Research on Marketing and Children: Upside Or Downside on the Product Life Cycle?

Scott Ward, Harvard University and Marketing Science Institute
ABSTRACT - Session papers are reviewed in terms of three criteria which might be applied to indicate whether research on marketing and children will form an important and lasting sub-area of consumer behavior research, or whether research in this area is merely a temporary "hot topic."
[ to cite ]:
Scott Ward (1979) ,"Research on Marketing and Children: Upside Or Downside on the Product Life Cycle?", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 427-430.

Advances in Consumer Research Volume 6, 1979      Pages 427-430

RESEARCH ON MARKETING AND CHILDREN: UPSIDE OR DOWNSIDE ON THE PRODUCT LIFE CYCLE?

Scott Ward, Harvard University and Marketing Science Institute

[ED. NOTE: Professor Ward is author of the famous "KID'S TV -- Marketers on Hot Seat" article, Harvard Business Review, July 1972.]

ABSTRACT -

Session papers are reviewed in terms of three criteria which might be applied to indicate whether research on marketing and children will form an important and lasting sub-area of consumer behavior research, or whether research in this area is merely a temporary "hot topic."

INTRODUCTION

Children are a minority group around the world, and consumer researchers who study children are also a minority group. As such, they must endure certain indignities--witness, for example, that the only pejorative session title at this whole conference is applied to this session! Instead of dignified-sounding labels like "multi-attribute...," "multidimensional..." or "information overload," we are branded with, "from tots to tarts!" But the individuals who prepared papers for this conference, and the individuals who showed up at the session presumably have some common interests in what they see as a worthwhile set of consumer research problems. Is it worth the time and effort?

Recently, Alan Andreasen examined some historical trends in consumer research and he demonstrated that areas of research interest have a life cycle like products and people (Andreasen, 1978). The frequency of articles and papers dealing with marketing to minorities, for example, followed a sort of product life cycle curve over a several-year period. One thinks about other "seasons" of consumer behavior research topics: personality and buying behavior, diffusion of innovations, and so on. I think we'd pretty much agree that consumer research that somehow involves children is still in the "growth" stage of the life cycle; the question is, will studies in this area lead to a more or less permanent place in the systematic study of consumer behavior (and be freed of cute labels for convention sessions)?* Is the evolving model closer to child psychology--a more or less permanent and important aspect of the study of human psychology--or is the model closer to narrow areas of study which are important, but temporary, since their only source of demand derives from current environmental issues and conditions?

I submit that the answer to this question depends upon the answers to three more basic questions:

1. Does the study of marketing and children give the field of consumer research a "new look?" That is, does it advance our knowledge in areas which the producers and users of consumer research find important and useful?

2. Do studies in this general area exhibit--and contribute to--conceptual clarity, and the development of theory?

3. Is there attention to relevance in research, to marketing practice, and/or to public policy decisions?

Let me take these general questions one at a time, and evaluate the papers at this session as specific object lessons.

1. DOES RESEARCH BRING A "NEW LOOK?"

First, studies of marketing as it relates to children and families clearly have the potential for giving the field a "new look," and one which would seem to be potentially important and useful. The focus is on inter-individual processes--a trend away from earlier traditions in consumer research which focused exclusively on intra-individual processes while ignoring social processes. There is also great attraction in studying consumption behavior as it "actually occurs." We are not bound to contrived experiments with college sophomores, or to yet another tedious questionnaire mailed to house-persons. There is intuitive, as well as empirical support for the notion that many purchases are influenced by, or because of, "family" considerations, and children themselves are consumers whose attraction to marketers increases as they age. True, the birth rate is down, and singles and oldsters are emerging as other consumer phenomena worth studying, but the family seems to be a venerable social institution, with important economic consequences.

The authors of papers presented here recognize that previous research on family consumer behavior has largely focused on husbands and wives and ignored children. To the extent that research contributes to conceptual models which include offspring, we will have a "new look," indeed.

2. CONCEPTUAL CLARITY AND THEORY DEVELOPMENT

What about the second question: research in this area exhibiting conceptual clarity, and usefully applying, and contributing to, theory? We are blessed with a surfeit of theoretical approaches to the study of children which can help us understand their consumption behavior, and their role in family consumption behavior. Zigler and Child (1969) in their seminal review of socialization research, identify no less than seven major theoretical approaches to study in this area. Add to that various theoretical notions in areas designated something like "family sociology," and you have even more. The dominant theory in child psychology these days is cognitive development theory, so it is not surprising that this theory dominates research pertaining to marketing and children. It's useful in describing and explaining the thoughts and behaviors of children at different stages of the early life cycle, and the neo-Piagetians have usefully extended the theory to suggest some implications for consumer researchers who are by and large more interested in applications, e.g., the possibilities of training different-staged children to more fully understand advertising messages. Consumer research can not only "borrow" from existing social science theories, but usefully extend these theories--if we choose to do so.

A first step toward useful applications and useful extensions of theory is carefully thinking through concepts and their operational measures we employ. What do we mean, for example, when we speak about children's influence on family decision-making, as Jenkins and Nelson have done? Are we talking about (and measuring) relatively direct efforts of children to influence specific purchases? Or are we talking about the oft-cited phenomenon of "passive dictation," that Wells has discussed (Wells, 1965) in which some parental purchases are partly predicted by their perceptions of children's unstated preferences? Are we talking about parental decision-making which occurs de facto because children are present, but in the absence of any "influence" from them (e.g., baby cribs and term insurance), or decisions adults make as parents although their children may not even be able to vocalize preferences before the fact (e.g., preschoolers being dragged around the Smithsonian)? If there is significant inter-subject variation in the definition of "influence" of children as Jenkins found, then it seems to me that the researcher ought to narrow and standardize that definition, so we know what results mean. That task, in turn, depends upon one's underlying, conceptual understanding of that thorny concept, "influence." The same can be said, incidentally, about the concept of "involvement," as in Nelson's paper. It's necessary to ask, influence on what, or involvement in what. Jenkins is interested in children's influence on a range of general and specific product and service categories that adults buy. But why were the decision categories chosen? They were probably used in previous research on husband-wife decision making, but it is difficult to understand why children would be expected to have much "influence" on major appliance, automobile, furniture, and savings decisions, especially if the concept of influence implied to some parents a notion of a relatively specific and direct request by children. Parent's decisions to buy furniture, insurance, and the like may be affected by the presence of children, and "influenced" in this way, but it's difficult to envision a kid asking Dad for a Maytag dryer or a Sealy Posturepedic.

Some conceptual and measurement issues also attend Nelson's study of children as information sources in the family decision to eat out. It is not clear how involvement is conceptualized, and how valid and reliable the measurement of the variable is. Depending on how respondents were defining the term, the following scenarios might each result in parents saying their kid was quite important, or highly involved, even though different processes--with different implications--were going on. Consider: the family is driving along and the kid is throwing a fit in the back seat because he/ she wants to go to McDonald's. But a second scenario is: the family is driving along and one parent decides to stop because he/she intuits that his/her kid likes McDonald's and would enjoy having a hamburger. A third scenario might be: the family is driving along and even though the parents would vastly prefer to stop off at the Cote Basque, they stop at McDonald's because that's where they have the time and money to feed their brood.

Are these trivial distinctions? I think not. If we are to have a satisfactory conceptualization of intra-family influence processes, then the nature of that influence must be clearly defined and delineated. From a policy perspective, critics of marketing to children are always asserting that children have enormous influence on parental purchasing, and they are envisioning the back-seat tantrum variety. Other less malignant processes may be going on. On the other hand, it is misleading to conclude that the husband-wife researchers have been right all along, and children have little influence on purchasing, if we ask about purchases for which children should not be expected to have much influence, or about purchases which are defined so broadly (grocery products) as to be of little use.

There are some nitty-gritty issues in the Nelson and Jenkins studies. Briefly, in Nelson's study, data are based on retrospective reports from one family member's perspective; sample procedure was not random; the study did not control for which family member answered the questionnaire, and, as Jenkins' study shows, this can make a difference. Finally, I am not sure how valid and reliable the six decision stages are, despite their popularity in other research. There would seem to be a lot of situational variables involved, which would lead to a lot of variation in how respondents interpreted their decision stages. In Jenkins' study, my main criticism--aside from my concern as to why the particular decision categories were chosen--is the lack of an underlying theoretical conceptualization. I had that uncomfortable feeling one gets whenever reading explanations of regression equations which show results--statistically significant, no less--but which are difficult to explain because there is no a priori theory, or meaningful hypotheses. The author has some underlying hunches, as evidenced by including F-scale measures and self-confidence items in the equation, but these need to be articulated and used as a basis for some hypotheses, or at least, expectations for the data.

3. RELEVANCE TO PRACTICE

Now let me turn to attention to relevance to management practice and/or to public policy issues--the third criterion I suggested to evaluate whether studies in this area hold the long-term promise of establishing an important sub-field of consumer research, or whether we are on the downside of the PLC. In my opinion, the Bjorklund and Bjorklund study exhibits the most thorough and interdisciplinary conceptualization of any of these papers, and the study shows a great deal of procedural rigor. However, it is also the paper that gives me the most trouble in trying to determine its relevance. There may be some relevance to toy companies which use observational techniques to assess the amount of time children play with prototype toy products, but this possibility is not discussed by the authors. They imply that toddler satisfaction with their toy environment has an effect on consumer socialization processes later, but there is no discussion of exactly how this might occur, or how their data bear on the process. Bjorklund and Bjorklund do cite studies which relate play satisfaction to developmental outcomes and cognitive performance, but they do not say what these developmental outcomes are, or how they might relate to consumer socialization processes. Finally, I have some difficulty with the assertion that time spent with a toy are valid measures of satisfaction: there is a tautology here, it seems to me. Admittedly, toddlers can't verbalize "satisfied'' attitudes, and even among adults, conceptualization and measurement of consumer satisfaction is currently a major interest in consumer research. So perhaps we can't be too critical of the satisfaction measure, although one might suspect that the novelty of a cornucopia of toys boggled the little kids, and it might have been better to observe play time and behavior after children had been given a chance to familiarize themselves with the toys.

Jay Lindquist broaches--but does not quite fully address--a potentially important marketing management and public policy issue. In the current regulatory climate, the thought has occurred to many advertisers who target messages against child audiences to get "out of the kitchen" and move some of their media dollars to advertising vehicles other than television. Assuming that children's attitudes toward advertising in these vehicles is related to advertising effectiveness, managers should be interested in Lindquist's study. But just think how much more interested they would be if there were some information on media patterns among children in the age ranges surveyed, i.e., some information to help the advertiser evaluate the trade-off between reach and cost efficiency using television advertising, versus the possible credibility gains in using print media. In this study, for example, some available data from commercial sources would have been helpful to give readers a feeling for how many children read what kind of magazines, comics, etc., versus viewing and listening patterns. It also would have been helpful if the author had established whether or not children ever heard of, read to some extent, and like, any or all of the publications (or other publications) they were shown as examples in class.

But let's get back to the managerial question of whether there are effective and efficient alternatives to television advertising to reach and influence kids. The overall finding in Lindquist's study is that children are most "negative"--in terms of overall attitude--toward television advertising. But a danger in using total attitude scores (TOTATT) is that some of the most interesting and useful information is obscured. If you look closely at Table 6, at the components of attitude, you find something interesting. Take sixth-graders, who are the best readers, have the most money to spend, and, perhaps, influence some parental purchases more than younger kids. It turns out that these children (and, in fact, kids at all grade levels) feel TV advertising more objectively describes advertised products than child magazine advertising; also, sixth graders feel TV advertising is more persuasive, and there's not much difference in their attitudes about believability and trustworthiness. If an advertising manager looked beyond the main reported result, he might decide that it isn't so hot in the kitchen after all.

There is a double-edged sword here in that too little attention has been devoted to measure development and testing, and it's useful to provide further testing of Rossiter's attitude scale. But the availability of measures sometimes blinds us to the most interesting and useful aspects of phenomena, while it steers us away from the tedious work involved in measure development. In this case, the most compelling data would seem to relate to how children process information in advertising in different media. Beyond attitudes, do children select, evaluate and use information differently, depending on the medium in which it appears? There is a good deal of literature in this area, and one hopes that, in future work, this literature will be explored and used as a conceptual basis for more explicit analysis of children's processing of advertising information in various media. This would help to form a basis for hypotheses, which would avoid the problem in this study that there is little basis for explanation of results. There are some hunches, but it is far more compelling to articulate hypotheses on the basis of an underlying conceptual framework than to offer explanations post hoc.

A RE-EVALUATION AND FUTURE DIRECTIONS FOR RESEARCH

What can we learn from these studies about the shape and direction of knowledge about marketing and children? With the exception of Nelson's study, the papers here are all billed as preliminary or exploratory work. Consumer research has been criticized for having so much work in the "exploratory" stage, but anyone who has tried to do research with kids knows there are some unique pitfalls not encountered in more established modes of research with sophomores or with adults. It is advisable to go slowly, and the criticism is only valid if exploratory efforts are not followed by mere definitive studies.

All of these studies pay attention to previous research. The Bjorklund's are particularly creative in reviewing and using work outside of marketing; Lindquist, Nelson and Jenkins each employ measures and frameworks used in previous research. There are problems in doing this, and I've tried to point out, but at least these authors have attempted to blunt another criticism often made of consumer research: that we do not pay enough attention to developing standard and reliable measures, and that we ignore useful frameworks in previous research,

There is also much to be said for the specific focus of the studies here. Jenkins and Nelson look in detail at children's influence on specific family decisions, and Lindquist focuses on a compelling management question. But the feeling I have is sort of like watching potters at work, shaping objects by holding them up against one of those big round wheels. These studies are just "one turn" away from the kinds of finished products that can really make substantive contributions to the field. In some cases, the "final turn" involves conceptual work, in others, greater attention to appropriate methodologies to the research questions.

For example, a recurrent problem concerns the reliability and validity of data from children--especially data based on verbal responses. The younger the kids, the worse the problem. The choices are to observe kids in some natural or artificial state, or to ask them questions. On the other hand, we can ask parents to recollect and reliably report things about kids. The papers here illustrate all of these approaches, and the authors candidly acknowledge the problems attending each of them. The choice of a research approach depends on careful delineation of the problem for research. For example, the question of children's influence on parents' decisions suggests that we ask parents some questions. The more specific the questions, and the more discreet and objective the response categories, the better the data. Are we interested in the process of influence? What aspects of the process? Parental responses to kinds of direct and indirect requests by kids, or buying decisions parents make for kids regardless of their requests? Is the interest in the range of goods and services children influence? Or is the interest in the independent variables which predict different frequencies, types, or processes by which children influence parental purchases?

What about research problems which demand that we gather data from kids themselves? The issue is really not experiment versus survey, or verbal versus behavioral measures. The issue again seems to me to be the problem you're interested in. In some of our current research, we are interested in children's processing of information in television advertising differing in the type of appeal used, and the degree of repetition (Wackman, et al, 1977). So we have asked children questions, but before we did, we tried to plan the research so we had an a priori conceptual framework to use in coding the open-ended responses. We hope the results will be pertinent to policy decisions about types of appeals which foster or inhibit children's learning from commercials, and the further applied question of how educational materials could be designed to help children understand, evaluate and use advertising content better.

To sum up, the methodological choices are pretty straightforward once the objective of the research is set. And my belief is that the objective ought to be related to some identifiable decision alternatives. The studies here are pertinent to management decisions such as media scheduling and appeal formulation, depending on advertising effectiveness in different media reaching kids, or depending on what we can tell managers about how kids affect parental buying decisions in families. And these questions are relevant to public policy questions too, since they bear on the issues of the extent to which children bug parents to buy things, and the nature of parental mediation of children's influence--as Nelson and Jenkins show, its a complex process, indeed.

It's my feeling that attention to relevance distinguishes consumer research from more discipline-oriented research. Moreover, the most useful research proceeds from a basis of conceptual clarity. To the extent clarity and relevance guide the choice of research questions concerning marketing and children, I feel this sub-area of research can be a productive and methodologically sound on-going tradition in consumer research, rather than a temporary rash of exploratory studies.

REFERENCES

A. R. Andreasen, Remarks at Tenth Annual Paul D. Converse Awards Symposium, University of Illinois, Champaign-Urbana, Illinois, May, 1978.

E. Zigler and I. Child, "Socialization," in G. Lindzey and E. Aronson (Eds.), The Handbook of Social Psychology, Vol. 2 (Reading, Mass., Addison-Wesley, 1969).

W. D. Wells, "Communicating with Children," Journal of Advertising Research, 5 (June, 1965), 2-14.

D. B. Wackman, E. Wartella and S. Ward, "Children's Information Processing of Television Advertising," National Science Foundation Grant, 1977.

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