Contents - Previous - Next

Part III: Applying data to programs

Unit 18. Developing a Health Communication Strategy with and for Women


To develop health education messages and a strategy for communicating these messages for women and their families that will help them to seek timely and appropriate care for their health problems. The development of this strategy should be accomplished with the full input and cooperation of the community.


Community participation is integral throughout the process of developing health interventions for women. This unit, and the ones that follow, describe one process for using the information collected using this protocol.

The development of interventions for women can be broken into seven stages:

Stage I

Organize new or locate existing groups of community members to provide input, feedback, and guidance to the intervention development process.

It will probably be necessary to organize several different types of groups:

a) Local women;

b) NGO/service organization staff;

c) Community leaders and other persons of local authority;

d) Local health providers (including traditional healers).

In addition to these different kinds of groups, it will be necessary to work with individuals. In particular, the key informants developed during the process of conducting this protocol may be important.

These groups should not only provide input regarding interventions developed for their communities, but should also exercise some authority and control over the interventions themselves.

Ideally, a set of the data collected using this protocol will be left with the group(s) which they can use as a resource for planning. In situations of low literacy, it may be necessary to summarize the information in more visual formats.

Stage 2

Develop a set of intervention ideas during "idea-generation sessions with community groups. (Further described below)

Stage 3

Conduct further data collection (if needed) using qualitative and quantitative methods to assess the feasibility/form of each intervention.

Further data collection may involve doing additional, but much more focused interviews with a small number of key informants. For instance, if you felt that face-to-face communication during home visits might be an appropriate mechanism for disseminating health messages to women, it might be necessary to explore some basic issues about current home visiting patterns. When are women available to be visited? Do women currently visit each other? What kinds of interactions are deemed appropriate for home visits?

Stage 4

Pretesting intervention ideas in a modified form to community members. These may be conducted as small focus groups and/or individual key informant interviews.

Stage 5

Eliminate unacceptable ideas for interventions (based on the focus group/key informant sessions) and/or further alteration and refinement of existing ideas.

Stage 6

Pilot testing of the different interventions in the community.

Stage 7

Initiation of full scale health communications strategy.

In addition, Units 18 and 19 of this protocol contain guidelines for developing interventions. These two units represent Stage 2 activities.

Idea Generation Sessions to Develop a Health Communication Strategy for Women

A series of idea-generation sessions should be conducted with different community groups (described above). These groups should go through a series of steps in the development of a health communication strategy:

1. Assessment of protocol data.

2. Identify target behaviors.

3. Identify target individuals/groups.

4. Select appropriate vocabulary.

5. Create messages.

6. Select media.

7. Put it all together into a unified communication strategy.

These steps are described in detail below. For most of the development steps, specific units from this protocol have been identified, which can supply information. The level of involvement of each of the different groups for each of the steps will necessarily vary depending on local circumstances. However, as a general principle, greater involvement by community people will mean a greater chance for success both in the short run and in terms of sustainability.

1. Assessment of protocol data

During the idea-generation sessions, the following analyses should be reviewed:

Form 1.1 Diagram of Study Community

Form 2.2 Sample Data Presentation Matrix

Form 3.2 Free Listing Tabulation of General Women's Health Problems

Form 4.2 Free Listing Tabulation of Women's Illnesses

Form 4.3 Free Listing Tabulation of Signs and Symptoms of a Women's Illness

Form 5.3 Tabulation Sheet for the Pile Sorts of Women's Illnesses

Form 6.2 Type of Practitioners

Form 7.2 Ranking of Practitioners

Form 7.3 Tabulation Sheet for Practitioner Characteristics

Unit 9 Diagram of Factor Influencing Women's Decisions About Their Health

Unit 10 Flow Diagrams for Different Illness

Form 10.3 Case Summary Form for an Illness Category

Form 10.4 Illness Summary Sheet

Form 11.3 Overall Tabulation of Women's Activities

Form 11.4 Women's Daily Activities Composite

Form 12.2 Illness Severity Tabulation Sheet

Form 13.3 Comparing Interrelationships Between Illnesses

Form 13.4 Creating an Ethnomedical Model of Illnesses

Form 14.4 Consensus Picture of Woman's Body: Location of Internal Organs and Important Illnesses

Form 15.2 Tabulation of Responses

Form 16.2 Tabulation Sheet for Diagnosis with Scenarios

Form 16.3 Tabulation Sheet for Home Remedy Treatments with Scenarios

Form 16.4 Tabulation Sheet for Outside Home Care-Seeking with Scenarios

Form 17.3 Tabulation of Events in Health Treatment Setting

Use the interpretation suggestions provided with each unit as an initial means of looking critically at the data.

As specific ideas for communication strategies arise, the following aspects should be considered: cost, personnel involved, appropriate timing (what specifically would/could happen with each possible intervention); further data that should be collected to assess the feasibility/form of a particular intervention.

An initial assessment should be made of the kinds of information needed to communicate to women: What kind of education is needed? What are the existing levels and types of local knowledge about the problem? What health problems are women concerned about? What are the beliefs about the management of these health problems? How does this correspond to the appropriate biomedical concerns for women's health? How do women interpret the severity of symptoms? Are women prompt in seeking care? What constraints to prompt care-seeking can best be addressed through educational programs? (See Unit 19 for improving services.) What practices are harmful and should be discouraged? Which practices should be encouraged?

2. Identifying Target Behaviors

It is important to determine a set of key behaviors of women that place them at greater risk of health problems. By focusing on a limited set of target behaviors, interventions can be both focused and interlinked in a way that will enhance communication. Selection of target behaviors may be assessed in several ways. While it is clear that all these criteria cannot be met for every health behavior of interest, some guidelines for selection are:

- Behaviors are clearly (causally) linked in a biomedical sense to a health problem of concern.

- Behaviors occur across all or most areas of the study site.

- Behaviors occur with some minimum frequency with all areas.

- Behaviors are locally recognized to be problematic.

Examples of the kinds of behavior that might be targeted include:

- Delay in care-seeking for white discharge.

- Lack of sanitation in bathing.

- Lack of sanitation in menstrual practices.

- Not doing a full course of treatment for a RTI.

- Multiple sexual partners.

- Unprotected sexual intercourse.

- Specific food avoidances which may place women at risk of nutrient deficiencies.

Sources of Data: Units 2, 5, 6, 7, 9, 10, 11, 16, 17.

After identifying a series of target health behaviors, use the guidelines mentioned above to help prioritize the list of behaviors, as well as issues of cost, personnel, and so on.

3. Identifying Targets/Individuals/Groups

It is efficient and effective to select appropriate subgroups within the population on which to focus intervention efforts. In some cases, individuals for a behavioral intervention will not be the person performing the behavior, but will be someone with authority over that person. Think about whose behavior needs to change. Then think about who influences this behavior. For example, suppose a married woman has excessive menstruation and wants to seek treatment from an allopathic doctor. However, her mother-in-law says that allopathic medicine will make her infertile and says she should visit the local healer instead. In this instance, the mother-in-law exhibits influence over her daughter-in-law's treatment decisions, and may be an appropriate target for health education.

Some other issues to consider include: Which women are more knowledgeable about their health problems than others (older, younger, married, unmarried, certain religious or ethnic group, less educated, lower economic status)? Which women are more vulnerable to health problems, especially untreated health problems? Are there other groups of people (men, folk healers) who need to be targeted by a program?

Sources of Data: Units 3, 4, 5, 6, 7, 9, 10, 11, 16, 17.

4. Selecting Vocabulary

Use vocabulary that is familiar to your target audience. Specifically, choose vocabulary for health messages which includes local illness term/ signs, etc. Look back at the interviews with women and local health providers and ask: How do women describe their health problems? What words do they use with each other? With providers? Which words should be used for face-to-face contacts (more familiar or more respectful)? Which are more appropriate for mass media campaigns?

Use words that seem to be less ambiguously applied to specific health problems. In many cases, women are more accurate and consistent in their assignment of labels to signs and symptoms than when they use local illness terms. Use terms which show more consensus/frequency between respondents - this means they are more salient to people and will be more likely to be acted on consistently.

Sources of Data: Units 2, 3, 4, 5,6,9, 13, 14, 15.

5. Creating Messages

Use your knowledge of appropriate local vocabulary and subgroups to be targeted to develop a series of messages. Messages/health education materials should make use of one or more of the following general guiding principles:

a. Work within or around existing local belief systems. Try not to directly contradict local beliefs as this may undermine the credibility of your messages. For example, acknowledge local beliefs about hot foods and their relationship to safed paani - and then talk about other potential causes.

b. Consider use of culturally appropriate metaphors. Use locally appropriate metaphors/ analogies to explain key health concepts (Nichter 1989; Herman and Bentley, 1993).

The selection of appropriate metaphors should:

- Involve material things/activities that are familiar (ideally on a daily or weekly basis) to all targeted individuals.

- Pull together multiple related concepts. Can be connected with other concepts you wish to communicate to women.

- Be rigorously tested for understanding (and potential misunderstandings) on key informants. For example, in South Asia effective analogies for health and nutrition matters frequently involve the use of agricultural concepts. Fertilization of fields is similar to feeding a pregnant women a healthy diet.

In Northern Ontario, diabetes education for Ojibwa-Cree Native Americans has used the metaphor of the snowmobile to teach about the function of insulin. Just as the key allows you to use the gas to make a snowmobile run, so does insulin allow you to use the food energy to make your body run. This example also was successful in pulling in the related concept of eating a good diet. Just as you need the proper fuel and mixture with oil for a snowmobile for it run, so does a human body need appropriate foods in the right combination.

c. Clarify disease etiologies using appropriate local vocabulary.

d. Emphasize existing positive behaviors. It is important to let people know what existing behaviors they should continue. Supporting common knowledge is also a good way to build rapport with local people and establish an equivalent relationship.

e. Interlink health education messages. Messages should be interlinked, mutually supportive, and build on each other.

Besides these general guiding principles, it would be useful to create a project-specific set of more focused guiding principles. These specific principles may relate to cultural features you have uncovered using this protocol. For instance, you may decide to make it a principle to reinforce existing perceptions about the local food classification system because it appears to advantageously allocate micronutrient-rich foods to women during menstruation.

Sources of Data: Units 1,2,3,4, 5, 6, 7, 9, 10,11, 12, 13, 14, 15, 16.

6. Select Appropriate Media for Communication

Two main mediums for health communication are possible: mass media and localized media. Mass media, such as radio, television, poster campaigns, should be considered whenever resources permit. The information collected in this protocol is especially useful for eliciting potential types of localized media. These include:

a. Identifying informal social/communication networks of friends, neighbors, and relatives and determining key (focal) actors within these networks. These focal actors may make effective lay health communicators.

b. Use of school children as a means of carrying home messages and information about issues such as basic sanitation (most likely, not about care-seeking for RTIs).

c. Identifying more formal organizations for passing on information, such as churches, women's group, etc.

d. Identifying local newspapers, radio stations, etc., that may be used for disseminating health information.

7. Putting it Together

Message, targets, and media must all be combined appropriately into a package which is economically and logistically feasible. The overall strategy should follow a logical pattern in terms of reinforcement and frequency. Reinforcement refers to the number of different media used to communicate a series of health messages. Frequency refers to the number of times a message is communicated.

These two concepts heavily overlap in practice. Base reinforcement and frequency on the earlier prioritization of significant problematic health behaviors. Those behaviors of higher priority should be more reinforced (i.e., communicated using more forms of media) and more frequent (within a particular media form).

A third aspect of an overall communications strategy is sequencing, creating a logical progression of messages (assuming there are multiple messages). More complex messages requiring integration of multiple concepts should occur later in the progression of messages.

Following these seven steps to develop a preliminary health communications strategy, continue with Stages 3 to 7 (further information gathering, pretesting, refinement, piloting, and implementation).

Unit 19. Improving Services Provided to Women


To develop interventions to improve the quality of health services provided to women.

This unit presents some
Ideas on interventions to improve health services to women. These ideas should be developed, refined, and pretested using the same seven stage framework described in Unit 18. The suggested activities described below fall into stage 2 of the framework.

Idea Generation Sessions to Improve Services for Women

1. Develop educational materials for health providers

Initial discussions with health providers should assess responses to some of the following questions: Do they feel they need education about local social, cultural, economic, and environmental conditions? Do they have experiences in working with women with misunderstandings? How will helping them understand local beliefs (e.g., local ethnomedical model), perceptions, and reasons for health behaviors help them to do their jobs?

The format for this kind of intervention is highly variable. It may involve developing curriculum or orientation materials for providers during initial training or periodic in-services. It may involve developing a series of field trips or guided data collection experiences in the community.

Products developed should enhance understanding and communication skills of health providers. When developing materials, select appropriate forms of popular local vocabulary for health care workers, so they can communicate effectively with women.

Sources of Data: Units 2, 4, 5, 7, 10, 12, 13, 14, 16, 17.

2. Develop ways to help providers ask questions about women's health problems more effectively

In many settings, discussion of women's health problems are considered sensitive and difficult. Women may use euphemisms to describe their illnesses. For example, in many parts of India women will commonly use body ache or more generalized reports of weakness when they have white discharge.

Appropriate educational materials here may involve the development of one or more question patterns, much like the WHO sick child algorithms for the diagnosis and treatment of women's health problems. These could instruct the health providers how to ask about signs and symptoms. It would tell them what euphemisms are used locally for women's illnesses. To do an adequate job of developing such an algorithm may take additional epidemiological studies comparing women's reports of their own signs, symptoms, and local illness terms to a series of "gold standard" biomedical tests.

Sources of Data: Units 2, 4, 5, 10, 14.

3. Identify and develop ways to overcome local constraints to care-seeking and to successful treatment outcomes

As a first step, the groups must identify different types of local constraints: economic, social, cultural, logistical, transportation, hours of operation of the clinic/health center, and reputation/gender of the health provider, and so on. The sensitivity surrounding women's reproductive health problems is likely to be a major constraint in many settings.

While reviewing the data, groups should try to answer the following questions: What constraints could be dealt with by changing the way services are provided? Which must be dealt with by changing local beliefs and practices? How can health providers increase rates of successful treatment outcomes ("compliance") by identifying constraints m improving the condition or women's health and by suggesting locally appropriate strategies to deal with these constraints? How can women be encouraged to come in for treatment? How can they be encouraged to come back for follow-up? How can existing services be modified to meet women's needs, e.g., make better use of village health workers.

Sources of Data: Units 1, 2, 3, G. 7, 9, 10, 11, 16, 17.

As in Unit 18, following the generation of a series of ideas, the intervention development process should continue with Stages 3 to 7 (further data collection, refinement, pretesting, piloting, and implementation).

Unit 20. Writing a Full Ethnographic Report on Women's Health


To write a report which describes the results of data-gathering operations, so that the information can be used for program development, future research, and information dissemination to a wider audience. A suggested format is given below. It should be modified according to the audience and program needs.


1. Title. A concise title which describes the contents of the report.

2. Abstract. The abstract should be no more than 250 words and should address the following points. Alternatively, a three to five page Executive Summary may be included in the front of the report. The abstract or executive summary should include:

a. Main objectives of the study

b. Description of the research area and population

c. Methods used

d. Summary of results

e. Major conclusions

3. Goals and Objectives.

These should be no more than five to six total.

4. Introduction. The introduction should state the specific topics of investigation and perhaps provide some background information. You may wish to summarize the methods employed and why they were chosen.

5. Study Site. Describe the study site and population fully. Explain how the site was selected and how it compares with the surrounding region and to what extent the results are generalizable. The following should be included:

a. Economic characteristics of the population, main occupations, availability of cash, etc.

b. Socio-cultural characteristics. Ethnic or religious groups in the population, educational levels.

c. Description of available health facilities (perhaps include a map) and transportation.

d. If available, perhaps include a short description of the main health problems (from a biomedical perspective) of women in this area (not more than a paragraph).

6. Methodology. Depending on the audience, you may wish to put this section at the end of the report. The section should include:

a. How respondents and informants were selected; sample size.

b. Characteristics and training of researchers and data collectors.

c. Techniques and instruments used, and any problems that were encountered.

d. Limitations of methods and of the data.

e. Human subjects clearance; process of informed consent used.

7. Results. This section will be the largest of the report and should include:

a. It is recommended that your results contain at least three main sections: the Context of Women's Health, Women's Illnesses, and Women's Health Seeking Behaviors. Additional sections may be created as deemed necessary (such as Health Providers, etc.). Within each of these main areas, subsections should be created dealing with more specific topics. For instance, within Women's Illnesses may be subsections on Types of Illnesses, Perception of Severity, Perceived Causes, Perceived Progression, Preferred Treatments, etc. Summary paragraphs, figures, and tables for each of the topical subsections considered should be developed.

b. Diagrams and cognitive maps often convey information more clearly than lengthy explanations. In particular, present local ethnomedical model(s) of women's health problems. For each illness term, discuss the associated signs/symptoms, severity, causes, preferred home and other treatments, progressions, and biomedical equivalent. This can also be summarized in a table, but should be discussed in detail.

c. Present the data from each of the units (however, do not overload the body of the report with multiple tables from each of the units). Complex tables may be placed in an appendix at the end of the report and referred to by number in the body of the results section.

d. Present case studies. (There should always be some good illustrations) (cases) in key informant interviews (remember to use pseudonyms.) Actual quotations from field notes can make an idea come to life.

e. Combine the results from data sources that touch on the same issues. For example, in a section on "Choosing health services," your writing will include data from structured key informants, responses to structured interviews, plus your direct observations.

8. Describe the Intervention Strategy Developed.

This section should include information on the process used to develop an intervention strategy, how it was implemented, what kinds of information were most useful for developing the interventions, how effective it has been, and how interventions were modified over time.

9. Conclusions. Discuss the meaning of these findings for the following:

a. Recommendations for health communications for women (in lieu of section 8 if no interventions have been initiated at the time of preparing the report).

b. Recommendations for health provider education (in lieu of section 8 if no interventions have been initiated at the time of preparing the report).

c. Recommendations for improvement of health services for women (in lieu of section 8 if no interventions have been initiated at the time of preparing the report).

d. Further research needed (specifically discuss how the findings can be used to improve epidemiological studies).

e. What issues need to be addressed by programs immediately?

10. Acknowledgments.

11. Glossary of local terms and their approximate English equivalents.

1 2. References.

13. Appendices with extensive data tables, data collection instruments.


Allocate time and resources for proper distribution of the report and your findings. Distribute the report within the organization as well as to other organizations working in similar communities or on similar topics. Try to publish findings in academic journals, local newspapers, and other forms of media. Perhaps most importantly, find ways to disseminate the results of the research to the community in which it was conducted.

References and Suggested Readings

Ethnographic Methods in General

Bernard HR, et al. 1986. "The construction of primary data in cultural anthropology." Current Anthropology 27(4):382-396.

Bernard HR. 1994. Research methods in anthropology: Qualitative and quantitative approaches. Second edition. Thousand Oaks, CA: Sage Publications.

Chambers R. 1992. "Rural appraisal: Rapid, relaxed and participatory." Discussion paper 311. Brighton: Institute of Development Studies.

Johnson JC. 1990. Selecting ethnographic informants. Qualitative Research Methods, Series 22. Newbury Park, CA: Sage Publications.

Kirk J. Miller M. 1986. Reliability and validity in qualitative research. Qualitative Research Methods, Series 1. Beverly Hills, CA:: Sage Publications.

Pelto G. Pelto P. 1978. Anthropological research. New York: Cambridge University Press.

Stanton B. Clemens J. 1987. "Twenty-four-hour recall, knowledge-attitude-practice questionnaires, and direct observations of sanitary practices: a comparative study." Bull WHO 63(2):217-222.

Stone L, Campbell JC. 1984. "The use and misuse of surveys in international development: An experiment from Nepal" Humn Org 43(1):27-37.

Direct Observation

Bentley ME, Boot MT, Gittelsohn J. Stallings RY. 1994. "The Use of structured observations in the study of health behavior." Occasional paper no. 27. The Hague: IRC International Water and Sanitation Centre.

Gittelsohn J. 1991. "Opening the box: intrahousehold food distribution in Rural Nepal. Soc Sci Med 33(10):1141-1154.

Gross D. 1984. "Time allocation: A tool for the study of cultural behavior." Ann Rev Anthropol 13:519558.

Mulder MB, Caro TM. 1985. "The use of quantitative observational techniques in anthropology" Curr Anthropol 26(3):323-335.

In-Depth Interviewing

Scrimshaw SCM, Hurtado E. 1987. Rapid assessment procedures for nutrition and primary health care: Anthropological approaches to improving programme effectiveness. Los Angeles: UCLA Latin American Center.

Spradley JP. 1979. The Ethnographic Interview. Chicago: Holt, Rinehart and Winston, Inc.

Focus Groups

Carey MA. 1994. The group effect in focus groups: Planning, implementing and interpreting focus group research. In: Morse JM, ed. Critical issues in qualitative research methods. Thousand Oaks: Sage Publications.

Morgan DL. 1988. Focus groups as qualitative research. Qualitative Research Methods series, no. 1 G. Beverly Hills: Sage Publications.

Systematic Data Collection

Weller SC, Romney AK. 1988. Systematic data collection. Qualitative Research Methods series no. 10. Newbury Park, CA: Sage Publications.

Werner O. Schoepfle GM. 1987. Systematic fieldwork. Two volumes. Beverly Hills: Sage Publications.

Qualitative Data Management and Analysis

Borgatti S. 1992. ANTHROPAC 4.0. Columbia, SC: Analytic Technologies.

Fielding NG, Lee RM. 1991. Using computers in qualitative research. Newbury Park, CA: Sage Publications.

Huberman AM, Miles MB. 1994. Data management and analysis methods. In: Denzin NK, Lincoln YS, eds. Handbook of qualitative research. Newbury Park: Sage Publications, pp. 428-444.

Miles MB, Huberman AM. 1994. Qualitative data analysis: An expanded sourcebook. Second edition. Newbury Park, CA: Sage Publications.

Application of Information to Programs

Atkin CK, Freimuth V. 1991. Formative evaluation research in campaign design. Chapter 6. In: Rice RE, Atkin CK, eds. Public communication campaigns. Second edition. Newbury Park: Sage Publications

Coreil J. Augustin A, Holt E, Halsey NA. 1989. "Use of ethnographic research for instrument development in a case-control study of immunization use in Haiti" Int J Epidemiol 8(4)(Suppl 2);s33-s37.

Gittelsohn J. Pelto PJ. 1995. "Suggestions for the appropriate use of qualitative research for developing health intervention programs" (in preparation).

Gittelsohn J. Harris S. Burris K, Kakegamic L, Landman LT, Sharma A, Wolever T. Logan A, Barnie A, Zinman B. 1995 "Use of ethnographic methods for applied research on diabetes among Ojibway-Cree Indians in northern Ontario." Health Edu Quart, (1995, in press).

Griffiths M, et al. 1988. Improving young child feeding during diarrhea. A guide for investigators and program managers. Washington, D.C.: The Weaning Project, Manoff International, Inc.

Herman E, Bentley ME. 1993. Rapid assessment procedures (RAP): To improve the household management of diarrhea. Boston: International Nutrition Foundation for Developing Countries.

Nichter M. 1985. "Drink boiled water: A cultural analysis of a health education message." Soc Sci Med 21(6):667-669.

Nichter M. 1987. "The ethnophysiology and folk dietetics of pregnancy: A case study from South India." Humn Org 42(3):235-246.

Nichter M, Nichter M. 1989. Education by appropriate analogy. In: Anthropology and International Health: South Asian Case Studies. Boston: Kluwer Academic Publishers.

Smitasiri S. Attig GA, Valyasevi A, Dhanamitta S. Tontisirin K. 1993. Social marketing vitamin A rich foods in Thailand: A model nutrition communication for behavior change process. Bangkok: UNICEF.

Stanton B. Clemens J. 1987. "An educational intervention for altering water-sanitation behaviors to reduce childhood diarrhea in urban Bangladesh: formulation, preparation and delivery of the educational intervention." Soc Sci Med 24(3):275283.

Contents - Previous - Next