Contents - Previous - Next
II. Objectives of the study of women's health
III. Guidelines for using the protocol
IV. Introduction to the research methods used in the protocol
V. Format of individual units
VI. Setting up
VII. Site selection and representativeness
IX. Limitations of the data
X. Materials required
This document contains guidelines and procedures for carrying out a community-based ethnographic study of women's health. The main goal of the protocol is to provide techniques for applied social research to facilitate program development and implementation in organizations working in women's health. The results of the ethnographic study are intended to assist in:
1. Identification of the main health problems of women as they perceive them.
2. Developing recommendations for appropriate communication with women on health topics and the most appropriate health education messages;
3. Selecting appropriate forms of local terminology for health care workers to communicate effectively with women;
4. Identifying constraints to improving women's health and suggesting locally appropriate strategies to deal with these constraints.
5. Preparing an ethnographic report that will be useful for the dissemination study findings.
6. Improving surveys that examine patterns of morbidity and mortality in women by suggesting ways of adapting questions to consider/acknowledge community perceptions and practices.
The protocol is a tool for developing and improving the capacity of an organization to conduct applied social science research. The first part of the protocol is centered around an intensive training period in ethnographic methods, during which preliminary data on women's health are collected. The main body of the protocol focuses on a series of data collection exercises with which to generate a sizable body of data on local perceptions and practices. The final section of the protocol presents several ways to apply the data to programs and future research activities.
This protocol differs from other rapid ethnographic manuals is several important ways:
1. It addresses a wide range of women's health problems, rather than focusing on a specific disease (e.g., AIDS) or cluster of highly related illnesses (e.g., diarrhea).
2. Emphasis is placed on appropriate training of data collectors. A module is provided for the training of data collectors during the first 10-14 days of the project.
3. Primary unstructured data in the form of text is used to a greater extent than in other ethnographic methods.
4. Specialized computer software packages (e.g., DtSearch and Anthropac) may be used with the methods in this manual to generate data collection forms, manage, and analyze data.
5. Data analysis procedures are included within each unit and also across units.
6. The last section of the manual shows how the different research components can be used in program development and implementation.
The objectives of the ethnographic study of women's health are:
1. To develop recommendations for appropriate communication with women, and to enhance development of effective health care advice through description of ethnomedical model(s) of women's health problems, including:
i) Identifying the local belief systems and terminology by which women label and interpret their health problems.
ii) Identifying symptoms that women focus on when evaluating their own health problems.
iii) Identifying what women perceive to cause their health problems, directly or indirectly.
iv) Identifying the symptoms and associated beliefs concerning their seriousness that lead women to seek treatment from health centers and/or other medical practitioners (triggering symptoms). This information will permit selection of the appropriate language for encouraging women to seek health care when necessary.
v) Identifying women's knowledge and beliefs regarding the use of home remedies and other home management of their health problems. This information will help avoid "superfluous" home care advice.
2. To describe the various paths through which women seek health care, how their health care-seeking is related to their perceptions of their health problems, and any constraints that may prevent or delay their doing so.
i) Identifying economic, political, geographic, social, and cultural barriers to women's access to appropriate health care.
ii) Provide information for managers and health workers regarding the cultural conditions that may influence community responses to program activities.
The objectives of the study are to be met through the completion of four interrelated research activities:
1. Compile lists of terms that women use to label their own and other women's health problems, and to describe signs and symptoms of illnesses.
2. Construct "ethnographic models" or "cultural maps"1 of women's illnesses from the perspective of the cultural belief system(s) in the specific study region.
3. Analysis of the signs and symptoms of illness, and other factors that women take into account when deciding to seek or not seek treatment.
4. Review women's utilization of health care resources in the study population. Identify the usual and preferred sequences of care-seeking and household management strategies.
1 Key Definition
Explanatory Model: Cultural knowledge, beliefs, and attitudes with respect to a particular illness or other aspect of health. The "explanatory model" for an illness includes the presumed causes, processes, preferred treatments, and expected outcomes.
The protocol is designed to be used by any organization (government agency, academic institution, or NGO) that is planning to develop an intervention or epidemiological study for women's health. It is assumed that the users of this document will be generally familiar with the cultural setting, including the local language(s), in which the research is conducted.
The protocol is organized as a series of units in three parts: Part I, training exercises; Part II, data collection; Part III, applying data to programs. This system is somewhat flexible. Programs with limited financial and personnel resources may only wish to undertake selected sections of the protocol. Programs with greater resources may wish to go further. In any case, Part I should be undertaken before Part 11.
Within Parts I and II, exercises are organized under three main topics:
1) Exploring the Context of Women's Health by understanding women's status within the society, their control over resources, and the priority given to their health problems.
2) The Domain2 of Women's Illnesses will be covered by listening to women talk about their health problems in order to construct "explanatory models" or "cultural maps." These domains include how women label and interpret their health problems, the specific signs and symptoms that women focus on, and the perceptions of severity.
3) Exploring Women's Health-Seeking Behaviors (HSB) will identify the health care resources that women do or do not utilize and their usual and preferred sequences of care-seeking. It will also consider the criteria by which women decide to seek treatment, and the role of home remedies in women's health care.
2 Key Definition
Cultural Domain: A group of words or concepts, that are perceived as "going together" from a particular culture's perspective. Examples of such domains include "illnesses," "healers," "treatments for disease x," etc. (Spradley, 1979).
Qualitative data management and analysis is an ongoing process, which is integrated into the first two Parts. Part III shows how data collected in Parts I and II can be applied to the design, implementation and evaluation of women's health intervention programs (See Figure 1 for a typical time-line).
A short description of the three parts is presented below:
Training Exercises (Part I)
Part I (Units 1-8) presents a training program on the use of this protocol. Suggested content and format of the training is presented in Appendices A and B. The training exercises include a series or basic activities that can be performed in approximately two weeks of fieldwork by two to three data collectors. The primary emphasis explores the context of women's health and the description of women's illness domains. Units 1-3 address the context of women's health, Units 4-5 investigate the domain of women's illnesses, and Units 6-7 chart HSB. Unit 8 guides the researcher in drawing preliminary conclusions and identifying specific areas for further investigation in Part II (Units 9-17).
The data collected in this short period will highlight key areas of concern. To generate programmatic/ policy decisions regarding women's health, however, the additional strategies outlined below are needed.
Data Collection (Part II)
The second set of activities will require approximately four to six weeks of fieldwork by data collectors. These exercises, which represent the basis of ethnographic research on women's health, cover three key areas: the context of women's health (Units 9-11), the domain of women's illnesses (Units 12-15), and women's HSB (Units 16-17). The exercises in this phase serve to "round out" Part I's exploration of the context of women's health. The domain of women's illnesses, for example, is investigated in much more detail. Emphasis is placed on what women actually do when they become ill. In contrast to Part I, the data collected here are a sufficient basis on which to formulate programmatic decisions. Some of the Units (1 1,15,16,18) labeled "Advanced" may require considerable time or are focused on specialized topics. If possible, all of the activities should be completed, but certain portions may be scheduled for a future time.
Applying Data to Programs (Part III)
This part of the protocol describes how the results of data collection can be applied to the design and implementation of effective women's health programs. Part III focuses specifically on developing health education programs, improving health services, and writing a full ethnographic report (Units 18-20). This section is designed to assist the researchers in the process of both synthesis and application in drawing conclusions from the many different data sources.
Typical Time-Line for Using the Protocol
TRAINING - DIDACTIC
TRAINING- FIELDWORK EXERCISES
MAIN DATA- GATHERING ACTIVITIES
Ethnographic research attempts to understand human beliefs and behavior from the point of view of the people being studied (this is known as taking the emic or "insider" perspective). Researchers usually have a different point of view, (known as the etic or "outsider" perspective). This protocol is designed to collect emic data systematically, in a way that has etic significance. Ethnographic research goes further than merely describing people's behavior, by seeking to understand why people do what they do. Through understanding the reasons for people's behaviors (including the obstacles people encounter in seeking health care), planners can adapt programs to cultural conditions.
Ethnographic research is characterized by a flexible and iterative approach. Hypotheses are continually developed, tested, and redefined through a series of feedback loops during data collection. Unlike survey research, analysis is ongoing during the process of data collection. This allows the researcher to identify areas that that need further investigation while still in the field. False leads or unproductive areas of research can be eliminated. Methods are usually adapted as needed and new methods may be developed to address specific questions.
This protocol uses many different methods to address the same three topic areas. Using several techniques to investigate the same question is a form of triangulation³, which helps ensure the validity of qualitative data.
3 Key Definition
Triangulation: The use of multiple methods to investigate the same research question. Triangulation looks for "convergent validity," in which several methods give the same results. For example, in biomedicine, anthropometric measurements, analysis of blood samples, and clinical examination might all be used to determine that a person is malnourished.
Within the first two Parts of this protocol, each Unit contains the following four sub-headings: Purpose, Data Collection, Data Analysis, and Interpretation Suggestions. Examples of completed data collection and data analysis forms are provided with many of the units. Blank samples of forms are provided at the end of the book. These blank forms can be photocopied.
The statement of purpose briefly presents the goals of the exercises described in the unit, how they relate to other exercises in the protocol, and the significance of the unit for the organization's programmatic efforts.
The data collection subsection of each unit usually contains the following elements:
1. Instructions on how to prepare data collection materials.
2. How to locate and select informants.
3. Suggested sample size (e.g., 10 - 15 respondents).
4. How to present questions to informants.
5. How to record responses on data collection forms.
6. Samples of completed data collection forms.
The section on data analysis may contain some or all of the following types of information:
1. How to expand the field notes (i.e., suggestions on what to include, degree of detail, etc.).
2. Suggested codes4 for textual data.
3. Methods of data presentation, including how to tabulate data on specific forms.
4. Samples of completed data analysis forms.
4 Key Definition
Codes and coding: Codes are short or abbreviated words (or numbers) that label important topics/ concepts that you wish to mark for easy identification (for later retrieval and analysis) in field notes and interviews. For example, a relevant topic is reasons for delay in seeking treatment. A code for this might be DELAY.TRT.
Note that the data collected in each unit tend to be rich and detailed. Only a few of many possible options for data analysis are discussed in this section of each unit.
presents suggestions on how to interpret results of the data
analyses and how to apply these interpretations to programs. In
many cases, this section presents questions the investigator can
use to guide his/her reading of the data. When researchers
interpret qualitative data about a study population's beliefs,
perceptions, and behaviors, they of en implicitly compare the
results with their own or other "outsiders" perceptions
about the cultural beliefs and practices in the community under
The typical study, for example, one that is connected with an existing health project in a community, would require the following personnel: 1 Project Manager, I Trainer, 2-3 Research Assistants/Data Collectors. Ideally, all staff should be familiar with the local area and be fluent in both the local language and English (since the protocol is currently available only in English). Some attention should be given to the age and gender of the staff on the project. In many societies, it is inappropriate for men to discuss health related topics with women. Older married women, in addition, may be reluctant to discuss their problems with young unmarried women. The staff composition should balance these characteristics as much as possible. If the project manager were an older man, for example, the research assistants should include at least one or two married women.
The Project Manager (PM) is responsible for administering the study and will, ideally, use the results of the study to develop future survey instruments and interventions. The project manager supervises all aspects of the research, analyzes and integrates the data, and writes the final report. The PM is responsible for incorporating the results into future program activities. It is preferable if the PM is familiar with the se of microcomputers. (Alternatively, a computer-proficient individual might work on the project, as needed, to develop the computer database.) The PM should give 100% of his/her time during the training period and a minimum of 25% during the rest of the project. The PM will develop the ethnographic field guides, supervise the research assistants, review the data collected, and identify which topics require further investigation. The PM also makes all decisions regarding modification of the methods and units based on the data collected. The PM should participate in data collection activities during the training period. During the second phase, the PM might conduct key informant interviews and be involved in other data collection activities, as time permits. It is important to note that the more time the PM can spend both collecting and reviewing the data, the more closely the research can be adapted to local beliefs and practices.
The Trainer should have a social science background and experience in using ethnographic methods and computer applications. Ideally, the trainer should have experience in using this protocol and some knowledge of the local cultural context.5 The trainer must work on the project for 100% of the time during the training period. After that, it is anticipated that the trainer would no longer be involved in the project, except perhaps as an informal consultant. The responsibilities of the trainer include an orientation to the protocol and conducting the didactic training sessions on ethnographic concepts and methods. The trainer would also provide training in the field by participating in the data collection procedures discussed in units 1-7. In Unit 8, the trainer would work with the project staff to develop a coding scheme and identify areas that need further research.
5 For suggestions and recommendations for possible trainers, please write or e-mail:
Dr. Joel Gittelsohn, Dept. of International Health, Johns Hopkins University, School of Hygiene and Public Health, 615 N. Wolfe St., Baltimore, MD 21205, USA. JGITTEL@SPH.JHU.EDU
There should be 2-5 Research Assistants (RAs). Each RA should be fluent in the local language and comfortable writing English. They should have some research experience and should not be viewed by the community as service providers (i.e., the RAs should not be involved in any intervention in the study site). The most important qualifications of the RAs are their ability to develop rapport with informants and to probe effectively for more detailed information. If the computer is used, the RAs should have some experience with computers, especially word processing. RAs should devote 100% of their time for the duration of the project. The RAs are responsible for preparing the data collection materials (photocopying forms, preparing and maintaining files, making cards for pile sorts, etc.), and conducting most of the data collection activities.
This would involve identifying key informants, conducting in-depth interviews, observation, pile sorts, and paired comparisons. They may also enter the data into the computer and do some tabulations.
There may be a need for a Typist/Data Enterer. This person should be proficient in a word processing program if the notes will be entered into a microcomputer and should be fluent in both the local language and English. This person would be responsible for typing expanded field notes or entering them into a microcomputer and filing the notes according to a filing scheme. This person must be carefully trained to type the notes exactly as written without translating or changing terms.
In cases in
which at least a portion of the local population speaks a
different language from the investigators, every effort should be
made to employ research assistants from the local population as
they may also be asked to act as translators for the
investigators. Translators/interpreters must also be carefully
trained, so they do not simplify, "gloss over," or
otherwise "edit" the statements of community people.
The translator must translate phrases literally. Specific terms
and phrases for illnesses, signs and symptoms, and progressions
should be recorded in the local language.
Guidelines for Choosing a Study Area (s)
1. The study area should represent the target population. The study may also focus on particularly vulnerable groups in the target area for example, those who are poor, low caste, more isolated, and the like.
2. It may be advisable to include more than one area in order to compare different groups, for example, two villages, slum areas, neighborhoods, or communities. For instance, the "catchment area" of a health center may include more than one group or community.
3. The location should be logistically feasible (e.g., closer, on main road), so long as it does not compromise representativeness, by being substantially different (e.g., poorer) than the area further from the road.
Confirmation in other Sites
Many programs offer health services to women in multiple settings (rural and urban, different geographical locales, etc.). In order to implement programmatic activities, organizations may need to collect ethnographic data in several locations. The following guidelines are recommended:
1. Repeat several data-gathering activities in the second or third location. Often, it is not possible and may be unnecessary to repeat the whole instrument on a second population. In most cases the number of data-gathering units utilized and the total number of times each unit is performed can be reduced.
2. Since the purpose of this exercise is confirmation and not new exploration, the second study population (i.e., confirmation site) should be of the same ethnic/cultural background as the first study population.
3. If the group wishes to expand the research to cover a new ethnic group, it may be necessary to repeat all or most of the original data-gathering activities.
Unlike quantitative studies, which emphasize the use of random or systematic samples, the methods described here employ primarily purposive sampling. Random sampling involves selecting units from a sampling frame that uses randomly computer generated numbers and allows one to generalize the results to the population from which the sample is drawn with measurable statistical precision. The main purpose of a random sample is to ensure that the sample is representative of the total population. Purposive sampling involves selecting individuals based on their level of cultural knowledge of a particular research topic or area. Although purposive sampling does not allow generalization with a known precision, every attempt should be made to ensure adequate coverage of important sub-populations (ethnic groups, poorer groups, etc.). For example, one might sample four to five respondents from different economic or ethnic groups.
The methods in this protocol involve two types of informants. Key informants are purposively selected and are interviewed several times (at least 3-5 times) in order to provide in-depth information about particular topics. Respondents may be either purposively or randomly selected and are usually interviewed only 1 or 2 times. Their interviews are intended to ensure coverage of information. The two types of informants are intended to provide a balance between coverage and depth of information collected about women's health.
How do we locate key informants and respondents? Often the selection of key informants is based on the recommendation of knowledgeable community members. Commonly, community leaders are first asked to recommend persons who would be "good to talk to ..." about the topic of interest. These initial referrals usually lead to secondary and even tertiary referrals. A key informant is a person with whom the interviewer builds a social relationship of communication over time. Every referred person is a potential key informant, and it should be clear after 1 or 2 interviews whether this person will serve as a true key informant.
Part of the
decision about random or purposive selection of respondents
depends on the availability of current lists of
persons/households in the study community. If a reasonably
current community list can be obtained, random sampling of
respondents can be used to strengthen the representativeness of
findings. Many of the groups using this protocol will have access
to household and/or individual lists based on health records.
When a community list is not available, a systematic sampling
framework sample may be used, such as mapping the study community
and interviewing persons in every fifth household.
This protocol is intended for use by health professionals and social scientists. It is designed to be carried out as a series of tasks which address program specific questions relating to women's health.
1. It is not intended to be a thorough investigation of all cultural-behavioral aspects of women's health in a community or region.
2. It will help to identify the characteristics of women who are likely to be at higher risk for illnesses as well as those who are less likely to seek treatment. However, it will not provide data on the incidence or prevalence of the illnesses or beliefs under study, or the proportion of women who fall into a given category.
3. It will identify salient characteristics of illness recognition and the steps in deciding to seek treatment but will not develop predictive models of behavior and decision making processes.
The following materials are required for different phases of the work:
1. Didactic Training:
- Overhead projector and transparencies (optional)
- Chalkboard and chalk
- Copies of data collection forms
- Tape recorders and tapes
2. Data Collection:
- Small tape recorder, tapes, and batteries
- Access to photocopier
- File folders
- Small field notebooks
- Composition books for expanded field notes
- Copies of data collection forms
- File cabinet
3. Management and Analysis:
- Typewriter and paper
- Computer (optional)/ZyIndex or DtSearch software/Anthropac software/(see Appendix C-4 for ordering information)
- Word processing software
- Hand held calculator
Contents - Previous - Next