Communication Issues in Different Public Health Areas

F. Gerald Kline, University of Michigan
Peter V. Miller, University of Michigan
Andrew J. Morrison, Market Opinion Research
ABSTRACT - This paper deals with information selection by adolescents in two cities from several media concerning three public health topics: drug use, alcohol use, and family planning. A three factor ANOVA design with repeated measures on two factors is used to investigate the adolescents' message discrimination on the topics. We find significant topic, channel, and city differences in this behavior, as well as city-topic and topic-channel interactions. Implications of these findings for policy makers are discussed.
[ to cite ]:
F. Gerald Kline, Peter V. Miller, and Andrew J. Morrison (1976) ,"Communication Issues in Different Public Health Areas", in NA - Advances in Consumer Research Volume 03, eds. Beverlee B. Anderson, Cincinnati, OH : Association for Consumer Research, Pages: 290-294.

Advances in Consumer Research Volume 3, 1976      Pages 290-294


F. Gerald Kline, University of Michigan

Peter V. Miller, University of Michigan

Andrew J. Morrison, Market Opinion Research


This paper deals with information selection by adolescents in two cities from several media concerning three public health topics: drug use, alcohol use, and family planning. A three factor ANOVA design with repeated measures on two factors is used to investigate the adolescents' message discrimination on the topics. We find significant topic, channel, and city differences in this behavior, as well as city-topic and topic-channel interactions. Implications of these findings for policy makers are discussed.


The interest in "information processing" aspects of consumer activity in recent years has paralleled a similar development in the study of mass communication. Dissatisfaction with the traditional view of the media's workings in society -- the so-called "hypodermic needle" metatheory -- has led to a more intensive consideration of what uses audience members make of media offerings, rather than on the characteristics of communicators or messages which might make them powerful persuaders. Both theoretical reasoning and empirical productivity impel this recent orientation. The former treatment of the audience as an homogeneous, passive collection of individuals left much to be desired in the way of explanation for their perverse tendency to ignore, or misconstrue, varied communication efforts (Bauer, 1964). When the null hypothesis on media effects continued to prevail in areas like political persuasion research, furthermore, it became apparent that new tactics -- including a recognition of audience potency in the communication process --were required to search out the effects that did exist. Hence, we find more attention being paid these days in the mass communication research literature to the potential receivers of mass media messages, focusing on their interpersonal environments and intrapersonal predilections and abilities.

One salient on this front has been a renewed effort to identify or conceptualize the factors of interest in the communication process. Some of this work has concentrated on criterion variables, as in the recent research on "agenda-setting" by the media in the political arena. Here new dependent variables stress the media's potential ability to structure the ground rules of debate rather than the media's supposed persuasive power. [The media may not be able to tell us what to think, but they may tell us what to think about (Cohen, 1963).]

Agenda-setting research forces us to consider, from the audience's point of view, those rather broad pieces of information which may be culled from the media concerning societal problems. It is the audience member who elucidates the media's potential effects by the way he cuts up the big issues of the day. This is in sharp contrast to the research of earlier days which focused on the communicator or aspects of his message to define the universe of possible effects.

Another audience-based approach has been pioneered in England and Israel. The "uses and gratifications" perspective on mass communication examines the reasons why people choose with media content and how they use that content (Blumler and Katz, 1974). The typical flow of causality from media stimulus to audience response is simply reversed, in this case, to emphasize the receiver's needs as the force behind his choice of certain communicator offerings. The approach also gives us a better handle on possible media effects, in that the audience predispositions may be considered as mediating variables which soften or enhance the import of a particular communication (Kline, Miller and Morrison, 1974).

In this paper, we consider yet another reconceptualization which has followed from the emphasis on the receiver in the communication process. Message discrimination, as a measure of media exposure, incorporates different receiver processes than the traditional "time-spent" estimates. It also fits nicely into the information processing orientation of many consumer researchers by highlighting one of the stages in that process -- the input or "first order" stage of selecting pieces of information for further consideration (Ray and Ward, 1975).

Message discrimination tells us something about both the receiver and the sources and subjects of his reception. We learn about audience priorities for subject-matter and channels of reception, as well as the frequency and strength of messages in various channels. Disentangling the contributions of the receiver's priorities and the availability of information on certain topics in certain channels is not a simple task, and one we will not completely resolve in this paper. Rather, our analysis leads us to speculate on the channel, topic, and "media environmental" differences which may, in conjunction with receiver preferences, account for variance in the way teenagers select information from the mass media on several public health topics.

The subject-headings of interest are drug use, alcohol use and family planning. We have included an additional topic area in the analysis for comparison purposes, occupational information. In examining the amount of information selected from various media on these topics by adolescents, we hope to elucidate the interface between the macro-perspective of the communicator or public health policy maker and the micro-orientation of the teenage audience member in the transmission of public health information.


The data presented in this paper come from two waves of interviews undertaken as part of a year-long study of the social context of adolescent media use. Nearly three hundred households containing at least one adolescent between 14 and 17 years of age were samples in each of two cities (Flint, Michigan and Toledo, Ohio) which we attempted to match on media market characteristics. Three random groups were formed in each city so that we could have control and experimental groups and additional controls for pretest sensitization in a multi-wave field experiment. The resulting scheme, similar to the Solomon Four-Group design (Campbell and Stanley, 1966), was used to assess the impact of messages about family planning, alcohol use, drug use and occupational information. Teen radio stations were the vehicles for these experimental messages, which we produced.

In this report, we will restrict ourselves to non-experimental aspects of the study and concentrate on making topic, city, and channel comparisons of message perception and recall by the adolescent samples. Reports on other aspects of the larger project may be found in Kline, Miller and Morrison (1974), Miller, Morrison, and Kline (1974), Miller, Kline and Morrison (1975), and Morrison, Kline and Miller (1975).

Our analysis focuses on city, channel and topic differences in message discrimination. We chose a three factor ANOVA design with repeated measures on two factors to investigate these effects. The choice of design was indicated by the fact that our sample of respondents had been interviewed about all four topics, and had reported on their media message discrimination across seven channels within each topic. Thus, we treated the topic-channel survey questions as repeated measures on the topic and channel factors. The third factor, city of residence, naturally did not contain a repeated measures component. (See Winer, 1962, pp. 319-337 for details of the analysis procedure).


Message Discrimination

This concept was introduced into the literature of mass communication research by Clarke and Kline (1974) as a replacement for the standard "time-spent" measures of media exposure. The measure entails asking a respondent in a survey interview what he has seen or heard about a particular topic in a particular medium within a particular time frame (the maximum being one month). The interviewer records the reply and probes to see if any additional thoughts can be obtained, and the process is repeated for each of the media under consideration. The open-ended responses are coded into "messages" according to an inductively-derived coding scheme, which allows for several messages to be coded per question. An overall message discrimination score is constructed for each respondent by summing all of the messages coded for all topics and media.

Thus, we conceive of message discrimination as the selection of units of content for attention and recall out of a universe of competing stimuli. In order to evidence media exposure, the respondent must recall something he actually saw or heard while in contact with the media. The emphasis is on actual, or engaged exposure, rather than potential exposure addressed in measures of how much time was spent with a given medium during a given period. The discriminations are considered first-order cognitive processing, a necessary but not sufficient condition for integration of information into a larger cognitive structure (Kline and Davis, 1973).


The topics chosen for the study reflected our judgment about subject-matter which would be salient for a teenage audience, and concentrated on public policy issues whose unique character might be investigated. The three public health areas, for example, have been highlighted as major problems among youth in American society. Drug use is a familiar and seemingly ubiquitous issue, while alcohol use has recently become a source of even more concern to some public health experts. The shortage of correct information among adolescents regarding family planning practices is well documented, (Weiss, et al., 1974) and the taboo character of this subject-matter adds another interesting dimension from a public policy perspective. Rather than consider these three topics in isolation, we chose to offer a comparison with occupational information which, while perhaps equally important to teenagers, does not have the "health problem" character.

Channels and Cities

The media we investigated included a range from the truly "mass" type to more specialized channels; the list includes television, newspapers, magazines, radio, books and pamphlets, film and billboards. Our anticipation was that this variation in media would allow us to make some inferences about the informativeness of the channels across the different topics. Since the group contains some media usually not considered in the context of public health communication research, we may also shed some light on the nature of these channels in this specialized area. The cities were chosen by us both for their geographical location for research purposes, and because of the kinds of media mix which existed in them. We hoped for a relatively good matching, such that the two places would be approximately equal in the amount of information in their environments about the topics we considered. Furthermore, we wanted to find places which would not cross-feed information to one another. While we were successful to a degree in these goals, it remained important to consider the possibility of city differences in attention to the topics (differences in media environment") or differences in the populations studied which might produce differential salience for individual topics. From a policy view point, city differences would indicate a necessity to consider these factors when planning public information campaigns and research on them.


We would expect the combination of audience priorities and channel and topic penetration to result in differences in message discrimination by these two factors. It is likely that teenagers pay greater attention to certain channels (e.g., television) and have greater interest in certain topics (e.g., drug use). It is also likely that certain topics receive more "play" across channels, and that the frequency of information on all four topics would vary by channel.

Furthermore, we might anticipate topic by channel interactions, as some topic information is likely to be played, and received, in some channels more than others. To the extent that city differences occur, (we expect no differences) we would expect them to be the result of greater media attention to a particular topic in one city versus the other. City by topic interactions might also occur in such an event, but city-channel interactions would be quite unexpected.


The results of the repeated measures analysis of variance are displayed in Table I. We find significant main effects for all three predictor variables -- city, topic, and media channels. Significant interactions occur between city and topic, and between topic and media channel-variables. We now turn to a more detailed look at each of these significant effects.

Cities and Topics

In designing a field experiment such as the one described here, we attempted to match two cities in terms of both existent mass media and demographic factors. We hoped, then, that the baseline message discrimination levels between the two cities for any particular topic would be minimally and non-significantly different. However, as shown in Table II, there is a significant overall difference in mean messages discriminated between the two cities as well as for overall topics except family planning. We might interpret this as resulting from a higher message display in the Toledo media for alcohol, drug, and job topics. The small difference for the family planning topic could possible be attributed to intensive efforts by the Mort Foundation family planning program to get family planning information disseminated in the Flint area.



Topic differences are of more interest to us. It appears that occupational and drug messages are discriminated by adolescents much more than alcohol and family planning messages. Of course, this finding leads to a consideration of what topic characteristics might enhance information receptivity in the adolescent population. Certainly, sheer display of information might be a factor. Many kinds of messages, including want ads, television dramatizations and news stories may be perceived as occupational information, and so its "play" in the media is likely to be greater because of the size of the topic category.

On the other extreme, family planning information apparently still has something of a taboo character about it, especially as family planning matters are discussed (or not discussed) in the mass media. There appears to be more messages discriminated about drug use than the alcohol topic. This difference may be due to the greater news display about drug-related incidents (crimes, latest scientific findings, and the like) than one finds for alcohol related happenings. Moreover, messages about alcohol use may predominantly be tied to advertisements for alcoholic beverages or public service announcements warning against the use of those same beverages.

As mentioned earlier in the paper, it is difficult to disentangle the contributions of media emphasis on certain kinds of content and audience information priorities in producing a message discrimination score. Some control on content in the media would help, and where we have conducted content analyses of the channels (with regard to family planning), we have found a strong monotonicity between proportions of messages discriminated about the topic and media "play" during that time. Whether that finding would hold up for the other topics is unclear, since family planning is such a special kind of subject matter, and since our content analytic efforts have not yet extended to the other topic areas. Ideally, an experimental method would be employed in which content available could be strictly controlled and measured, and audience priorities for information manipulated. The "signal detection" literature would provide a model for such an analysis (Swets, 1964). Meanwhile, in our "actuarial" accounting of message discrimination differences by topic, we are forced to concede the ability of both audience predispositions and sender emphases in flavoring the outcome.



The city by topic interaction is due, again, to the equal numbers of messages discriminated about family planning in the two cities as compared to the other three topics. This interaction is seen clearly in Figure 1. The key finding here is the topic differences that exemplify the necessity for message producers to view these topics differently in terms of designing mass media campaign strategies in different locations.



Media Channels and Topics

We have noted above some reasons why the clear differences in messages discriminated among the four topics may have occurred. It is now of interest to determine why, across topics, significant differences in mean number of messages discriminated occur among media channels. If one arrays the media from most to least total messages discriminated across topics, the array takes on a "mass -- 'focused' -- mass" media character.

The data in Table III show that television and newspaper dominate as primary message discrimination channels while two other mass media, radio and magazines, appear to have had the fewest messages discriminated in them.

Among these four mass media, this may reflect the news and, to a lesser extent public service characteristics of television and newspapers and the entertainment focus of many magazines and radio stations attended to by this adolescent sample.



The "focused" media of books and pamphlets, billboards, and non-television films falls in between these two extremes. We perceive these media to be "focused" relative to these topics because of single topic, narrow-theme nature of pamphlets, billboards, and school movies, in particular. Some of the message discrimination differences for both the "mass" channels and the "focused" media, naturally, may be due to some purposive use of channels on the part of adolescents for certain kinds of information. While serendipity may explain the sighting of a drug use message on television, an occupational information message from the newspaper is likely to reflect a scanning of the want-ads. Similarly, purposive use of the "focused" channels is likely for certain information, particularly in the case of books and pamphlets. We would expect adolescents to have subjective probabilities, in other words, on the likelihood of encountering various kinds of content in these channels, which may be reflected in the differences in message discrimination across channels.

The topic-channel interaction is the joint product of message producer's current use of mass or focused media to display messages about particular topics and the media use habits of the adolescent audience. Occupational messages were picked up primarily from television, books and pamphlets, and billboards. On the first of these, of course, a great deal of occupational information is evident in the form of actors playing out a variety of occupational roles. Job messages seen on billboards can be attributed in the main to military advertising (Miller, Kline, Morrison, 1975), while occupational pamphlets are a prominent feature of most school counselor offices. The lack of messages seen in newspapers may indicate that 14-17 year old adolescents are not yet at a stage where they are very concerned about future job possibilities, or employment statistics. Drug topic messages were most perceived in newspaper and television media and, to a lesser extent, in books and pamphlets and films. On the one hand, this may reflect wider news coverage of drug related items in these mass media news shows. On the other hand, we may also see the influence of school pamphlet and film campaigns on drugs, as well as drug use as a central theme in movies seen at local theaters.

Alcohol messages are seen extensively on television compared to all other media, perhaps as a consequence of public service announcements. This may hold true for the relatively few family planning messages discriminated. From the adolescent information user's viewpoint, parents and peers may be more likely choices as information sources about these topics. Figure II illustrates the channel differences in message discrimination and the channel-topic interaction.




It is clear that message discrimination is not the only criterion we would like to examine with regard to communication in the public health area. We might be interested, for example, in some 'higher-order' product of information processing, such as cognitive restructuring, new knowledge, or actual behavior shifts. The latter would be particularly of interest to policy makers, one expects.

At the same time, it is evident that a report of the actual information selection by teenagers concerning some pressing public health issues -- along with a consideration of some major factors which may account for a variance in this behavior -- has some merit in its own right. The initial selection, storage and recall of bits of information on these topics from the whelter of stimuli available in the media is a process which may, in large measure, define what higher order effects are possible. Without initial contact with the stimuli and some attention to them we cannot attribute to the media any shifts in knowledge or behavior which may occur in the population of interest. And we have seen how media exposure measures often used in mass communication research do not really capture the initial processing character of attention to and perception of certain media messages, but focus instead on "potential" exposure -- a self-report of time spent with a medium. Hence, it seems reasonable to examine message discrimination in the light of some geographic, subject-matter, and channel differences, and the interactions among them In so doing, we bring both communicator policy decisions and audience selection priorities together in a single measure, and we make inferences about the macro- and micro- factors which determine initial information processing about these important topics.


Raymond Bauer, "The Obstinate Audience: The Influence Process from the Point of View of Social Communication" American Psychologist, 19(May, 1964), 319-328.

Jay Blumler and Elihu Katz, The Uses of Mass Communications: Current Perspectives in Gratifications Research (Beverly Hills: Sage Publications, 1974).

Donald T. Campbell and Julian C. Stanley, Experimental and Quasi-Experimental Designs for Research (Chicago: Rand-McNally, 1963).

Peter Clarke and F. Gerald Kline, "Media Effects Reconsidered: Some New Strategies for Communication Research' Communication Research, 1 (April,. 1974), 224-240.

Bernard Cohen, The Press, the Public and Foreign Policy (Princeton: Princeton University Press, 1963).

F. Gerald Kline and Dennis Davis, "Cognitive Similarity and Facilitation of Communication," Paper presented to The World Congress on Human Communication, Barcelona, Spain, November, 1973.

F. Gerald Kline, Peter V. Miller and Andrew J. Morrison, "Adolescents and Family Planning Information: An Exploration of Audience Needs and Media Effects," in Jay Blumler and Elihu Katz, eds., The Uses of Mass Communications: New Perspectives on Gratifications Research (Beverly Hills: Sage Publications, 1974).

Peter V. Miller, F. Gerald Kline, and Andrew J. Morrison, "Adolescents' Learning About Military Occupations in the Mass Media," Proceedings, Research Conference on the Social Psychology of Military Service, Chicago, April, 1975.

Peter V. Miller, Andrew J. Morrison and F. Gerald Kline, "Approaches to Characterizing Information Environments," Paper presented to the Annual Meeting, International Communication Association, New Orleans, April, 1974.

Andrew J. Morrison, F. Gerald Kline, and Peter V. Miller, "Aspects of Adolescent Information Acquisition on Drugs and Alcohol Topics," in Ronald Ostman and Hamid Mowlana, eds., Communication Research and Drug Education (Beverly Hills: Sage Publications, 1975).

Michael Ray and Scott Ward, "The Relevance of Consumer Information Processing Studies to Communication Research, 2(July, 1975), 195-202.

E. Swets, Signal Detection and Recognition by Human Observers: Contemporary Readings (New York: John Wiley, 1964).

Eugene Weiss, F. Gerald Kline, Everett M. Rogers, Mitchell E. Cohen and Jennifer A. Dodge, "Communication and Population Socialization among Youth: Communication Behavior and Family Planning Information," Paper presented to the Population Association of America, New York, April 1974.

B. J. Winer, Statistical Principles in Experimental Design (New York: McGraw-Hill, 1962).