Contents - Previous - Next

Exploring women's health-seeking behaviors

Unit 6. Key Informant Interviews with Women: Health Concepts and Health-Seeking Behaviors


The purpose of these key informant interviews is to obtain detailed information about women's health-seeking behavior from local women. These data will provide important contextual information for use in more structured assessments of women's health-seeking behavior. The information gathered in Units 1 to 5 should be used to construct an Ethnographic Interview Guide using terminology and questions that make sense to the Informants.

Data Collection

1. Develop an Ethnographic Guide for Interviews on Health-Seeking Behavior.

A sample Guide follows (Form G. 1), and should be modified for use in the research. The questions provided are examples of some that you might use in conversation with a woman about her health-seeking behavior, but should not be followed rigidly. It will be impossible to get though all of these topics in 30 to 45 minutes. If the informant is knowledgeable, several interview sessions may be useful to explore other issues raised in this field guide.

It is not only important to learn what health problems women face, but what resources they have to deal with those problems, both external (such as clinics) and internal (such as home remedies, sympathetic help from husbands, and personal strengths).

2. Translate Questions into the Appropriate focal Language. Use data from Units 2 to 4 to help identify appropriate local terms.

3. Select Informants. At least five purposively-selected women key informants should be interviewed as part of this exercise.

Data Analysis

Microcomputer Application (optional)

Key informant interviews can be managed and analyzed with ZylNDEX and/or DtSearch. A detailed method for conducting a computer search using these programs is presented in Appendix C-3.

1. Analyze data on the developmental stages of women. On a piece of paper, draw a time line of an average woman's life cycle, according to local perceptions. The time line may or may not have a scale measured in years. This picture should be a kind of diagram that represents a synthesis of responses from all your key informants from all the different interviews you conducted. It could look like this:


Naming Ceremony

First Menstruation


Birth of First Child

Birth of First Grand Child


2. With the developmental cycle of a woman (in local terms) constituting the x-axis of a scale, you may wish to plot other kinds of information collected during the interview (i.e., activities performed, kinds of foods eaten, perceived vulnerability to illness, etc.). From the interviews you may get a sense of when women are seen as having the most health problems. The same technique could be used to look at the developmental cycle of a woman's pregnancy experience.

3. Results of the interview may be tabulated. Form 6.2 gives an example of a type of tabulation that could be developed using information provided by informants on different types of health practitioners they use for different illnesses.

Sample Ethnographic Interview Guide: Health-Seeking Behavior

Insert the following questions into Form T1.1:

1. General issues related to health problems of women in the community

"I would like to learn from you about health problems of women in this community, and what women do about them. I was wondering if you could share with me some of your insights about this community and from your personal experience."

"What are the most common problems that women face in this community? What are the most serious problems that women face in this community? Have you or your family faced these problems?"

"How do you know if somebody is not doing well?

"How does one know a healthy woman? A sick woman?

"What are ways that women cope with problem X ?"

2. Questions about this woman's situation

Exploring women's health by developmental stage.

"Please tell me about your life when you were younger.

What was life like for you as a girl? What were some of the good things you remember? What were some of the problems you remember? Were you ever ill? Could you tell me about this illness and what effects it had for you?

Tell me about your monthly cycle (find appropriate local term). What happened when it first started? How is it now? What do you do nowadays when you get your monthly cycle? What do you not do when you get your cycle?

How many times have you been pregnant? How many births have you had? How many children do you have now? How have these pregnancies affected your health?

What was life like for you with your first pregnancy? Tell me what you understand about the process of pregnancy. Where does the baby collie from? How did you know you were first pregnant? Where does the baby grow in your body? How does it grow? What did it feel like for you to be pregnant? What did you think? What did you nor think about? What did you wish you knew then. that you know now? What are questions that you still have about pregnancy?

Looking back at your life, you mentioned that you had (list the various problems, possible problems). Which of these do you consider illnesses? Which are normal things?"

3. Health and Health-Seeking

"How is your life going right now? How would you say your health is? What things have made you healthy? What things have made you weak/sick?

Tell me about other illnesses you have had. (Try to get a full "story" of one or more "illness episodes."

What were the first symptoms of illness X? What caused those symptoms? What did you think when you noticed those symptoms? What did you do? What kinds of foods did you eat to become more healthy? Why do you feel those kinds of foods were good for your health?

"Where do you go or whom do you consult when you are ill?"

Keep probing, following the sequence of events that the woman noticed, and her thoughts and actions at each step. This will give you a global view of her perceptions of each illness, the health resources that are available to women in the community, costs, and distances to these health providers.

4. Financial Considerations

Depending on the research questions, it may or may not be appropriate to find out about how healers are paid. Questions of motivation for going to different types of practitioners can be explored at the same time as exploring questions of payment.

5. Terminology

What are the words or ways to talk about being shy or embarrassed? Who do women talk to about things they are shy about? What are the different situations or things women are shy about?

6. Women 's Status/Power

Ask for narrative accounts of what happens during meals.

Who eats first? Get a "play by play" description of who is served, how much, etc.

Who makes decisions about cash? How are household jobs divided up (may be important to determine status and authority among women)?

What ways are men and women treated differently in your household? Who is the most important person in your household?


How do others treat that person?

What responsibilities does each person have to the household? What benefits does that person gain from the household?

Who is the second most important person?

How do they address that person?

What do they ask that person to do for them?

What benefit does that person get from being in the household? What happened the last time your child was sick?

Tabulation of Practitioners who Serve Women (sample)

Type of Practitioners

Illness Term: Safed Paani

Practitioner Name



No. Informants

1. Dr. Ray


1 hour away


2. Dr. Malla


1 hour away


3. Madame Sen


in village


4. Good Health Clinic

Private clinic

next village


5. Government clinic

Government clinic

next village


6. Mr. Pandit










(Add lines and paper as needed)

Interpretation Suggestions

1. How does the information on local perceptions of women's developmental cycle differ from the ideas in your organization? What features/stages do they stress as most important?

2. Does it appear that some household members are more important than others? How so?

3. Who are the least important and the most important persons in the household? How do the others in the household treat the "most important person? How do the others in the household treat the "least important person?"

Unit 7. Paired Comparisons: Women's Illnesses and Choice of Practitioner


This task is designed to yield information on locally preferred choices of health care for women's illnesses. Since the task is presented to a woman in the form of questions about what she believes she would do, rather than what she actually did, the data that result from this task are not measures of behavior per se, but of ideal or normative behavior.

Data Collection:

1. Constructing the Task Materials

Microcomputer Application (optional)
The paired comparison data collection forms can be constructed and printed using ANTHROPAC

a. Identify a set of 3 to 4 women's health problems. These may be the top health problems mentioned by women during the free listing.

b. From a list of alternative sources of health care (see Units 2 and 6) select up to six specific choices. If there are more than six types, try to reduce the number to six. For example, the list may include: a government health center or clinic, a private physician, a community health worker, one or two traditional healers, a pharmacist, and perhaps a non-qualified private medical practitioner. Try to include a range of different types of health providers.

Results are likely to be more valid if the actual names of individuals or clinics are used. For example, the set of alternative health care choices might be: Dr. Ray (a private physician), Dr. Malla (a non-licensed, non-qualified practitioner), Madame Sen (a traditional healer), the Good Health Clinic (private homeopathic clinic), the Government Health

Center, and Mr. Pandit (a pharmacist). When choosing one provider over another for inclusion, preferably select those known to treat women's illnesses. For instance, the local traditional bone setter would be an unlikely choice as one of the six. In urban/periurban areas where there are many providers, choose those more commonly mentioned.

c. In the task, respondents will be asked to choose between alternative resources, presented two at a time. This procedure provides a fuller and more accurate picture of people's evaluations of alternative choices than asking them simply who they would go to if a woman had a particular illness.

The task would be very tedious and lengthy if every respondent was asked about all combinations for all illnesses. Therefore, each respondent should be presented with all combinations of practitioner choices for only one of the illness terms.

d. Prepare sets of work sheets (lists of all possible pairs among the six health care alternatives). For each illness, each individual work sheet should present the pairs (of health care choices) in a different, randomized order. The following chart illustrates the combinations of all possible paired comparisons for six different providers:

Example of Randomized Sets of Paired Comparisons

Health-Care Options

All Possible Pairs


Dr. Ray

A and B

B and C

C and E


Dr. Malla

A and C

B and D

C and F


Madame Sen

A and D

D and E

D and E


Good Health Clinic

A and E

B and F

D and F


Government Clinic

A and F

C and D

E and F


Mr. Pandit

2. Selecting Respondents

Conduct paired comparisons on at least five respondents per illness term being investigated. If there are four illnesses and five respondents for each illness, up to 20 women would be required. This number should be adequate to identify a pattern, if there is a high level of cultural consensus. However, if one (or more) of the illnesses elicits very different choices from different women, it will be important to interview additional respondents regarding that item.

3. Preparing the Forms (Randomization)

Each time the Paired Comparisons task is performed, a new order of presentation of every pair should be drawn up. This seems to reduce bias in responses. Fill out the first two columns ahead of time and arrange the order of the respondent pairs differently on each form. (Alternatively, 3 to 4 different orders could be developed and then the form could be photocopied.)

4. Suggested Wording

a. "In order to help me understand more about what families can do (what kind of help they can get) when women become sick I'd like to ask you some more questions."

b. "One of the illnesses I'm trying to learn more about is (illness name).

c. "In our discussions with people here in (community name), we have found a number of places families can go to get help when someone is ill. For example, the Government Clinic, Doctor Ray, and Mr. Pandit the pharmacist. Give the names of all six health providers selected for the exercise.

d. "If a woman thinks she might have (illness name), there are several places she could go for help. But suppose that she could only go to one or the other of these two places." "If she could only go to Dr. Ray or to Mr. Pandit, which one should she choose?" [Following the order on the work sheet, present the first two providers. Circle the response on Form 7.1 ]

e. [Then ask why she made that choice.] "Why do you feel that she should go to _____ in stead of ______?" [Write down the response in the section entitled, "Explanation for Why Practitioner was Selected."

"Now, suppose that the choice was the Government Clinic or Madame Sen. Which would you recommend?" Again, ask "Why?" after she makes the choice. Continue this process for each pair.

Use Form 7.1 to record answers from each respondent.

Data Analysis

Microcomputer Application (optional)

Paired Comparisons data can be entered and analyzed using the ANTHROPAC software.

1. Separate the forms by illness type/term.

a. Add up the number of "votes" that an individual respondent gives for each practitioner choice on Form 7.2. Since each provider appears in five pairs (if a total of six practitioners was used) the "winner" (most likely choice) from an individual's perspective should be chosen five times, the next choice gets four "votes," and so on. If the informant is consistent in his/her choices, a possible pattern for one individual looks like this:

Results of Paired Comparisons: One Respondent

Dr. Ray

5 votes

Mr. Pandit

4 votes

Government Clinic

3 votes

Good Health Clinic

2 votes

Madame Sen

1 vote

Dr. Malla

0 (never chosen)


15 votes

b. The paired comparison task results in a clean rank-order if the respondent is consistent. [Respondents are not always consistent, of course.] As a result, simply adding all of the votes of all of the respondents should give a clear rank ordering of people's choices of practitioner.

Recording Form for Paired Comparison of Practitioners and Illness Term (sample)



Name/lD No:


Marital Status:


No. of Children:

Years of Education:

Ethnic Background:


Illness term

Directions: As you conduct the interview, circle the respondent's choice of provider and write down the reason given for each pair. Remember: Each worksheet should present the pairs in a different randomized order.

Pairs Of Practitioners

Letter Code

Practitioners Names

Explanation For Why Practitioner Was Selected


Dr. Ray -Dr. Malla



Dr. Ray - Madame Sen

better medicines


Dr. Ray - Government Clinic

no waiting


Dr. Ray - Mr. Pandit

closer to the village


Dr. Malla - Madame Sen

better medicines


Dr. Malla - Good Health Clinic



Dr. Malla - Government Clinic

Government clinic is dirty


Dr. Malla - Mr. Pandit



Madame Sen - Good Health Clinic



Madame Sen - Government Clinic

Government clinic is dirty


Madame Sen - Mr. Pandit

better medicines


Good Health Clinic - Gov't. Clinic

much better doctors


Good Health Clinic - Mr. Pandit



Government Clinic - Mr. Pandit

better medicines


Dr. Ray - Good Health Clinic


Results of Paired Comparisons: For 5 Respondents (Hypothetical Data)








# Total

Dr. Ray







Government Clinic







Madame Sen







Mr. Pandit







Good Health Clinic







Dr. Malla







2. Tabulate the Choice of Practitioner. Use Form 7.2 for this procedure. Write the particular illness being considered for this analysis on the top row of the form. Write each practitioner's name along the top row (next to the appropriate letter A, B. C, ...). Then for each informant, write the number of votes in the appropriate space. When complete for all informants, write the total number of votes at the bottom of Form 7.2.

3. Tabulate reasons for choosing practitioners or not selecting a particular practitioner for a specific illness on Form 7.3

a. Answers to the question "Why?" concerning peoples' stated choices for health care services are generally short statements, such as: "Cheaper," "Her medicines always cure," "They have more medicines," "She was educated in the city," "You don't have to wait all morning," "They speak our language there," and other similar statements.

b. Examine all the responses, and group them into categories. The following categories will cover most of the responses, but it may be useful to create others:

1. Economic reason (e.g., cheaper, easier to pay, "you get your money's worth")

2. Belief in efficacy of modern drugs ("she has the strongest [beer] medicine." "he gives injections)

3. Belief in efficacy of traditional medicines and practices (e.g., "traditional medicines are beer")

4. Convenience, social access (e.g., "you don't have to wait all day," "they speak our language there")

5. Special positive qualities of the practitioner (e.g., "She was educated in the city")

6. Special negative qualities of the practitioner (e.g., "They are very rude to us when we go there") In order to tabulate on Form 7.3 this will have to be rewritten as "Other is very rude to us when we go there."

c. Collate the responses to show the cross tabulation of reasons with practitioners (or practitioner types) and summarize the results in tables and descriptive paragraphs.

Interpretation Suggestions

1. From Form 7.2: The practitioner who received the largest number of votes is most likely to be seen for that particular health problem (according to their reports). Are there some practitioners who really stand out as "preferred treaters" for this illness? Or are the number of votes relatively close? A close vote indicates one of two things: either there is no preferred practitioner for this illness, or there may be particular subgroups in the population that use different practitioners. If you suspect this to be the case, go back to the original data collection forms and see if the types of people who select one practitioner over another differ according to some observable characteristics (economic status, ethnic background, etc.). You may wish to tabulate responses for each sub-group separately.

2. From Form 7.2: Is there disagreement on ranking of providers by informants? For example, see Madame Sen on the previous page. Is there any pattern of those types of informants who prefer one type of practitioner to another?

3. From Form 7.3: What are the reasons that certain practitioners are selected over others for specific illnesses? Is it a matter of experience/skill or use of particular medicines? Often the allopathic practitioners are favored because they possess "hospital medicines" or give injections, while traditional healers are chosen due to their skill and experience in treating certain kinds of illnesses.


Contents - Previous - Next