Contents - Previous - Next
The key-informant methodology is the single most powerful ethnographic data-gathering tool. Although individuals vary considerably in their natural interviewing skills, the techniques of key-informant interviewing can be taught to persons ranging from the highly educated to near-illiterate community people. The most effective training methods are usually a combination of role-play, trial-and-error, and continued practice under supervision. The following summarizes the main elements of key-informant interviewing:
i. The informants must be interviewed more than once, so that a social relationship develops between the interviewer and key-informant.
ii. The key-informant is regarded as an expert, who imparts important information to the interviewer. The interviewer acts the part of someone interested in learning from the informant. (For this reason some anthropologists such as O. Werner and R. Schoepfle suggest the term consultant in place of key-informant.) The interviewer should not respond to information from the key-informant/consultant with value judgments or expressions of criticism.
iii. The interviewer must record as much of the information presented by the informant as possible, usually by writing in a notebook, sometimes supplemented with tape-recording. The act of recording the information is part of the demonstration that the data from the expert are important.
iv. The interviewer seeks to get key portions of the testimony in the informant's exact, or nearly exact words. Words, phrases, and (sometimes) whole sentences are written down exactly, when possible. Sometimes the tape-recorder is used as a backup device, to check on the interviewer's notetaking. In most cases the interview must take place in the informant's native language and the interviewer must also be fluent in the local dialect.
v. The interviewer must avoid too much reliance on asking a series of focused questions. Instead, he or she tries to get the informant to narrate, list and enumerate, and expand his or her explanations of various topics.
vi. The interviewer tries to get the informant to lead the discussion into elaborations, explanations, and sometimes into whole new topics.
vii. If the interviewer has some prepared questions, or lists of topics, these are never used to structure the entire interview. The prepared questions are introduced occasionally, secondarily, for example, when the informant has finished explaining a topic. One exception to this rule occurs when the key-informant is asked to respond to a pretesting of a structured interview that is being developed for use with a wider range of respondents.
viii. Second and subsequent encounters with the informant provide the opportunity to bring up materials from the previous interview-probing for more information. Often the interviewer repeats portions from a previous interview to verify or her understanding and to prompt the informant for more detail.
ix. Probing for more detail, information, examples, or cases, is a major element of successful interviewing technique. This probing, prompting, urging of the informant is a counterpoint to "listening to the narrative explanations."
x. Ideally, the relationships of researchers to their key-informants continue throughout the duration of the project. In later phases of information gathering, it is very useful for the interviewer to try out hypotheses with his or her best key-informants. That is, the researcher has developed a model or general idea of the particular cultural/behavioral domain that is the focus of study. For example, perhaps the research team has a series of ideas about the personality of leafy green vegetables and how they differ from other vegetables. This idea can be brought up with the key-informant, who is asked to comment, correct, and perhaps expand the idea into other food groups. In some cases researchers will ask their key-informants to comment on drafts of reports, or on descriptions written from past interview materials.
xi. In an increasing number of cases, ethnographers seek to have true local participation in the data-collection and interpretation processes. In such cases the key-informants become participating information-gatherers, paid or unpaid.
Probing and Prompting
One of the central skills of interviewing (whether it is key-informants or just one-shot informal discussions with persons you meet in the community IS getting the informant to continue to elaborate, give examples, or suggest new areas of exploration with little interference and minimal structuring from the interviewer. The main technique involves unobtrusive probing. Here are some miscellaneous examples used in the midst of open-ended, conversational interviews:
"That's interesting, please go on."
"Why did people stop using .________________(e.g., a food or a method of food preparation)?"
"What did you do then?"
"Could you explain that a bit more, I didn't get the part about the___________(some element of action or content)."
"That's fascinating...can you think of any other examples?"
"What do you call that method of food preparation?"
"Please tell me about what you ate when you were young and how food and diet have changed since then."
"Can you think of any other dry season foods?"
In many instances you can simply repeat a part of the previous statements of the informant, such as:
"I see. So the people who live closer to the river are the ones who do most of the fishing...."
Simply restating the informant's words often leads to further elaboration and explanation. It cannot be too strongly emphasized that your attitude of interest, expressed in body language and verbal reactions, is often the most effective prompt that leads to further explanation and detail from the informant. Sometimes the most effective prompt is in the form: "This is so interesting that, if its OK with you, I'd like to come back tomorrow. Then we can have more time to go into these details."
Getting Lists of Things (Items) in a Domain
One of the more important techniques in open-ended interviewing is that or asking informants for lists of things. The most mundane, ordinary topics can suddenly come to life when an informant tries to list a series of elements, types, or other things. In addition to gathering lists of foods (as defined in the free list, section IV-B) it will be useful to get lists such as special holiday/celebration foods, types of snacks, foods that are good for you when sick, etc.
Of course we do not usually say, "Please give me a list of __________." More likely, we say "You just mentioned that chicken soup is good to give children when they are sick. Can you tell me what other foods are good to give to sick people, for different sicknesses, or for different kinds of people...?"
Usually, you will have to do some prompting and encouraging: "OK, that's three different types...any others? What about any special foods for the sickness that pregnant women get?"
It is worth repeating that the best, most detailed and systematic information comes from key-informants, after you have established a social relationship in which they come to understand and appreciate the kinds of information you are trying to gather. They become emotionally involved in the process themselves and will often make special efforts to think about the information, and will even seek out more examples and details from their friends.
So, the bottom line is, cultivate long-term communications with your best key-informants.
Fieldnotes are your primary data records from informant interviews as well as from direct observations out in the research community. Here are the different kinds of items that should be written in fieldnotes:
i. Observations of crops and other food sources that one sees while walking about the community.
ii. Lists of the different types of food being sold on a market day, along with notations about prices, quantities, condition of the food, behaviors of people in the market and other details.
iii. Detailed, play-by-play description of unstructured interviews with key-informants.
iv. Observations of the cooking areas seen in peoples' homes and lists of cooking equipment, stored foods, and other details.
v. Step-by-step descriptions of cooking processes as told to you by informants (e.g., preparation and cooking of unusual wild foods, sauces).
vi. Notes on past history, geographical features, ecological data (climate, rainfall, etc.), ethnic and language features, and other background materials as gathered from written sources and from open-ended interviews.
This discussion will cover writing of notes from both informal observations and interviews. Formally structured interviews (survey interviews), on the other hand, are generally recorded directly on the prepared interview forms.
Your field notebook should be small enough to keep in your pocket or purse. The notebook should not be overly conspicuous, but it is often a good idea to make it obvious to community people that you (and your team members) are taking notes. Writing things down from interviews demonstrates to the informants that you are serious about wanting the information and you regard them as experts.
Whenever possible, your information gathering team members should write down notes directly into a notebook when doing open-ended interviewing. In addition, you should have your information gatherers write notes on descriptions of places (e.g., marketplaces, cooking and food-storage areas); sketches or diagrams of a complex food processing device; sketch plus written description of home gardening area with its crops; and other special information. It is also useful to include notes on such things as: "Graciela is a very good key-informant...but it's important to find her early in the morning, or else late at night, as she goes to sell in the marketplace every day...."
Many of the items you write in fieldnotes are the same information your research assistants already know because they are members of the local community. But writing these things down in fieldnotes makes the information available and organized for direct use in planning interventions.
From "Jottings" to Fully Written Notes
Obviously you can't write down everything the informant says. But you can jot down key words and phrases, to keep a running "log" of the ideas and answers as your informant responds to your questions, probes, and encouragements.
The jottings are intended to jog your memory when you expand them into fuller statements, as soon after an interview as possible. Here is an example of jottings from an interview and then the full notes, written up later:
[Interview with elderly woman (excerpt)]
Earlier..."all people collected them (calchan) [wild greens, stalk, leaves] had time. Children. Especially older women. Now. Too far. All summer. Cooked with meat. Poor people potatoes. Lazy. "Young women won't cook 'em..." Edge of fields...river...people. "Weaker now..." "Poor food..."
Expanded notes as written up afterwards:
The old grandmother in the___________ household told me that when she was a child and even as a young married woman, people gathered the wild greens called calchan, which were available throughout the summer. "All people collected them...." The main locations were at the edges of the cornfields and also along the riverbank. They gathered the stalks and leaves. Sometimes they sent the children to gather them, but mainly it was older women who did not have small children to care for. The wealthier families cooked the greens with meat, but most families were poor and cooked them with their potatoes.
When I asked why people do not gather these wild greens nowadays, she said that "people are lazy" and then she added that "...the younger women won't cook 'em." because they view the greens as being inferior food and "poor food" (that is, food for the poor people). She also said that she believes people today are weaker now and less healthy because they do not eat the healthy foods "from nature" like they did when she was young.
In this example we see that the expanded fieldnotes are approximately four times as many words as the originally jottings. Also, the jottings are almost undecipherable. Only a person who heard the original interview could make sense of these jottings.
Some Additional Guidelines
i. Look for key words and special vocabulary. Be sure to write down those key words in the exact language of the speaker. Note especially any words that are special to this population and which identify special roles, (e.g., the name of the food) and special words for actions, behaviors, or ideas that might come up in structured interviews or behavioral messages during the actual intervention, especially words such as "poor food" that reflect strongly held attitudes.
ii. Check with the informant to be sure that the special vocabulary is actually used by the target population, that it is an insider vocabulary. You may find that the older informant uses a word (e.g., calchan, that is not familiar to the younger generation, or, perhaps it is the word used by the people in one ethnic group only.
iii. Particularly important ideas of the key-informants should be written down in their exact words (at least key phrases showing how they expressed the idea).
iv. Expand and write out full fieldnotes from your jottings as soon as possible after an interview. If you do not have the opportunity to write out the full notes immediately, try very hard to get all the notes written out the same day of the interview. Otherwise you may lose much of the content.
Often it is useful to check some part of your jottings during the interview. Sometimes you can say to the informant. "Let me just check this thing I have here in my notes...let's see...you said that ____________"
It is a good idea to go over your notes right after an interview, if you can find a private place. Just go through and add in a few details to the written notes. That will help to make sure that you will remember clearly when you sit down to write out the notes more fully.
As you write out your notes more fully, try to preserve as much as possible of the play-by-play flow of the interview. Do not try to write a polished essay, just go through the specific information-what the informant said-in the order that it occurred in the interview. Often during the interview you will need to change direction to get more data on a point made earlier. Also, your key-informants usually wander off the topic and return to earlier points, filling in earlier gaps. Usually it's best that those later additions be written in the order in which the interview actually happened.
The most preferred way to write out your notes fully is with a microcomputer. If possible, your team members should become familiar with using a word processor system in the microcomputer, so that all the interview notes and other fieldnote materials will be accumulated in computer files. That makes it much easier to go through the files of notes to find particular points of information.
A second alternative is to write the notes out on a typewriter. In any case the fieldnotes should be written out on paper so that they can be stored, sorted, and organized efficiently. If the notes are typed, be sure to make at least one extra copy, and it's better to have two extra copies of all of the materials. Those should be stored in different (secure) places.
If your team members cannot type, they must write out the notes fully by hand, and those handwritten notes should be checked by others to be sure that they can be read easily. In some cases, if you have typists available, it is possible that your information gatherer can dictate out loud from the fieldnote jottings, while the typist writes out the full text.
v. You can include in your notes some of your own interpretation of things that your key-informant said. You should also include comments such as, "At this point the informant seemed to be covering up something and quickly changed the subject. Next time, I should ask her some more about this. This time it seemed to be a touchy and emotional subject."
When you write such personal impressions, you will of course, make sure they are clearly your ideas and not the words or ideas of the informant.
vi. Allow at least two hours of writing time for each hour of interview. That's right-two hours! It is slow work, because often you will be remembering many different things that your key-informant has said, at the same time trying to make sense of the very short and cryptic jottings in your field notebook.
vii. In some cases you may be able to use tape-recordings of some key-informant interviews. Even if you are tape-recording the interviews, you should still take good notes and write out those notes fully, as described here. The tape-recording can serve as a backup source. Remember that it costs a great deal of money and time to transcribe from tapes to typewritten form. Most projects cannot afford to pay for the transcribing.
viii. If you do use a tape-recorder, be sure to transcribe the interview(s) into the computer or write them out on paper as soon as possible. Never leave tape-recorded materials in unwritten form, as it is very difficult to go back to find information on tapes. Often you won't be able to hear everything clearly, especially after several weeks have passed after the interview.
ix. Keep all field notebooks and your written out fieldnotes in safe and secure places where unauthorized persons cannot get them. Even seemingly noncontroversial, innocent-sounding information about foods and diet may be considered personal and private and you should always protect the confidentiality of informants' statements.
Keeping Fieldnotes Organized
Fieldnotes can quickly become an unwieldy mess unless you take pains to keep them organized. Even a few weeks of fieldwork will produce several hundred pages of notes that can be very difficult to manage unless some system of indexing is used.
i. In most cases it is best that your fieldnotes are kept in chronological order. That is, at least one set of fieldnotes kept more or less in the calendar order that the materials were collected. The pages of fieldnotes can be numbered consecutively. It may add some additional structure if you start new page numbers each month. So, if you started in October you would have notes from zero to one up to perhaps zero to 500. Then comes N- 1, N2, and so on.
ii. It can help you to find things if you put key words at the beginning of each interview, or at the beginning of each separate note. Some people put key words at the top of each page.
iii. If you are entering all your interview notes and other fieldnotes in a computer, you do not need to list key words when those words are in the text itself The normal search programs that you use with the computer can quickly find any words that are in the interviews. Only list key words that are important index words, or identification of topics, if they are abstractions that do not occur in the statements themselves.
For example, in the case of the bit of interview above, you would use key words such as prestige foods, SES, traditional or historical.
iv. During the first two weeks of information -gathering, as your interview notes begin to pile up, start listing the main topics in the notes-key words that you will use as an index.
v. If you are not using a microcomputer for writing and storing your interview notes, then you will need to write more key words at the front of each interview-including important words that are found in the text itself.
vi. Even if you are storing all notes in the computer, you will want to have hard copies (paper copies) in your files. As mentioned above, store your paper copies in two different, secure, locked places.
vii. With two separate sets of (paper copies) of interviews and observations, you can keep one set in the chronological order in which the work was completed and notes written; the other copy can be sorted out into the topical areas, key words that you begin to organize into different file folders.
viii. One way to keep good organization of your fieldnotes is to keep one set of files that are directly linked to the generic questions and other important questions.
Suppose you had a really good interview with an excellent key-informant and she told you information that applies to five or six different generic questions? Then, make extra copies of that interview, so you can put a copy into each of the five or six file folders that are answers to those generic questions.
The Field Work Log
In addition to the fieldnotes and diary, it is also recommended that the Project Manager maintain a field log. The log is a running account of your work schedule. The log should reflect your planned schedule for the next two weeks or more, as well as the record of actual times and places of interviews and other data gathering. Keeping a tidy logbook can help fieldworkers stay on schedule and maintain a sense of progress in the face of the daily frustrations brought by bad weather, hard-to-trace informants, and other problems typical of field work.
Some people like to keep a wall chart that shows the current status of the research, plus the expected schedule to complete data collection.
Here are some main points to follow in training your data-gathering team about translation from the local dialect or language to the national and international language.
i. Preserve vocabulary of key words and phrases in the original form, as used by the local people. This applies to the names of crops, foods, meals, dishes (types of prepared foods), attributes of foods, and other key words that emerge in interviews, as well as terms related to vitamin A deficiency.
For example, if local people have special words, or nicknames, for food items, these can be presented as used, rather than substituting the national language equivalent.
ii. Do not assume that words that sound like equivalents in the national language have the same range of meaning. For example, the word sopa in Mexican food culture sounds as if it refers to the same kind of food as soup in English. However, when we learn that a platter of noodles is also sopa we realize that the word cannot be freely translated, without further explanation.
The word tomati in Hausa almost always refers to tomato paste in arid regions of Niger, rather than a fresh tomato. In speaking of fresh tomatoes the adjective generally is added for clarification.
In many parts of the world foods are categorized in words that refer to hot or heaty. In some contexts the word may actually refer to the temperature of the food, while in other contexts the word hot (and the opposite, cold) refers to an abstract quality or attribute of food in relation to maintenance of a complex balance of hot and cold qualities in the body.
Thus the label hot or heaty concerning food may require considerable explanation, instead of simply literal translation into the equivalent word in the national language.
iii. Complex local vocabulary items should be presented first in the indigenous language, followed by the literal translation, followed by further clarifying explanation.
Example: In Hausa some people say: "/Shine/ /mini/ /koshi/." Literally: "/That food//makes me/ /full/."
The statement can refer to filling one's stomach, but in Niger where this manual was tested, the statement also referred to building up bodily reserves for a future time of food shortages. Thus statements in the field notes should always use the Hausa word, cowshi, rather than the French language equivalent, plein.
For important attributes or qualities such as cowshi it is useful to ask informants to use the word in different contexts. Interviewers can also try using the word themselves, asking the informants if this example is a correct use of the word.
iv. Investigators must watch for topics in which local assistants might not have full command of the national language, even though they have moderately fluent use of both languages in most areas of conversation.
Conversely, educated research supervisors and team leaders may believe themselves fluent in the local dialect, yet they may be lacking in local nuances. For example, in rural Haiti the term opresion, is considered by the local health professionals to be synonymous with asthma. There is, indeed, an overlap between the meanings of these two words, but in rural localities it turns out that opresion can refer to a wider range of sicknesses, including forms of bronchitis and pneumonia.
v. When passages of fieldnotes are translated into the national language, the key terms in the local language should remain imbedded in the text. For example, in Andhra Pradesh, India the language spoken in rural communities is Telugu, in which the abstract concept of hot/treaty is vedi. Since for research purposes the national language is often English, a sentence concerning this attribute should be written like this:
Fruits such as papaya and mango are rated high in vedi (hot/treaty).
vi. Team leaders can review the fieldnotes and reporting forms of the research assistants to see that local terminologies are carefully preserved and explained.
vii. Important features of local vocabulary are not only to be found in food names and their attributes. In some cases the local dialect will have special expressions or special slang for behaviors dealing with foods and eating. Also, there may be slang expressions or special words for types of persons corresponding to ideas such as picky eater, omnivore, etc.
As was stated in the manual, the selection of the list of key foods is crucial to the administration of the research procedures. The guidelines and criteria stated in section IV-C should be followed closely when determining the list. The following example illustrates the rationale behind the way in which a research team, led by Dr. Hilary Creed-Kanashiro, in Peru, established a final list of key foods.
The final list of foods included a group of local staple foods, cereals, and tubers selected on the basis of being most frequently mentioned in the free list, locally produced, and stored in the homes during most of the year. Oca is a local tuber which is more seasonal and less able to be stored, but has been included as an example of one of the locally-produced Andean tubers. Of the cereals selected, cebada (barley), trigo (wheat), and maize are most used. Quinoa has been included as it is regularly used during the whole year, is also readily stored, and has a high nutritive value. These also provide some income when sold, if they are produced in sufficient quantity. We did not include pan (bread), as it is infrequently used in comparison to the cereals or oca.
We included camote, although it is rarely consumed, as it is potentially an excellent source of carotene. It is the least expensive tuber, and children like it and eat it in large quantities on the few occasions that it is available. It is considered as a possible food that can contribute to vitamin A stores. It is also considered important during pregnancy and on giving birth.
Arvela seca (dried peas) has been selected as the most commonly used major vegetable protein source.
The animal products have been selected on the basis of frequency of mention, with the addition of breastmilk which was not mentioned in the free list exercise. Huevo (egg) is locally produced and most consumed. Pescado salado (dried fish, caballa), is considered cheap by the population and relatively easy to obtain from the Cajamarca market. It is able to be stored and is used fairly frequently. Carne de gallina (hen) is eaten more frequently than mutton (although not very frequently-often for birthdays and fiestas) and so was included.
With respect to fat sources, aceite (usually a combination of marine and vegetable oils) is bought from the Cajamarca market. However, it is not considered cheap and seb de carnero (mutton fat) is cheaper and bought when money is not available for the aceite. It is felt that it is used as frequently as aceite. Linasa was regularly reported in the free list and we considered it an important source of fat, usually prepared toasted with cereals. It is grown locally.
Of the fruits, we selected naranja (oranges) because they are considered inexpensive and are bought regularly and consumed most of the year on a weekly or fortnightly basis. Mangos are consumed in larger quantities when in season, February to April. Both of these foods are bought from the market. We excluded banana, although a staple fruit during the year, as the type that is most used is not a source of vitamin A.
Chiclayo (like a squash) is widely grown locally, frequently mentioned, and is used for sweet preparations and in soups.
We included aji seco (like chili) as an example of a food that is used almost daily in the aderezos (sauce used as a basis for the soups and the stews) together with oil and onion. Aji is rich in vitamin A although consumed in small quantities.
Zanahoria (carrot) is an important vitamin A-rich food. It is grown in the gardens and also bought from the market. It is used in small quantities but fairly regularly. Cochayuao (alga) is bought and is apparently consumed fairly regularly.
The remaining foods on the list are locally-grown green vegetables and herbs that are used in different food preparations. The first group are green leaves that are seasonal but could be important sources of vitamin A. They are mainly used in stews and soups, in quite significant quantities. Then there is a selection of smaller leaves used in a particular preparation caldo verde, typical of the region. Some of these are also used in smaller quantities to flavor soups.
A yellow flower has been included as a seasonal, but possible source of carotene. Finally manzanilla (camomile) flowers have been included as representative of the many infusions that are frequently consumed and which are all grown in the family gardens.
In the process of selecting the foods, we also looked at those foods that were mentioned in the interviews, to see whether there were important foods that had been underestimated in the free list. This influenced our decision when debating the inclusion of a few of the foods. We eliminated the hoja de olluco as there were many hojas and we found the reasons for using them were very similar to hoja de quinoa and replaced this with camote during the first interviews as we found that this was just a repetition of the hoja de quinoa.
Contents - Previous - Next