the determinants of malnutrition and aid in guiding the formative research (see Box 3, following page). The richer the sources available for existing background information, the more the formative research questions can be refined. Assembling the information in a format that facilitates the identification of knowledge gaps in order to define research questions is particularly useful. A practical tool for this purpose, the Infant and Young Child Feeding Practices Matrix, can be found in Annex C. This matrix lists the guiding principles (Annex A) and contrasts them with what is known about actual practices in a given country/setting and with suggestions for feasible new or improved practices. To provide support for the effort/program to address infant and young child feeding, the matrix also summarizes what is known about the motivations for current or recommended changes in behaviors as well as what appear to be the barriers to these ‘new’ practices. This matrix should be used throughout the research process; first, to organize existing information, and later, to add highlights from the research itself. 4 Box 3. How quantitative data can help guide formative research Using several quantitative data sources, the following profiles were assembled: Category and status Nepal Nigeria Nutrition status (%) and mortality ( /1000) Infant mortality (younger than 1 year) 41 96 Diarrheal deaths (children younger than 5 years) 24 13 Stunting (low height-for-age) 49 41 Underweight (low weight-for-age) 39 23 Wasting (low weight-for-height) 13 14 Anemia (children younger than 2 years) 74 76 Infant and young child feeding practices (%) Early breastfeeding initiation 35 32 Exclusive breastfeeding for the first six months 53 13 Breastfeeding at 2 years 95 32 Fed solids minimum number of times 66 41 Fed minimum number of food groups 61 51 Fed per all three recommended practices 47 22 In Nepal, a focus on diarrheal disease and low rates of early initiation of breastfeeding are two infant and young child feeding elements that arise from these statistics. Some general research questions might be: · What are the perceptions and practices related to hand-washing and hygiene in the home in relation to prevention of diarrheal disease in the young child? · What are the infant and young child feeding practices during and after a diarrheal disease episode? · What are the birth traditions preventing early initiation of breastfeeding and how might breastfeeding be begun within the first hour of birth? · When, why, and with what is supplementation occurring during the first six months and what would motivate a delay in supplementation? · What are the infant and young child feeding practices for each age segment, placing priority on consistency of the complementary food for the youngest and amount of food fed to the older ages? Why and at what ages is frequency of feeding a problem? What are the key food diversity problems, focusing especially on animal-source food and fats? In Nigeria, some of the ‘problem’ practices are similar to the Nepal profile (e.g., lack of early breastfeeding initiation), but other practices are quite different (e.g., exclusive breastfeeding). Some general research questions based on this profile might include: · Beginning with the birth of the baby and following women for the first 18 months, when is supplementation occurring, with what, and why? Also, how can supplementation be delayed? · Why are young children not receiving food frequently enough, and is it more prominent at younger or older ages? · What are the key food diversity problems, focusing especially on animal-source foods and micronutrient-rich vegetables? 5 III. Planning the specific formative research A. Choosing methods and research participants Deciding who will participate in the research and which methods to use are important elements of the formative research planning process. The goal is to choose respondents who can provide the most accurate and useful information about infant and young child feeding practices, who or what influences those practices, or who or what needs to be considered in facilitating change in the practices. Box 4 on the following page provides some examples of the kinds of information different respondents can provide. Research participants The most common groups of participants for infant and young child feeding formative research are: · Caregivers (usually divided by the age of their youngest child and/or by several factors from those listed below). · Husbands/Fathers of the young children, depending on their involvement in child rearing. · Grandparents of the young children, particularly grandmothers if living in the same household. · Other members of the community who might be involved in child rearing such as the community health worker or midwife, traditional healer, or school teacher. · Individuals outside the community, such as at the health center, birthing center, store, or market. The more that participant groups and the geographic areas where the research will take place can be further defined by environmental and cultural factors important in infant and young child feeding, the easier it will be to interpret and learn from the findings. When identifying which of these criteria should be applied, ask whether, for example, people with different religious beliefs feed their children differently on a regular basis. If yes, the different religious groups need to be included in the sample. If not, no distinction is needed. Typical factors that aid in defining the research sites include: · Rural or urban location. · Highland, lowland, or coastal area. · Market accessible, market non-accessible. · Ethnic groups. · Religious groups. Finally, segmenting the categories of research participants (caregivers, grandmothers, traditional birth attendants, etc.) according to additional criteria will further aid in interpreting the results of the research. For example, the research might segment caregivers using the following criteria: 6 · Nutritional or health status of the children: Caregivers with well-nourished children, caregivers with malnourished children or sick children. · Age of the youngest child: Families with infants in important age groupings for