This is a good point. There are populations in which (e.g.) a DDS of 6 might be high and others where it might be low. There is also seasonality to consider. Fruits (e.g.) come in and out of season in most places. I think that you can use a normative approach to identifying risk. One approach would be to estimate the location of the 25th percentile (Q1) and label those HHs at of below this level as being "at-risk". This approach is no good for between population comparisons as all populations will have similar proportions at or below Q1 but is should work within populations. I hope this is of some help.

I think what you are saying is that low dietary diversity in pastoralist communities as indicated by HDDS isn't necessarily reflective of the true food security situation in those communities. Mark has already commented that this is most likely the case and has offered an alternative way of analysing the HDDS score to show "at-risk". Another possible approach around this is instead of using dietary diversity (as indicated by HDDS) as a proxy for food security, you might consider using the Household Hunger Scale or HHS (see [url]http://www.fantaproject.org/publications/hhs_2011.shtml[/url]) as an alternative. Reflecting on what we know about pastoralist communities, I think that if you were to detect even just a modest proportion of moderate hunger in these communities, this can already be indicating a considerable situation of food insecurity.

The use of HHS as a food security indicator is a good idea. It uses a simple and validated question set. I think that you would still want information on diet and dietary diversity. I have done this in Somali pastoralists using a rapid semi-quantitive method (in-depth interviews, informal group discussions, some stratification be wealth rank, multiple data sources). Below is an example report section to give you a flavour of the sort of data that this sort of approach can yield. This is a quick approach. --- EXAMPLE --- There are usually two or three meals per day. The number of meals taken depends upon household duties, food availability, and the presence of the mother at the homestead. Older male children and men in pastoral and agro-pastoral communities tend to be away from home during the day and will usually take only the morning and evening meals. Men, women, and children in riverine communities may, depending on the agricultural cycle, be away from home during the day and take only the morning and evening meals. Examples of typical family meals (by 24 hour recall) are shown in Table 18. Tables 18 : Three examples of typical family meals Morning Sorghum porridge, tea with sugar Daytime Sorghum porridge with sugar (mother and children only) Evening Sorghum porridge with yoghurt and sugar Morning Sorghum porridge with vegetable sauce, tea with sugar Daytime Sorghum porridge with sugar (mother and children only) Evening Boiled whole-grain sorghum with beans, oil, and sugar Morning Tea with sugar and milk Daytime Boiled whole-grain sorghum with beans and sesame oil Evening Sorghum porridge with milk Wealthier families may have pancakes (a batter of wheat flour, sugar, and water fried in oil) taken with tea with sugar for the morning meal. Milk tends to be less available during the dry season. When milk is not available a sauce of onions and tomatoes, occasionally with pumpkin, aubergine, or sweet pepper is used as a milk substitute. Beans are usually available close to harvest time. Pastoralists purchase vegetables directly from Riverine communities, in village markets, or the market in Belete Weyne town. Vegetable are seen as a poor substitute for milk in pastoral communities. Vegetables are seldom eaten in pastoralist and agro-pastoralist communities. One focus group of children reported that they had last eaten vegetables (onion and tomato sauce with rice) one month before data collection. Fruit was almost never eaten. Meat (mutton) had last been eaten ten days before data collection when a sheep had been slaughtered to signify the reconciliation of a family conflict. This group reported that they did not eat wild foods. This group also reported that boys were given more or better foods than girls but that this was an occasional occurrence. These children were occasionally given sweets (when their parents went to town). Sweets were considered a treat and not food. Diets in the Riverine zone follow the same basic pattern. Maize is usually available as an off-season (i.e. dry season) crop and tends to be eaten roasted on the cob or milled and eaten as a stiff porridge. Vegetables (onions, tomatoes, pumpkin, aubergine, sweet pepper) and fruit (lemon, grapefruit, watermelon, mango, and papaya) are grown as cash crops by a minority of households. These market-gardening activities tend to be led by merchants who pay for inputs (i.e. seeds, fertiliser, pesticides, fuel for irrigation pumps, labour, and transportation) and provide participating households a share of profits instead of a fixed payment for rental of land. Profits are reported to be too low for smaller landowners who concentrate on growing sorghum, maize, beans, and sesame (as a cash crop). Poorer households have little access to home grown vegetables and fruit. The quantity of vegetables grown decreases with distance from markets and only a small minority of household use vegetables from kitchen gardens in the more distant communities. This lack of access to vegetables in riverine communities is likely, in part, to be due to the lack of access to land associated with the destruction of irrigation infrastructure caused by flooding. Access to the machinery (e.g. bulldozers) required for large-scale rehabilitation of the irrigation infrastructure is difficult as is access to essential supplies (e.g. pumps, pipes, seals, filters, spare parts and fuel) required to operate a rehabilitated irrigation system at full capacity. Snacks are not common in pastoral and agro-pastoral communities. In riverine communities roast maize cobs and sorghum stalk (similar to sugar cane) are prepared as snacks. Wild food consists mainly of game and is occasionally consumed in all livelihood zones. This is usually dik-dik and ibis. Rabbit is occasionally taken particularly during illness. The consumption of wild food is subject to an array of traditional and religious restrictions. Wild fruits are also consumed when available as are resinous gums (as savoury chewing gum) from the sap of trees. Collection of resinous gums (e.g. frankincense) for sale provides a small income source for pastoral communities. Village markets in pastoral areas usually had one or more traders buying resinous gums for onward sale. --- END --- I hope this is useful.