Contents - Previous - Next

Overview of methods of data-collection described in the protocol manual

The FES protocol addresses a series of questions about the local situation of vitamin A and health. The questions are ordered as primary and secondary, with the secondary questions giving detail to the primary questions. The primary questions are:

What are the key foods? Secondary questions relate to food acquisition, details on location of markets and their prices, food grown in gardens, and gathered from the wild, etc.

What are cultural beliefs about key foods? Secondary questions include the qualities or attributes of the food, how these vary by age and gender, perceptions of vitamin A in food, suitability of food for children, etc.

What are the patterns of food use? Secondary questions include pregnancy, lactation, infancy, and childhood; perceptions of food use and health/illness; specific foods recommended during these times; degree of breastfeeding, etc.

How is food prepared and stored? Secondary questions relate to length of cooking, extent of drying, use of spices, use of vitamin A-rich food in these ways, etc.

What are the signs and symptoms of vitamin A deficiency? Secondary questions refer to local terms for deficiency states, beliefs about the diseases, treatments used, perceptions on degree of illness, etc.

• General questions relate to women's work, family food distribution, food purchasing, presence of infections, and exposure to nutrition and health education.

The methods of data-gathering, that will be described below, are:

• in-depth interviewing of key-informants;

• direct observation in stores, markets, and agricultural production areas;

• structured interviews with small samples of respondents;

• food frequency and 24-hour dietary recall interviews;

• use of written resources to obtain background information about the community and region, demographic and health status data; food species and composition data, past nutrition surveys; ethnographic studies and other descriptions of economic conditions, life styles, and food beliefs.

Selecting the specific locality for the study requires attention to several criteria. These include:

• The area chosen should be characterized by well-documented vitamin A deficiency.

• The area should be representative of a major cultural/ecological region of broadly similar conditions (including nutrient deficiencies), and in which similar food availability and food practices are found.

• The road and communications network systems should be such that data-gathering logistics are not overly difficult.

• The prospects for developing interventions programs, through existing health service networks of other agencies or community groups should also be considered. Frequently the site of a FES data-gathering operation can serve as a suitable location for experimentation with a pilot intervention program.

Key-Informant Interviews

While the goal of the study is to obtain the specific information required to answer the study questions, the manual directs investigators to begin gathering information about foods and food use by engaging in broad discussion with people in the community. Talking with the people leads to the identification of persons who have a lot of knowledge about local food culture and practices, and who are able to formulate clear ideas about why people do what they do with food in the local area. Persons who can talk effectively about their own food use and diet, as well as about general practices, are sought out as key-informants.

Free Listing

An important early step in key-informant interviewing is the application of the free listing technique. The informant is asked a series of questions that are designed to yield a list of foods about which other information can be collected. Examples of the initial questions to elicit food lists include:

• "Please tell me all the kinds of food that you grow here in your garden?"

• "Please tell me all the kinds of food that you can gather from the fields and forests - things you do not plant, but grow naturally and can be eaten?"

• "Please tell me the names of the foods you buy in the market each time you go there?"

• "Are there foods you buy only sometimes, or rarely?"

The lists, gathered from several informants, provide different kinds of information:

• The names or labels given to food, as well as synonyms for food. (The key-informant will sometimes volunteer information about different words with a comment such as, "Some of the people in the lower village call that _________.")

• As no one remembers every food, a composite list begins to form an exhaustive picture of the local food system.

• Certain items, that are mentioned first and frequently, are usually the most important, the central features of the food system. Less frequently mentioned food items are usually the peripheral sources. Researchers will often find important vitamin A-rich food in the peripheries of the food system.

Describing the Food System

Two data-collection modules are used to obtain information about the general community food system: the community food system data tables and the market survey.

Food system data tables provide a clear framework for organizing the data-gathering. Filling our community food system data tables is started early in the course of the study. The purpose is to obtain systematic information about vitamin A-rich food, as well as other main sources of nutrients in the local diet. In many target areas investigators will be able to use existing (recent) dietary studies and food composition tables as a starting point.

An important feature of the FES methodology, as noted above, is that local names must be carefully collected for major food sources, so that conversations and interviews with the population are phrased in appropriate language. Investigators must be alert to the possibility that some basic food materials, and some types of prepared food may have different local names in other communities. That is one of the reasons key-informants are of central importance in this stage of data-gathering.

The food system data tables also call attention to seasonality, prices, and the importance of the specific food for the target groups of women and children. Table 3.1 is an example of a data table for one food item.

The instructions in the manual point out that there may be some formerly used, or little-used (wild or cultivated) sources of vitamin A that only a few, older key-informants would remember or have information about. Some of them may be thought of as famine foods by informants. Special effort is required to search for these little known foods.

The market survey consists of direct observation plus interviewing. This technique has an important advantage-the investigators do not have to rely on reports from others, but can see with their own eyes what foods are available in the local markets. Questioning concerning prices of food can, of course, be a more difficult task. The manual suggests that local persons who are familiar with local market conditions are best for finding out food prices in the weekly or daily market scenes.

The market survey is repeated at least twice during the six or eight weeks of data-gathering. Also, questions about prices in the marketplace require investigators to obtain information concerning seasonal variations. A great amount of useful information about available food, the origins of the food, and potential for vitamin A intakes, can be gained through observations and casual conversations in the markets. Fixed stores that sell food are also an integral part of the market survey data.

Collecting Structured Data from Small Samples of Respondents

Food use, dietary practices, and food systems in general are what John M. Roberts has labeled "high concordance systems," (Roberts et al., 1981), although he applied the concept in a different cultural domain. Food systems can change rapidly in response to modernization. However, for most communities, especially in rural areas, strong cultural preferences and powerful economic/ecological forces control many aspects of food availability, hence the patterns of food use are strongly structured. Just as any reasonably cosmopolitan diner knows how the food in a Chinese restaurant will be different from that in an Italian trattoria, so the food styles or menus, in rural villages in central India have a particular, predictable, clearly defined format, with only secondary, peripheral variations from one household to another. Of course socioeconomic status affects how families eat, and wealthier families generally have more variety in their diet, but their food patterns (menus) are not in a different world from their less fortunate neighbors. They have the same basic pattern, with the addition of foods that cost more; in most parts of the world this means more animal foods.

Because of high concordance in food culture, it is not necessary to have large samples to describe the basic patterns of food use in relation to vitamin A sources. Instead of surveys of hundreds of households or hundreds of mothers, we can be fairly assured that the basic patterns will be evident from samples of twenty to thirty respondents.

Small samples are adequate to achieve a descriptive goal only when they are representative of the community. Sample selection must give adequate representation to the poor as well as the more affluent portions of the communities. Where feasible, a random sample of households should be used. In areas with important ethnic variations, the samples have to be larger to represent the various ethnic groups.

The following sections describe data-gathering modules that are used with small samples of women:

Cognitive Mapping Procedures

This set of interviews makes use of cognitive mapping techniques from anthropology and cognitive psychology to describe emic perspectives on food.

Pile Sorting of Foods

This technique uses cards or slips of paper-one card or slip of paper for each food item. Each card has the name and a simple picture of the food, for ease of recognition by nonliterate respondents.

The pile sorting technique is done best with about thirty items, and the manual suggests that the investigator select up to fifteen to twenty vitamin A-rich foods (including some that are little used or even totally neglected) and at least ten or tweleve other foods. These should be staple foods that are included to make up a representative sampling of the entire food system. The respondents are asked to sort the food items into piles or groups that go together. To assist them in understanding the task, various analogies can be suggested. For example, the sorting out of laundry is sometimes an effective analogy. The respondents are told that there are no right or wrong answers, and that they can use as many, or as few, piles or groups as they wish. The method has proved to be effective in all of the test sites.

Example of an Entry in a Community Food System Data Table

Food Category: Green leafy vegetables (indigenous)
Local Name & Other Common Names: Chomtee (K'ekchi); chilete dulce; quilete (Chiapas)
Scientific Name: Lycianthes Synanthera - Bitter
Part(s) Used: Leaf, preferably young
Preparation: Leaf boiled in stew and soups, primarily for the flavor


Nutrient Composition/100g (Edible Portion by Part)




Energy, kcal, kJ

Protein, g

Fat, g

Retinol, mg

None Available

Beta carotene, mg

Total carotene, mg

Retinol equivalents

Ascorbate, mg

Riboflavin, mg

Niacin, mg

Iron, mg

Calcium, mg


Wild, Hunted/Gathered, or Cultivated: Gathered in rural communities only (N/A in urban)

Home Harvested or Purchased: Only available periodically in rural market (Q0.14 - 03/90 market price in Carcha, Guatamala)

Seasonality of Use: Year round availability, with greatest intake in February and April when the maize fields are cleared of indigenous greens and weeds

Cost of Production, if Known: ________________________________
































clearing of maize fields

Importance Value to the Community by Age/Gender:

No specific preferences, except some parents claim that children dislike greens.
Only used as flavoring in small quantities, and not very frequently, hence minor potential as source of Vitamin A

Miscellaneous information: Little information exists on this species, with at least one variety previously classified as a distinct species (ref. Gentry and Standley, 1974)

In addition to giving the emic sense of food groups, the sorting permits data-gatherers to ask about the meaning, or criteria of similarity, among foods in the same group. Often the respondents' explanations are more revealing and important than the actual composition of the groups themselves. For example, in some societies wild or domestic green vegetables may be used as a relish, sauce or flavoring, or used in much smaller quantities than in places where the same food is considered a vegetable.

Food Attributes

Getting people's views about the attributes, qualities, or selection criteria for food is fundamental for planning dietary intervention, as they provide the basis for communicating about specific food items. For example, in an educated middle-class urban population, you expect to find that people are interested in vitamin-rich food for their children, or food with iron and calcium, whereas these traits may be less meaningful to rural people.

Of greatest interest to participants in vitamin A interventions in deficient areas is gaining an understanding of how people select or judge various foods for themselves and for their children. In many parts of the world, for example, food may be labeled as "hot" or "cold," in ways that affect decisions about feeding them to pregnant women and children. Such attributes can affect the success or failure of intervention programs that encourage people to increase consumption of particular food items.

The same list of food developed for the pile sorting, described above, can be used when interviewing for food attributes. The list of twenty-five to thirty-five food items is the basic format. Each respondent is asked to tell about the qualities or characteristics of each food. In some areas it is useful to ask: "What would you describe as the 'personality' of this food?"

The attributes or criteria that were given for the various piles or groups in the structured pile sorting can be used as prompts or probes for getting respondents started with this topic.

Rating by Predetermined Attributes

In this technique respondents are directed to sort the cards by categories that are predetermined by the investigators because of their relevance to potential interventions. These categories may include "good for the eyes," or "good for pregnant women."

Sorting by Degrees

This mapping procedure is another technique for examining attributes. Respondents are asked to sort the food cards into three or more piles representing degrees of a previously identified attribute. For example, researchers ask the respondents to sort the food cards into those that are "very hot," "very cool," and "intermediate."

Household Food Acquisition (Structured Interview)

Many of the daily food items regularly used in an area are available in markets and stores, but are also homegrown. Some families rely mainly on their own production; others are busy with other occupations, so they buy most of their food. The patterns vary from one community to another. They also vary among households, even among families of similar socioeconomic status. Therefore, it is important to assess the range of variation in the target population.

The list of food is the same as used for the pile sorting and other structured interviews. Before beginning this set of structured interviews the field workers have already established all the different ways by which people obtain their food (gathering, hunting, growing, buying in store, buying in market, exchange for labor, etc.). Households may obtain some food items from several different sources. For example, for some households the response to "rice" might look like this: Rice-homegrown 50%, exchange for labor 25%, buy in market 25%.

The form used for collecting food acquisition data also includes the usual price (per usual quantity), season, and which member of the household acquires that food item. In many regions certain foods are acquired, grown, or purchased by the women, while others are acquired by the men. Also, the role of children in acquiring the food items can be important.

Besides getting the overall picture for the acquisition of different important food items, a major aim of this element of data-gathering is to find out if some subgroups in the community have different food acquisition patterns that would affect any planned intervention program. Information about seasonal variations is also important to consider in planning interventions.

Individual Food Intakes: 24-hour Recall and Food Frequency Check Lists

This step in the data-collection, again from a small sample of households, draws on standard procedures from nutritional science. The intent of the data-gathering is to get a systematic view of the actual (reported) food intakes of the children from six months to six years of age, as well as women of reproductive age. Special attention should be paid to the dietary intakes of pregnant and lactating women in the community. It is recommended that researchers seek out small samples of pregnant women in order to supplement the data gathered from key-informants.

Obtaining Information on How People Recognize Vitamin A Deficiencies. Hypothetical Scenarios

This section of the manual is designed to find out whether people recognize signs and symptoms of vitamin A deficiencies, what explanations are offered concerning these manifestations, and what treatments, if any, are utilized. The technique of data-collection is the presentation of hypothetical scenarios about which the respondents are asked a series of questions. The manual instructs the data-gatherers to modify the scenarios to fit local cultural conditions (including, of course, the names of the persons mentioned). The scenarios include:

• a three-year-old boy with nightblindness;

• an eight-month-old infant who is being bottle fed, and now has diarrhea, sore mouth, rash, and other symptoms;

• a five-year-old child with a white foamy patch on the white part of the eye;

• a woman who is seven months pregnant and suffers from nightblindness.

In the manual, the interviewers are instructed to record all home treatments, as well as actions or medications that would be obtained from other sources, including health providers.

For ease in completing the data summaries, instructions are given for categorizing or grouping the responses. For example, "remedies" might be grouped into the following categories: (1) Food, (2) Rest and/or bed, (3) Medicine (capsule, tablets), (4) Medicine (injection), (5) Tea, (6) Magico-religious ritual, (7) Other.

Additional Guidelines and Information in the Manual Appendix

The manual has further comments and guidelines designed to be used in the training of field data-gathering personnel. For example, there is a list of "Do's and Don'ts" that cover major ethical issues, as well as other elements of good data-gathering practice. Data-gatherers are particularly urged not to criticize people about their food habits, hygiene practices, and other behaviors.

The Appendix also gives suggestions concerning selection of field sites and data-gatherers; training of field personnel; notes on translation from local to national languages; comments on using microcomputers; further notes on working with key-informants; a section on recording and organizing field notes; checklists of important background information concerning the research location; guidelines concerning the vitamin A contents of various types of foods; suggestions about selection of representative samples of respondents; and finally, a glossary of terminology.

It is our feeling that the presentation, in FES manuals, of specific, steep-by-step procedures, gives field interviewers and their supervisors confidence that they can indeed assemble the information needed for the practical objectives of the study. For some people, the data-recording formats are at first confusing, even daunting. However, when they are approached one step at a time, the mystery disappears, and field teams are able to manage the data easily. Having a specific format, labeled and ruled off in specific fields and boxes, gives a clear meaning to the vague and mysterious directive, "record the qualities or attributes assigned to each of the foods...."

The table of contents of the manual, including the list of appendices and data forms are given in Appendix 1 at the end of this book.

Contents - Previous - Next