 657-69.

Nelson, Keith E. and Kosslyn, Stephen M. (1976), "Recognition of Previously Labeled or Unlabeled Pictures by 5-Year-Olds and Adults," Journal of Experimental Child Psychology, 21, 40-45.

Newcombe, Nora E., Rogoff, Barbara and Kagan, Jerome (1977), "Developmental Changes in Recognition Memory for Pictures of Objects and Scenes," Developmental Psychology, 13, 337-41.

Ornstein, Peter A., Naus, Mary J. and Stone, Barbara P. (1977), "Rehearsal Training and Developmental Differences in Memory," Developmental Psychology, 13, 15-24

Perlmutter, Marion and Myers, Nancy A. (1974), "Recognition Memory Development in Two- to Four-Year Olds," Developmental Psychology, 10, 447-50.

Reese, Hayne W. (1962), "Verbal Mediation as a Function of Age Level," Psychological Bulletin, 59, 502-9.

Roedder, Deborah L. (1981), "Age Differences in Children's Responses to Television Advertising: An Information-Processing Approach," Journal of Consumer Research, 8, 144-53

Rogoff, Barbara, Newcombe, Nora and Kagan, Jerome B. (1974), "Planfulness and Recognition Memory," Child Development, 45. 972-77.

Rossiter, John R. (1976), "Visual and Verbal Memory in Children's Product Information Utilization," in Advances in Consumer Research, 3, 523-27.

Wheeler, Richard J. and Dusek, Jerome B. (1973), "The Effects of Attentional and Cognitive Factors on Children's Incidental Learning," Child Development, 44, 253-58.

----------------------------------------