Contents - Previous - Next

C. Select key foods

In the following modules you will use a list of twenty five to thirty food items. Selection of these local key foods must take place before beginning the structured interviews. The determination of the key food list is based on information gathered from your key informants, including the free list of food plus the data collected for the Community Food Systems Data Tables. Observations made by the research team made during Phase I are also likely to be useful.

The list of key foods is chosen using the following criteria, with the first four items on the list being of particular importance:

i. Include fifteen to twenty vitamin A-rich foods (if there are that many available in the area). For clarification see Appendix 8 on vitamin A-rich food. The list should especially include those that are eaten by a large part of the population, that are available at affordable prices, and for which vitamin A is known to be available to the body.

Team members should read through the Community Food Systems Data Tables, using the information as a guide in choosing these foods. In particular, you should examine the vitamin A content of the foods, the availability including seasonal variation and cost, and the extent to which the foods are used by the local population.

ii. Also include seven to ten staple foods that constitute part of the daily diet (see the Glossary for definition of staple foods). Staple foods are included to understand how vitamin A-rich foods fit into the local food structure or food culture.

Examine the distribution of responses to the free list exercise. Items that are mentioned most often are probably very common and should be part of the key food list. The Community Food Systems Data Tables should, once again, be used as a reference when selecting the seven to ten staple foods.

iii. Food eaten by the majority of the population should constitute about 80% of the food list. This will depend upon price, availability, and seasonality of the food, as well as local acceptance.
iv. Those items most frequently mentioned in the free list exercise are probably important to the local food culture and should be taken into consideration when establishing the list of key foods. Items that are only mentioned once or twice may not be widely available. However, they may be important for you to study if they contain significant amounts of vitamin A.
v. Try to include at least two vitamin A-rich foods that are not regularly eaten, but were formerly eaten by the population or are possible important sources of vitamin A. These may include foods that have been phased out over time or foods that have been recently introduced to the community. Information on local foods that were formerly important to the community should be gathered from village elders. The research team should consult with key informants when identifying sources of vitamin A that are relatively new to the community.
vi. Depending upon the situation in the area, it may also be important to include one or two fruits or vegetables that have been recently introduced to the area and are widely consumed, but low in vitamin A. For example, during the garden season cabbage and lettuce are extremely popular new foods in West Africa. It may be important to study how people view these and compare them to foods that are grown at the same time of year but which are rich in vitamin A, such as red sorrel or amaranth.

Consult the Community Food Systems Data Tables for detailed information on the items identified through the free list and make decisions based on the criteria outlined in this section. When selecting food items identified during the free list, consider how each food fits into the community food culture.

vii. Consider local food group categories. Examples of local food groups might include fruits, vegetables/ greens, meats, carbohydrates, and dairy products. When determining the key list you should select foods that represent the food groups consumed in the area. If a category is totally missing, you may want to add a food item from that category.

See Appendix 7 for an example from Peru on selecting the key thirty food items.

FORM 1.1 Free List Record Form for Food Names by Individual Key-Informant

FORM 1.2 Food List by Number of Key-Informants Who Mentioned the Foods

FORM 1.3 Food List from Most to Least Often Mentioned by Key-Informants

D. Identify food plant and animal species

1. Plant identifications
2. Animal identifications

There is usually little problem in identifying well-known foods by the names used in food composition tables, but indigenous crops and wild species probably cannot be identified without collecting specimens and/ or enlisting the assistance of a botanist or zoologist for their scientific identifications.

Plant and animal foods are generally referred to by two Latin names, e.g., Daucus carota. This specific pair forms the scientific name for the carrot, Daucus being the genus and carota being the species. Food composition tables may list foods by English or other names. Scientific names, properly applied, avoid confusion and the need to translate among different languages.

Interviews with key-informants will ideally give you a complete list of food items used in the area. Each food will be entered on a Food Data Sheet for the Community Food System Data Tables. Existing data and your interview data are then used to complete the table as much as possible. Very often, scientific identifications are required and assistance will be needed, as noted above.

1. Plant identifications

To facilitate the identification of plant foods, specifically, the following steps are recommended:

i. Careful recording of plant names in the common local language and any other languages known by the interviewee. Be sure to ask for the name of the plant, not just the part eaten, as they may be different. Many local cultural names can be found in ethno-botanical reports and it is useful to familiarize your self with any ethno-botanical literature available for the study area. Many regions were the focus of botanical studies from the late 1800s to the early 1900s. The descriptive results of these studies are often located in botanical journals or local museums. It is of value to ask individuals of both genders and all ages about the plant names and their location as the identification, collection and preparation of plants can be a collective effort within the family and/or community.
ii. Contact a botanist familiar with the local flora as early in the study as possible. In most countries such specialists are on staff of museums, herbaria, or departments of botany or pharmacology in the universities. It may be possible for the botanist or botany students to accompany you to the field at the appropriate time, especially if travel funds are available, although you should be prepared to bring samples to him or her. Botanists at major international institutions are also familiar with flora from many different regions, although the arrangements for working with them may be more complicated.
iii. Collect herbarium specimens for identification by the specialist. Botanists are accustomed to working with dried, pressed specimens, although the quality of the specimens you bring them will greatly determine success. A botanist can advise you on proper collection and specimen preparation techniques. For collecting plants in the field, a supply of large plastic bags will come in handy. To make herbarium specimens the plants, or parts cut from larger plants, are spread out between folded newspapers and pressed in a plant press, minimally two wooden frames (30 x45 cm) bound by straps or rope. Corrugated cardboard or metal may be placed between specimens to increase aeration during the drying process. Plants can be dried in the sun, particularly if they are few in number and the ambient humidity is low, or placed over an external heat source, such as a portable stove. Specimens should be representative of the plant and ideally include distinguishing features such as flowers and fruits. In the absence of flowers and fruits, the best substitute is a detailed description by the interviewees. Collect specimens that fill the 30 x 45 cm sheet or enough of smaller plants to fill the sheet. Careful note-taking can make the botanist's job easier. The form of the plant, whether it is a herb, a shrub, or a tree, and its size should be noted, along with other distinctive features such as flower color (color may fade as the plants are dried). The date and the place the specimen was collected along with the details on the habitat in which it was growing should be recorded. Good quality photographs may assist the botanist but they are not a substitute for good specimens. However, photographs or drawings are valuable additions to the food systems data tables and may help in your interviewing within the community.

2. Animal identifications

It is more difficult to collect unknown animal species for identification, but it is less likely that animal food will be unidentified. Possible exceptions are insects, fish, and small birds. Begin with the local names and photographs of the intact animal before it is prepared for eating. Some animals will have variations in color according to the season and/or gender. Additional descriptions may be obtained from the interviewee. Put a small ruler next to the animal in the photograph to characterize size of the species. Check these with local experts and zoologists (same location as botanists) for formal identification.

E. Determine nutrient content of food

If the food item of interest appears in the food composition tables, determination of their potential as vitamin A sources is straightforward. However, for many indigenous and wild species, data may not be available in this form. In many cases, the species have not been studied for vitamin A content, although data may be available in scientific papers that can be located through a careful literature review. Leafy, green vegetables and fruits, flowers, juices, vegetables, and tubers that are red or yellow in color, and animal organs and fats, and some milk products have the greatest potential as sources of provitamin A and vitamin A and should be considered the most carefully. Laboratory analyses are beyond the scope of a rapid survey such as this, although foods with high potential that have not been examined for vitamin A content should be ultimately studied. If the opportunities are available for collaborating with food analysts, you are encouraged to explore with them the proper manner to collect specimens for analysis.

If you are compiling vitamin A contents of food for your data tables, it may be valuable to consult with the national food analytical laboratories for advice on the accuracy of the data from your references. Analytical methods are quite variable in their accuracy and caution is advised.

Appendix 8 gives some guidance on general levels of vitamin A in food.

F. Select the sample of mother-respondents

Approximately twenty-five to thirty households should be selected for the structured interviews. The interviews will be carried out with mothers or other primary caretakers, of children ages six months to six years. Ideally, a random sample would be taken from the community to insure representativeness. However, random sampling is often not feasible unless a full (recently updated) list of all households is available. If such a list of all the households is not available, you can develop a method such as selecting every third or fourth house, (depending on the size of the community) in order to get a representative sample. See Appendix 10 for further discussion of the sample selection strategy.

Avoid including minority ethnic groups that comprise a very small percentage of the local population. On the other hand, if the community is composed of ethnic groups that constitute significant numbers, it will be essential to select sub-samples of these groups to be interviewed. If this is the case, it is important to include a large enough sample of respondents representing a particular ethnic group for the analysis to be meaningful. For example, if 25% of the population constitute a minority ethnic group and the rest of the community comes from the same ethnic group, you should select 50% of the respondents from the minority group and 50% from the rest of the local population.

In communities where the nuclear family is the dominant pattern, the principle caretaker of infants and children is generally the mother. In many areas, however, the grandmother or older siblings may play a key role in the care of children. Depending upon the situation, it may be necessary to conduct the structured interviews with other caretakers as well as the mother. If you determine that it is appropriate to interview more than one caretaker in a household, be sure to record the responses on separate data collection forms. However, if more than one interview is conducted in one household, for sampling purposes, they are counted as one household.

In selecting households (mothers) for the sample remember the following criteria:

i. There must be at least one child six months to six years old living in the household.
ii. The woman you interview should be a permanent member of the community, not a visitor, and preferably have resided there for several years.
iii. Each respondent (mother or other caretaker) should be available for a series of interviews over the next five to six weeks (anticipate approximately one hour per visit). When you first meet with the mother respondent, ask whether she will be available and willing to participate in the study.
iv. Mother-respondents should represent ethnic groups living in the community, as noted above.

For further suggestions, see the sample section in Appendix 1 which illustrates how the research team in Peru selected the study sample.

G. Manage the data

Preparation of the report from this research is greatly enhanced by the use of the data-collection forms and data tabulation sheets with the modules. The research team will need to establish systems to manage the forms which can quickly accumulate and become unruly. It is important that the research assistants include the respondent numbers on all individual forms and that the completed forms be filed in an orderly and simple fashion. The forms can either be separated according to respondent or procedure. This depends on what your particular research team finds most logical and convenient.

Many of the forms need to have the list of the twenty-five to thirty key foods written on the form prior to the interviews. Remember that the food items retain the same order on the food list and do not change. If a photocopy machine is available a tremendous amount of time is saved by making master forms that can be copied for the sessions with the respondents.

It is also useful to keep a table or log with a list of all of the respondents, with important background information and notes showing which procedures have been carried out with each individual. This table or log is updated on an ongoing basis and can be kept in a notebook and/or presented on a blackboard or flipchart. It is recommended that every evening after interviewing sessions the team meets to review what was accomplished during the day and to discuss the following day's events. At this time the research manager ensures that the forms have been properly filed and that new forms are prepared for the next day's activities.

Other Considerations

Remember that your goal is to capture emic language and information. Therefore, to preserve the integrity of the data you will want to conduct the exercises and record the responses in the local language. Translations and interpretations will need to be made later when the data is analyzed and for the writing of the final report.

You will find that the exercises are more effectively carried out with two people, one conducting the interview and the second recording the responses. Before beginning interviews with mother-respondents decide who will record the responses. This should depend upon the skills of the field assistants and determined in practice sessions. Determine which modules require two interviewers and which can be carried out by one interviewer.

Contents - Previous - Next