Media Magic: the Use of Media Monitoring Methodology to Develop Aids Education Interventions

Carol Schechter, Academy for Educational Development
Susan E. Middlestadt, Academy for Educational Development
Lynne D. Doner, Porter/Novelli
[ to cite ]:
Carol Schechter, Susan E. Middlestadt, and Lynne D. Doner (1993) ,"Media Magic: the Use of Media Monitoring Methodology to Develop Aids Education Interventions", in NA - Advances in Consumer Research Volume 20, eds. Leigh McAlister and Michael L. Rothschild, Provo, UT : Association for Consumer Research, Pages: 302-307.

Advances in Consumer Research Volume 20, 1993      Pages 302-307


Carol Schechter, Academy for Educational Development

Susan E. Middlestadt, Academy for Educational Development

Lynne D. Doner, Porter/Novelli


Press coverage of health issues often presents a mixed blessing for social marketers, particularly when the topic is AIDS. On one hand, the issue is always in the news. On the other, the messages conveyed by the news media rarely match the messages a program manager would like to communicate to his/her target audiences.

The influence of the media in shaping peoples' perceptions about AIDS is a factor that needs to be considered in the design of AIDS information or communication programs. The AIDS Communication Support Project, a technical assistance project funded by the Centers for Disease Control, attempted to construct a media monitoring system to help program planners include an assessment of the AIDS media environment in the development of their prevention and education programs. In my presentation today, I want to review with you some of the methodological issues we faced in developing the system, and demonstrate through three program examples how media monitoring can be used as a tool in the development of AIDS education programs.

Before discussing the methods and examples, however, I'd like to give you some background about the program and organization I work for, and the agency that funds our work, to give you some context for our media monitoring task. Along with Susan Middlestadt and Bill Smith on this panel, I work at the Academy for Educational Development, based in Washington DC. The Academy, along with two other organizations - Porter Novelli and the Johns Hopkins University - was privileged to be awarded a 5 year, $13 million contract with the Centers for Disease Control to assist their AIDS prevention programs.

The Centers for Disease Control (CDC) is a U.S. federal public health agency, headquartered in Atlanta, Georgia. The business of CDC is basically the prevention and control of disease. CDC refers to itself as the "Prevention Agency" and works largely in areas that are considered traditional public health - epidemiology, surveillance, and infectious disease control. They work primarily with and through state and local health departments. The emergence of AIDS as a serious health issue in the last decade has had a tremendous impact on how CDC does business, and approximately forty percent of CDC's current budget is now AIDS funding. Since a main focus of CDC is prevention, a large portion of this AIDS budget is devoted to developing and supporting HIV prevention programs at the national, state and local level.

The purpose of our project - The AIDS Communication Support Project - is to provide marketing and communications technical support to CDC, state and local health departments, and community organizations to improve their HIV/AIDS prevention programs. As part of our work, we began to look at developing a system to monitor AIDS messages in the media to assess how the media messages supported or competed with the educational messages we were trying to deliver through our various programs.

Content analysis is used for both theoretical and applied studies of mass communication. Content analysis of media coverage in an applied setting, however, is frequently not as formal or rigorous as in theoretical studies. It is used most often in applied settings as a process evaluation tool to assess the performance of a public relations tactic, such as a press kit or media event. Here the objective would be to capture the reach and accuracy of media coverage.

For example, a press conference sponsored by the National Cancer Institute on the 5-A-Day for Better Health program (encouraging Americans to eat five or more servings of fruits and vegetables each day) generated a very successful total of 700 newspaper stories and 41 million gross impressions in July 1992. Analysis of the media coverage indicated that NCI's nutritional messages for adults and children were reported fairly accurately.

In our project, however, we wanted to take content analysis one step further, to use it as a formative evaluation tool for designing programs. Our objectives were to identify and learn about AIDS messages disseminated in the media, and to assess their prevalence. Basically, we wanted to view the world through the eyes of our target audiences to explore the "reality" the average person is likely to construct based on his/her exposure to the mass media.


At first glance, content analysis methodology appears relatively simple: obtain all the media coverage on a particular topic, read it and code it for key messages, then analyze and summarize the results. However, in our attempt to analyze media coverage of AIDS, we had to address three inter-related methodological issues:

- scope of the topic,

- taxonomy, and

- sampling frame.

All became issues because of the extensive and complex nature of media coverage of AIDS. During a one month period in November, 1991, for example, a sampling of the 15 top media markets produced over 1000 stories on AIDS compared with 93 stories on AZT and 10 stories following a publicity campaign on National Breast Cancer Awareness month. The large number of stories on AIDS pointed out how naive our original assumption was that we would just collect all the stories, code and analyze them. With such a large number of articles, we had to decide whether to collect all mentions of the AIDS coverage or only substantial mentions; and whether to focus on specific sub-topics or all AIDS coverage.

Sampling also became an issue given time and budget constraints and the extensive coverage. Decisions to be made included selection of print and/or broadcast-channels, time intervals and media markets.

The third area - and most challenging - was the construction of a taxonomy. The taxonomy is the coding system employed to capture the different issues and subjects addressed by the media. With a topic such as AIDS, sheer volume of coverage introduces interesting complexities of scale. The original taxonomy we developed contained over 400 categories. Figure 1 lists the broad (superordinate) categories we developed and Figure 2 presents detail under one category.

After working on the taxonomy for some time, it became apparent that we could not construct a meaningful taxonomy for all HIV/AIDS coverage. We would have to zero in on specific topics, focusing only on issues relevant to programmatic decisions. Let me move to the discussion of specific program examples to illustrate what I'm talking about.




The three examples I'm going to present all use stories related to Magic Johnson's dramatic announcement of his HIV status last November. Magic's announcement had a major impact on the amount of press coverage, the placement of stories and the messages addressed by the news media. HIV/AIDS stories began appearing in all sections of newspapers, including sports, style, as well as news sections. Analyzing coverage before and after this event presented an ideal opportunity to test issues in media monitoring methodology. Figure 3 shows a comparison of the number of stories in the top 15 media markets during three 4-week time intervals: 1 month before Magic's announcement, immediately following, and approximately 8 months following. Not surprisingly, the number of mentions of HIV/AIDS more than doubled in the month immediately following the announcement and has now almost returned to original level. The following three examples show how this coverage can be analyzed in more detail and used to provide input into program decisions.

CDC's National AIDS Hotline

The first example relates to coverage of CDC's National AIDS Hotline. The hotline operates a 24-hour toll free number (1-800-342-AIDS). Before Magic's announcement, call attempts to the hotline averaged 7000/day with occasional spikes depending on media coverage of issues. Call attempts skyrocketed to over 150,000/day immediately following Magic's announcement.

The increase in calls led CDC to conduct varied analyses around the public's use and perceptions of the hotline. One question raised was whether the public associated CDC with sponsorship of the hotline, and media coverage was one source of information used to answer this question. Figure 4 shows CDC hotline mentions before and after Magic's announcement relative to mentions of other state and local hotlines during the same time periods. From these data, it is clear that mentions of hotlines, both CDC's and other hotlines increased significantly. Figure 5 shows how CDC's hotline was referred to in various stories. Notice that the high frequency of mentioning CDC in referring to the National hotline. As a result of these and other analyses, CDC changed the official hotline name from the National AIDS Hotline to CDC's National AIDS Hotline to improve recall and recognition of sponsorship.

Homosexuality versus Heterosexuality

The second example looks at coverage of homosexual versus heterosexual transmission of AIDS. Part of the challenge of AIDS education is to make people aware that HIV does not only occur in gay/homosexual populations. Certain sexual behaviors put people at risk for HIV, and disease transmission occurs among straight and bisexual people as well as gay men. After Magic Johnson's announcement, it seemed likely that HIV/AIDS press coverage would focus more on heterosexual transmission and/or women affected by the disease, compared to prior coverage centering on men and/or homosexual transmission. To test this hypothesis, HIV/AIDS media coverage in the top 15 U.S. media markets was further analyzed for type of transmission discussed. The right side of Figure 6 shows the proportion of HIV/AIDS coverage mentioning women and/or heterosexuality in conjunction with HIV/AIDS for the three time periods. Before the announcement, 25% of the stories mentioning women and/or heterosexual transmission; immediately after, this increased to 37%. Clearly, Magic Johnson's announcement influenced the discussion of heterosexual transmission factors in the media.

From a program planning standpoint, knowing that coverage shifted from focusing on HIV/AIDS as a "gay disease" affecting mostly gay or bisexual men to focusing on women and heterosexual transmission is important. A first step in a public education program is to raise awareness. Monitoring what the media are covering can help us gauge when the public is likely to be aware of a particular issue. Knowing that Magic's announcement and the resulting media coverage raised awareness of heterosexual transmission allows program planners to take advantage of this and further promote this message.

A next step in this particular program example would be to analyze the types of messages contained in the women/heterosexual stories and assess whether they are accurate and if they reflect program strategies. It would also be important to assess the prevalence of other messages carried by the media. Some of these messages may be harmful because they are incomplete or inaccurate. Others may alert program planners to emerging issues, allowing them to determine if their communications should address such issues.









Coverage of condoms in schools

The third example looks at media coverage of condom distribution in schools. School condom distribution is a controversial proposal that has been introduced in a few urban centers. Figure 7 shows coverage of this topic, before and after Magic Johnson's announcement, nationally in 15 markets and in Washington D.C. where such a proposal was introduced. Analysis of this coverage could be very helpful to school districts and municipalities considering such an initiative. The media coverage would alert program planners to the salient issues raised by opponents in other geographic areas and which issues were picked up by the press. Based on this assessment, program planners could do "advance work" in their communities around these issues with community gatekeepers in the schools, churches, media and other key segments of the community. Under this scenario, media analysis could be used to construct a case study of a controversial topic.


Content analysis in the form of media monitoring is a consumer research tool that has a number of practical applications for program planners. The three examples that have been presented today demonstrate its use in assessing sponsorship recall, determining when a message strategy shift may be warranted, and illustrating its use in providing case studies. The methodology, however, can be complex and imprecise, and careful design and planning is necessary to ensure its usefulness. Our recommendations based on this experience are that media monitoring is a valuable tool for program planning, but for it to be cost effective, researchers should work very closely with program planners, and the scope of the project should be carefully delineated to answer specific program questions.