Contents - Previous - Next

List of acronyms


Rapid Assessment Procedures


United Nations Children's Fund


Nigerian Family Health Services Project


United Nations Fund for Population Activities


Health Communication for Child Survival Project


Expanded Program of Immunization


Oral Rehydration Therapy


Information, Education, Communication


Focus Group Discussion


International Development Research Centre


Primary Health Care


United Nations University


Research Bureau Nigeria Ltd


Community Based Distribution


Federal Radio Corporation of Nigeria


Nigerian Television Authority


Integrated Community Health Development


International Institute for Tropical Agriculture


Local Government Areas

7. Understanding infant feeding practices: Qualitative research methodologies used in the weaning project

Historical perspective
Sample results

By Marcia Griffiths

Marcia Griffiths is president of the Manoff Group, Washington, DC1.

Marcia Griffiths describes her experience in a number of different countries in developing a qualitative approach to understanding infant weaning practices and designing effective programmes to improve them. The countries included Indonesia, Cameroon, Swaziland, Ecuador, Ghana, and Zaire. The formal RAP guidelines owe a great deal to her experiences in this programme, and particularly to her introduction of the use of focus groups for this purpose. - Eds.

A CRITICAL FEATURE of successful public health programmes is that managers know the programme's clientele and tailor policies and activities to meet needs in an acceptable manner. Less successful programmes are often explained with such phrases as "we never anticipated that people's reactions would be..." Social science researchers can enhance programme success by offering managers a more succint view of their clientele. However, managers seldom have large budgets for research. This means that effective research which is planned must be completed quickly, below cost, have immediate relevancy for programming, and yield new, useful insights on the clients' perspective. The methodology used in The Weaning Project, to explore and better understand young child feeding practices, met these criteria. It was:

1. Consumer-based, and as open-ended and free of researcher bias as possible;

2. Relevant for programming purposes, particularly the design of communications and training activities;

3. Adaptable to different situations (the same basic protocol was used in six countries);

4. Replicable or manageable by professionals with limited research experience; and

5. Relatively quick and affordable in many development projects. The methodology used in The Weaning Project is the product of nine years of project experience with consumer research, most of it related to the exploration of infant and young child feeding practices.

Historical perspective

In the mid-1970s, in Nicaragua and the Philippines [1] for the design of a weaning food and oral rehydration education program, local researchers were trained to apply basic survey techniques and some open-ended questions with consumers, not too different from the KAP surveys traditionally, and still, done by health educators. Fieldwork took about two months. Evaluations of the resulting educational programs showed their weaknesses were due primarily to the fact that the initial research had been too "researcher-determined" and missed many subtleties needed for message design.

In the late 1970s, research was designed for a nutrition education project in Indonesia [2, 3] in a modified ethnographic style, with open-ended, detailed studies (interviews and observations) in carefully selected communities in the program area. In addition to the ethnography, an innovative step was added: participatory research borrowed from marketing. Mothers were asked to try out potential recommendations to get their reactions to preliminary messages and to solicit their contributions to revising the proposed suggestions for changes in standard practices. This worked well. The resulting educational program was associated with improvements in practices, increased intake of calories and protein, and improved nutritional status among children under 24 months. However, this methodology required nine months of fieldwork and from design through analysis, it took almost a year and a half. The process was guided by a full-time expatriate nutritional anthropologist. Not all programs have that luxury.

Following on this success, the challenge was raised to reduce research time and to incorporate more techniques that would help better understand lifestyle context: aspirations, desires, fears, attitudes toward child rearing, etc. For work in the early 1980s a component of focused group discussions was added, but the step of the trial of practices and much of the other contextual information gathering, typical of ethnographic research, was eliminated. This work was more like a rapid assessment. In the Dominican Republic, the process took about two months [4]; in Ecuador, about three months, from planning through analysis [5]; in India [61, about six months before the special step of the trial of recommended practices was added. Again, almost full-time guidance from a nutritional anthropologist was provided. While the resulting programs in the Dominican Republic and Ecuador were relatively successful, there was a feeling among project personnel that the education would have been better if the research plan had allowed for more exploration of the reasons for mothers' practices and particularly their willingness to change.

Based on these experiences and other work in nutritional anthropology 179], The Weaning Project developed a protocol for exploring young child feeding practices and refined it in six countries.

The complete protocol was first implemented in Indonesia and Cameroon and included a large amount of foreign technical assistance [10]. Later, in Swaziland, Ecuador and Zaire, the most salient pieces of the multi-step protocol were chosen, modified, and implemented primarily by host-country researchers, with periodic technical assistance [11]. After these experiences, there was an opportunity to utilize the protocol in Ghana. Again, based on experience, the protocol was reduced further and this time implemented with only brief orientation from expatriate consultants.


Many of the decisions made in designing the methodology and writing the protocol were to allow researchers to go beyond the usual researcher-determined questions about feeding practices and to explore with mothers, in their terms, how they make decisions. To do this, techniques from market research, anthropology, and nutrition assessment were combined to help researchers understand, not only the importance of the different determinants of infant feeding practices, but also the lifestyle context in which infant feeding decisions are made.

The assessment methodology has several characteristics:

1. It is basically qualitative, with some quantitative analyses.

2. It has several steps, each of which builds on the preceding one, so there is limited duplication effort.

3. It is in-depth, to explore the reasons behind everyday practices, beliefs and perceptions.

4. It is rapid, although this depends on what "rapid" means. In the first countries where it was implemented in pilot regions, the process took a year. Now, the time has been reduced to six months for a national assessment.

5. Its implementation requires minimum technical assistance, although it does require a principal investigator with knowledge of qualitative research.

The protocol is divided into four parts, corresponding to the research phases. Each phase has several steps. Not all of the steps need to be done in every situation.

Generalized protocol



1. Find critical problems impeding proper feeding and care of children;
2. Identify resources to solve problems. (Resources include physical and financial resources as well as outlooks and attitudes.)


1. Literature review. A complication of relevant information from all previous research. Most of this research is quantitative in nature; therefore, the review serves as a springboard for the design of the qualitative study and as a check on the results ultimately obtained from the qualitative process.

2. Focused group discussions. These are extremely open-ended and explore maternal roles, sense of control and confidence, ideas about child rearing, aspirations for children, general feeding practices, and images associated with certain practices.

3. Ethnography. This is a community and household exploration of food availability, women's time availability, cultural norms about child feeding, ceremonies, people who influence feeding decisions, the details of food preparation, serving and consumption, childhood morbidity, etc. It usually includes child anthropometry and dietary recall.


After the Indonesia and Cameroon experiences, stage one was modified:

1. Unless there are persons skilled in focused group discussions with in the country, the discussions are eliminated because it has proved too difficult to train people to get quality information using this technique.

2. The ethnography has been collapsed to in-depth household interviews and observations and some key informant inter views. The extensive questioning on food grown and purchased, on relationships between family members and on ceremonies, was dropped. Although more limited, the work is structured judiciously the households are selected carefully. They include a range of different age children, usually under two years old, and favour children who are growing well or who are undernourished.

Depending on the scope of the work, this initial phase can take up to three months, including planning and training investigators.



1. Determine nutritional benefit or harm from current practices;
2. Identify modifications in practices and rationales for them.


1. Case histories.

2. Dietary analyses and group feeding histories.

3. Tabulations and content analysis of the different topics, including dietary analysis by geographic area, the age and/or nutritional status of the child, by the amount of time the mother spends with the child, etc.; and

4. Matrices to compare ideal and real practices and list major resistances or motivations that may influence a change in the practice.


This process is done at research headquarters and takes about a month. Most research methodologies end at this point. It is noteworthy that in The Weaning Project, this was not the case.



1. Determine what mothers are willing to try and why;
2. Confirm what mothers can do over a brief time period;
3. Retest the successful concepts with even more mothers.


1. Participatory research where the researcher returns to the homes of mothers who participated in the in-depth interviewing and provides information to these mothers about their children and discusses with them their willingness to try new practices. Then, with some mothers, the researcher actually asks them to do it - for example, to try to give their children one more meal, or a snack between meals, or to make a different weaning food. The researcher returns to the home to see if the mother has been able to follow the recommendation and if so, exactly what she has done. This step has proved easy to do and analyze once the recommendations have been decided upon. This step is indispensable.

2. Focused group discussions: The most successful ideas and practices to emerge from the trials are taken to other communities where focused group discussions are held to get the "top-of-the mind" responses to the new ideas from people who have not participated in the earlier work and to get reactions from health workers and clinic nurses - those responsible for disseminating the new information.


This phase takes four to six weeks from planning through analysis.



1. Review all of the information;
2. Synthesize information from the problem identification and concept testing phases;
3. Write a brief for programme designers.


This report serves as the basic reference for those making strategy decisions and developing the creative work. It is the link so often missing between researchers and programmers. The style in which it is written is abbreviated to assist programmers to find the facts quickly and to understand recommendations about what to do.

The report contains the following sections:

1. The environment or lifestyle of the population;
2. A summary of current infant feeding practices;
3. A list of the most promising practice improvements, the major resistances to change and the possible motivations to stimulate change;
4. A review of potential media - their reach and the frequency with which they are heard or used.

The total time for this type of assessment is about six months.

Sample results

In The Weaning Project new information was abundant. Results common to almost all of the projects include:

1. The importance of maternal self-confidence in child feeding [12]. A mother's level of self-esteem and confidence seems to determine the amount to which she is swayed by her child's response to foods. It appears that in general mothers with well nourished children have more self-confidence they introduce foods when they feel it is right rather than when the child takes them. If they stop breast-feeding early, it is more likely to be because they want to than because of their child's reaction; they are more likely to persist in feeding their child when the child refuses; and they are more willing to try new foods and practices.

2. The significant role of fathers. Fathers seem to be playing, or seem willing to have a larger role related to child feeding. This was unexpected. Fathers are often the ones to convince when it comes to the purchase of "special," calorie-dense foods for young children.

3. Using store owners and food vendors. These community members hold great potential for disseminating some of the program messages, especially those related to foods. These people are often the most stable and abundant "medium" in the community and, in many instances, are knowledgeable about food-related topics.

4. Establishing priorities among the different factors in child feeding by the age of the child. That is, concepts of nutrient density, feeding frequency, and food quantity are difficult for mothers to accept all at once. For them, each component is appropriate for different aged children. The message then must be tailored by the mother's perceptions. For example, for infants four to six months, it is important and acceptable to focus on the consistency of food (decreasing its water content); from seven to 11 months - on feeding frequency and food variety; from 12 to 24 months - on feeding frequency and food quantity.

5. The early onset of the weaning process. While this varies by country, in several places there was no period of exclusive breast-feeding. There is an increasingly popular tendency to introduce foods early to "accustom the child to food." A priority of The Weaning Project has been to address this problem, in part by enhancing women's confidence in their ability to breast-feed their infants.

6. The worst characteristics of daily feeding pattern occur during and following illness. If mothers already allow children to determine their own feeding patterns, they will do so even more during illness. If mothers give only a small amount of food regularly, they will further reduce the quantity for an ill child. However, mothers generally continue breast-feeding and do not withhold food because they think they should, but because the child "just won't eat."

The six months that researchers spent to gain these types of insights into weaning problems and solutions was cost effective. The resulting programmes have been successful in achieving improvements in practices and those improvements have improved dietary intake and young children's nutritional status [13, 14].


1 The Weaning Project was supported by the Office of Nutrition, Bureau of Science and Technology, USAID, USAID/Jakarta, CARE/Cameroon, USAID/Quito, and UNICEF/Switzerland.


1. Manoff International. Radio advertising techniques and nutrition education: a summary of a field experiment in the Philippines and Nicaragua. Final report To the Agency for International Development, The Manoff Group, Washington, DC, 1977.

2. Griffiths M. Nutrition communication and behavior change component, Indonesian Nutrition Development Program. Volume 1: Concept testing. Report. The Manoff Group, Washington, DC, 1980.

3. Griffiths M. Using anthropological techniques in program design: successful nutrition education in Indonesia. In: Coriel J, Mull JD eds. Anthropology and primary health care. Boulder, Colorado: Westview Press, 1990.

4. Griffiths M. A nutrition education strategy for the Applied Nutrition Education Project (ANEP), CRS/CARITAS, Dominican Republic. Trip report. Newton, MA: International Nutrition Communication Service, 1984.

5. Griffiths M. Una estrategia para la communicacion nutricional dentro del programa de mejoramiento PAAMI del MSP. Newton, MA: International Nutrition Communication Service, 1984.

6. MODE. Report for a research study on nutrition and health education development, state 1. For the Ministry of Human Resources Development, Government of India. Bombay: MODE Services, April 1986.

7. Scrimshaw SCM, Hurtado E. Rapid assessment procedures for nutrition and primary health care: anthropological approaches to improving programme effectiveness. Los Angeles: UCLA Latin American Center, 1987.

8. Van Esterik P. Integrating ethnographic and survey research: a review of the ethnographic component of a study of infant feeding practices in developing countries. Working paper no. 17. New York: The Population Council, 1983.

9. Pelto G. Ethnographic studies of the effects of food availability and infant feeding practices. Food Nutr Bull 1983; 6(1): 33-43.

10. Griffiths M, Piwoz E. Manual to assess young child feeding practices in Indonesia. Washington, DC: The Manoff Group, 1987.

11. Griffiths M, Piwoz E, Favin M, Del Rosso J. Improving young child feeding during diarrhea: a guide for investigators and program managers. Arlington, Virginia: PRITECH Project, 1988.

12. Griffiths M. How to improve child well-being? First increase mothers' self-confidence. Development communication report, no. 70/3. Washington, DC: The Manoff Group, 1990.

13. Hollis C, Seumo E, Mal Bappa A, Griffiths M. Improving young child feeding practices in Cameroon: project overview. Washington, DC: The Manoff Group, 1989.

14. Griffiths M. Improving young child feeding practices in Indonesia. project overview. Washington, DC: The Manoff Group, 1991.

8. The use of RAP in the assessment of growth monitoring and promotion in north Sulawesi: Indonesia

The concept of monthly weighing
Differences among villages
Management of growth monitoring

By Mahdin A. Husaini, Ph.D., Satoto, M.D., Ph.D., and Darwin Karyadi, M.D., Ph.D.

M. A. Husaini is affiliated with the Nutrition Research and Development Centre in Bogor, Indonesia. Dr. Satoto is affiliated with the University of Diponegoro in Semarang. D. Karyadi is Professor in Community Nutrition, Bogor Agricultural University, and Director of the Nutrition Research and Development Centre in Bogor, Indonesia.

This paper describes the application of the RAP guidelines by RAP trained nutritionists to a growth monitoring programme in Indonesia. Different from many other RAP studies, there was only a single contact with each household. To be successful the programme required that both health personnel and mothers understand the significance of the growth curve on charts for each child and that mothers received correct and useful advice when their child's growth was faltering. The study combined direct observation at the weighing post and unstructured interviews with parents, weighing post volunteers, health center personnel, and village leaders. The method clearly identified weighing posts that were functioning well, those functioning less well, and those functioning poorly. It provided clear and practical recommendations for improvement. This study reinforces the contention that RAP should be carefully considered when rapid, relatively low cost methods are needed to help inform decision makers on how to improve a health intervention. However, the importance of solid training in RAP remains critical. - Eds.

GROWTH MONITORING HAS been conducted for more than 20 years in Indonesia. In a relatively short time, Indonesia has established an integrated system potentially capable of reaching nearly all of the population with essential primary health care. Monthly weighing of children under the age of five at Posyandu (integrated service post) has been used as an entry point for many interventions such as diarrhoeal control, vitamin A capsule and iron tablet distribution, family planning, and immunization, which represent the essentials of primary health care for children1. Now, in all provinces, every village is close to a Posyandu2.

It would be naive, however, to assume that the system is working well in all provinces in Indonesia. Much needs to be done to improve training of kadres (village volunteers), to insure accuracy in weighing and plotting weight curves onto growth cards, in interpretation of weight curves, and in nutrition counselling. The challenge now is to make growth monitoring more effective and efficient by alleviating the main constraints that hinder utilization of the existing services.

The objective of this assessment is to identify technical and operational changes that can improve the quality of growth monitoring, and to provide recommendations for improvement of the Posyandus' services.


The assessment was conducted by a team consisting of one nutritionist from the Nutrition Research Development Center in Bogor as coordinator, and four nutritionists from the province. The provincial nutritionists were trained for one week in Ciloto, West Java to collect data using RAP methods and to prepare the guidelines. In the district, 92% of children had growth cards, 62% attended the Posyandu, and an average of 42% of target children gained weight in each session. Three Posyandu areas were selected representing the best, the good, and the fair according to local criteria, including coverage, percentage of children who gained weight every month, and regularity of reporting. The best Posyandu was in Lemoh village, a good one in Sukur village, and one in Walian village that was only fair.

Weighing sessions at Posyandus are provided once a month. The team visited each Posyandu and observed all activities. Observations were -made from the beginning to the end of a Posyandu session, which included registration, weighing, plotting of weight onto growth card, interpreting the results (weight gain, weight loss), nutrition counselling, and immunization. The team also noted the registration book containing the list of all under-five children in the area, the names of the children who visited the Posyandu each month, and the children who gained or did not gain weight. It observed the food supplements given at Posyandu to attract the children, and other activities, such as diarrhoea prevention, vitamin A, prophylaxis, distribution of oralit and iron tables, and antenatal services for pregnant mothers.

Focus group discussions (FGD) were conducted with a group of mothers who had children under three years, a group of kadres, and a group of key persons in each village. Each group consisted of six to 12 people. The participants in the FGD for mothers were chosen randomly from those who were visiting the Posyandu that day. The participants in the FGD for kadres were those who were active on that day, including the formally trained and untrained kadres (the ratio of trained and untrained kadres was 1:5). The participants for village key persons, FGD, included the head of sub-villages and hamlets, religious leaders, PKK women (family welfare movement), and members of the LKMD (village community welfare movement). The information gathered during the discussions included knowledge, beliefs, attitudes, and practices in monthly weighing, use of Posyandu services, breast feeding, and infant feeding, and the benefits provided by the Posyandu for people attending it. The focus group session lasted up to an hour and a half, following the technique described by Scrimshaw and Hurtado [1]. Each focus group meeting was guided by two nutritionists, one acting as facilitator and the other as recorder.

Home visits for interviews on socio-economic background, frequency of visits to Posyandu in the previous six months, and mothers' understanding of how to read the weight curve of her child on the growth card were carried out in 77 households in Lemoh village, 77 households in Sukur village, and 87 households in Walian village. At the village level, interviews were conducted with the Lurah (village leader) on all aspects related to Posyandu services.

Interviews were also conducted with the physician (head of Puskesmas = Community Health Centre), and his staff about the role of Puskesmas in Posyandu activities, the health and nutritional status of the population, his perception of and attitudes toward the problems, and the management and effects of the Posyandu on people in the area.

An interview was also conducted at the district level with the head of the Health District Office. At the provincial level, interviews were carried out with the head of MCH Division, and the head of Nutrition Sub-Division at the Provincial Health Office. The topics of the interviews at the provincial and district levels were similar in principle to those at the Puskesmas level, adapted their broader scope and responsibility.

Each day's field observations were followed by a debriefing sessions in which each individual's findings were reported and discussed by those participating in the field work for that day. The major observations and conclusions were written by the team, and the final report of the study was written by the coordinator of the team.

The concept of monthly weighing

Although the concept of Posyandu is community ownership and management for the welfare of the community, the success of the Posyandu depends on involvement of the Puskesmas. In this study, it was observed that when the physician in charge of the Puskesmas motivated community participation (as seen in Lemoh village) the quality of the services was better. Although the influence of agricultural, religious, and social sectors has not been assessed, the role of Puskesmas still needs to be strengthened, since the quality of services at the Posyandus is not entirely satisfactory across the region.

The family planning field workers who have been trained in growth monitoring and promotion are also active in helping kadres, especially in motivating mothers to use contraceptives and providing them with materials for this programme. The use of contraceptives in these villages is over 80%.

The PKK leaders, particularly the wife of the Lurah (village leader) and the Lurah himself also strongly affect Posyandu performance. The PKK organizes a competition among the Posyandus in the villages every year. An evaluation team consisting of PKK, Puskesmas staff, Bangdes (village improvement under the Department of Agriculture) worker, the religious leader, and other key persons, evaluates Posyandu weighing sessions, methods of plotting the card, interpreting the card, counselling, and reporting. The winner receives equipment such as glasses, dishes, pans, etc. for preparing food supplements. The Puskesmas also derives benefits from the Posyandu. The immunization outreach became much easier, the number of children who attend the Puskesmas is significantly reduced, and the Puskesmas personnel do not have to weigh the children because they are already weighed at the Posyandu.

The superior performance of trained kadres compared with untrained kadres was clearly seen during the observation of Posyandu and discussion with kadres. They almost always gave nutrition counselling the most difficult task for kadres (Table 1). The untrained-kadres almost never assumed this role, and the transfer of knowledge from trained and untrained kadres was not successful. To be competent in performing this task, the untrained kadres need to be trained in a formal way by Puskesmas personnel.

If the child has not gained weight, the kadre is supposed to ask whether the child is sick. If the child is sick, she recommends that the child be brought to Puskesmas; if the child is not sick, she advises giving more food, and foods that are more nutritious and diversified. There was great variation in the quality of this advice. Nutritious foods mean including more varieties of vegetables. If the child had gained weight, the kadre advised mothers to give the same foods in greater quantity to promote child growth.

Contents - Previous - Next