Sampling Procedure
The 2012/13 UNHS sample was designed to allow for reliable estimation of key indicators at the national, rural-urban, regions levels and separately for 10 sub-regions. A two-stage stratified sampling design was used. At the first stage, Enumeration Areas (EAs) were grouped by districts and rural-urban location, then drawn using Probability Proportional to Size (PPS). At the second stage, households which are the Ultimate Sampling Units were drawn using Systematic Random Sampling.
A total of 750 EAs were selected using the 2012 Uganda Population and Housing Census Mapping Frame. These EAs were distributed to the 10 sub-regions in equal proportions with consideration of the rural-urban domains. The 10 sub-regions stratified on the basis of common sociodemographic characteristics were as follows:
1. Kampala (comprising of the entire Kampala City Council Authority (KCCA));
2. Central I (Kalangala, Masaka, Mpigi, Rakai, Sembabule, Wakiso, Lyantonde, Bukomansimbi, Butambala, Gomba, Kalungu and Lwengo);
3. Central II (Kiboga, Luwero, Mubende, Mukono, Nakasongola, Kayunga, Mityana, Nakaseke, Buikwe, Buvuma and Kyankwanzi);
4. East Central (Bugiri, Iganga, Jinja, Kamuli, Mayuge, Kaliro, Namutumba, Buyende and Luuka);
5. Eastern (Busia, Kapchorwa, Katakwi, Kumi, Mbale, Pallisa, Soroti, Tororo, Kaberamaido, Sironko, Amuria, Budaka, Buduuda, Bukedea, Bukwo, Butaleja, Manafwa, Bulambuli, Kibuku, Kween, Namayingo, Ngora and Serere);
6. Mid-Northern (Apac, Gulu, Kitgum, Lira, Pader, Amolatar, Amuru, Dokolo, Oyam, Agago, Alebtong, Kole, Lamwo, Nwoya and Otuke);
7. North-East (comprising the districts of Kotido, Moroto, Nakapiripirit, Abim, Kaabong, Amudat and Napak);
8. West-Nile (comprising the districts of Adjumani, Arua, Moyo, Nebbi, Yumbe, Koboko, Maracha and Zombo);
9. Mid-Western (comprising the districts of Bundibugyo, Hoima, Kabarole, Kasese, Kibaale, Masindi, Kamwenge, Kyenjojo, Buliisa, Kiryandongo, Kyegegwa and Ntoroko);
10. South Western (comprising the districts of Bushenyi, Kabale, Kisoro, Mbarara, Ntungamo, Rukungiri, Kanungu, Ibanda, Isingiro, Kiruhura, Buhweju, Mitooma, Rubirizi and Sheema).
At household level, the target was to interview 10 households per EA. This implied a total sample of 7,500 households.