|Polygamy [message #1853]
|Fri, 04 April 2014 10:20
Registered: April 2014
Polygamy is fairly prevalent in many of the countries in which the DHS operates, but there does not appear to be any specific mention of how this situation is treated by DHS in the Sampling and Household Listing Manual.
For instance, in Uganda and Zimbabwe (where DHS has conducted surveys), extended polygamous families often live in "compounds" which consist of one or more related huts that are located in close proximity to each other within the compound and that are circumscribed by a physical boundary such as a fence.
In such cases where multiple wives of one husband live within these compounds, most often (but not always) wives live in their own huts with their own offspring (and possibly other members of the extended family) while the husband may rotate the hut in which he lives and eats over time. In most instances (but not always), wives within a compound cook separately for their own offspring or other extended family members using different pots than the other wives within the compound.
Given that the definition for "household" used by DHS relates to "living under the same roof and eating from the same pot", the above scenario introduces some complexity at the second stage of sampling where households are typically sampled, and it is not clear how this is dealt with even in the section within the DHS manual on "multi-unit structures".
For the second stage of sampling, one potential strategy could be to sample compounds rather than the individual huts within the compounds, since individual huts may or may not represent households units (according to the given definition) and it is only possible to determine household composition upon completing a roster for the compound.
Such compounds could be treated the same way "multi-unit structures" are typically treated where, after the roster for the compound is completed, one or more "household" (which might consist of multiple huts within the compound) could be selected for sampling.
An additional complexity that is introduced in relation to sampling compounds (instead of huts) within EAs at the second stage of sampling is that it is not clear (even after speaking with the census bureaus of these countries) whether the "household" counts in polygamous zones of the country provided by the census office actually relate to compounds or relate to huts for each EA. The census offices claim that these are "household counts" but it is not clear within polygamous zones whether it is compounds, huts or households (which could be groups of huts) that are being referred to.
This ambiguity creates the necessity for surveys to conduct listing exercises for sampled EAs - where they might have otherwise been spared from this additional burden if census EA counts on compounds were available.
The above text discusses one suggested way of treating the issue of polygamy at the second stage of sampling; however, there are a number of possible ways. Regardless, whichever way the DHS chooses to treat the issue should be well documented so that users will have a full understanding of the design and so that organizations who choose to undertake smaller scale regional surveys within such countries will draw on the expertise of the DHS and follow suit by using a standardized methodology.
Scientist, Survey Methods Advisor
Food and Nutrition Technical Assistance III Project (FANTA)