Online-only Table 1 RHoMIS applications included in this data set.

From: The Rural Household Multiple Indicator Survey, data from 13,310 farm households in 21 countries

Country (site/region) Lead Institute Year Households surveyed Short description of the sampling strategy
Burkina Faso International Livestock Research Institute (ILRI) 2015 200 Fifty households were randomly sampled from one village in each of the four project regions.
Burkina Faso International Livestock Research Institute (ILRI) 2016 400 Fifty households were randomly sampled from 4 villages in each of the two regions.
Burkina Faso TreeAID 2018 1030 Smallholder producers were selected randomly from 50 villages across Burkina Faso. The villages were selected according to already established relationships with partner organisations.
Burkina Faso TreeAID 2018 1270 Smallholder producers were selected randomly from 60 villages across Burkina Faso. The villages were selected according to already established relationships with partner organisations.
Burundi International Livestock Research Institute (ILRI) - International Institute of Tropical Agriculture 2018 333 Households were interviewed in the regions of Cibitoke and Gitega. In both regions households were randomly sampled from the available village lists obtained from the village elderly.
Cambodia International Center for Tropical Agriculture (CIAT) 2016 631 Respondents were randomly selected from village and commune lists obtained from local authorities by national partners.
Costa Rica Bioversity International 2017 204 Respondents were selected through random selection in the Caribbean humid tropic region of Costa Rica based on lists of producers available at the ministry.
Democratic Republic of Congo World Agroforestry (ICRAF) 2017 413 Households were interviewed along three of the four borders surrounding the Yangambi Biosphere Reserve. The two-stage sampling procedure included purposeful sampling (used to equitably select villages along the borders according to population sizes) and random sampling within villages to select households.
Democratic Republic of Congo International Institute of Tropical Agriculture (IITA) - International Livestock Research Institute (ILRI) 2017 438 Households were interviewed in the northern and southern region of Bukavu. Households were randomly sampled from the available village lists obtained from the village elderly.
Ethiopia TreeAID 2017 302 Households were interviewed in the region north-east of Meki. Subsistence oriented households were randomly sampled from the available village lists obtained from the village elderly.
Ethiopia Bioversity International 2017 249 Households were randomly sampled from beneficiary villages involved in an ongoing research project led by Bioversity International by sampling constant number of smallholder farmers per village.
Ethiopia TreeAID 2018 405 Smallholder producers were selected randomly from 8 villages across the Bosena Werda area in the central highlands in Ethiopia. The villages were selected according to already established relationships with partner organisations.
Ghana TreeAID 2017 223 Informants were selected randomly from 26 villages within the project area, where informants were either project beneficiaries (101 households) or were not beneficiaries – a ‘control group’ (122 households). The villages were selected according to already established relationships with partner organisations.
Ghana TreeAID 2018 391 Smallholder producers were selected randomly from 10 villages across the Navrongo region in northern Ghana. The villages were selected according to already established relationships with partner organisations.
India International Livestock Research Institute (ILRI) 2016 160 In an earlier household survey application in 2012 200 households were randomly sampled from a 10 by 10 km block. In this RHoMIS application 160 households were randomly selected from these original 200.
Kenya (Central-West) International Livestock Research Institute (ILRI) 2016 160 In an earlier household survey application in 2012 200 households were randomly sampled from a 10 by 10 km block. In this RHoMIS application 160 households were randomly selected from these original 200.
Kenya (South) International Livestock Research Institute (ILRI) 2016 160 In an earlier household survey application in 2012 200 households were randomly sampled from a 10 by 10 km block. In this RHoMIS application 160 households were randomly selected from these original 200.
Kenya Bioversity International 2017 316 Households were randomly sampled from beneficiary villages involved in an ongoing research project led by Bioversity International by sampling constant number of smallholder farmers per village.
Kenya World Agroforestry (ICRAF) 2017 385 Households were interviewed in the Kitui and Baringo regions, two regions different in agroecological zone and livelihood strategies. In each region, four divisions were purposefully selected based on population density. Within each division, two sub-locations were purposefully selected to capture economic and geographic heterogeneity. Forty to sixty households were randomly selected to participate in each sublocation.
Kenya International Livestock Research Institute (ILRI) 2017 169 Respondents were selected through a two-stage, stratified cluster random sampling, one at the ward level and the other at the household level for each county. A random sample of wards is selected with probability proportional to the size of the population and number of households in each ward relative to the total population in all the wards in the counties. The wards are selected randomly from a list of all wards in the counties. The counties represent the strata and when a ward is selected, all the households within the ward are also automatically selected, as in cluster sampling. For this study, a total of 169 households were selected as the representative sample from all the six counties in Western Kenya.
Laos International Center for Tropical Agriculture (CIAT) 2016 365 Respondents were randomly selected from village and commune lists obtained from local authorities by national partners.
Malawi International Maize and Wheat Improvement Centre (CIMMYT) 2015 160 The register of farm members of the National Smallholder Farmers Association of Malawi was used to randomly select 80 Households each in the + HHM (households exposed to the Household Methodology; this was in Lilongwe North) and -HHM sites (this was in Lilongwe South), respectively.
Mali International Livestock Research Institute (ILRI) 2015 200 Fifty households were randomly sampled from one village in each of the four project regions.
Mali TreeAID 2017 363 Households were selected randomly from 21 villages within the project area. The villages were selected according to already established relationships with partner organisations.
Nicaragua Bioversity International 2018 145 In the area of La Dalia 145 households were randomly chosen based on existing village lists within the climate smart village run by the CCAFS program.
Peru Grupo Yanapai - Colorado State University 2018 174 Households were interviewed in Huancayo, Acobamba and Ocros. In all three sites households were randomly sampled based on village lists available from census information.
Tanzania International Livestock Research Institute (ILRI) 2015 150 In an earlier household survey application in 2012 200 households were randomly sampled from a 10 by 10 km block. In this RHoMIS application 150 households were randomly selected from these original 200.
Tanzania International Livestock Research Institute (ILRI) 2017 996 Respondents were randomly selected from village and commune lists of cattle owners obtained from animal health service records.
Tanzania Bioversity International 2017 528 Forty-four villages were randomly selected from administrative village lists for data collection (20 villages each in Mtwara and Lindi, and 4 villages in Tunduru). At each village, 12 farming households were randomly sampled from lists provided by local extension officers.
Tanzania World Agroforestry 2018 841 Households were interviewed in the regions of Tabora, Dodoma, Mafinga and Zanzibar. In all four sites households were randomly sampled based on village lists obtained from the village elderly.
Trifinio (Guatemala, El Salvador, Honduras) Bioversity International 2015 300 Across the Trifinio area, 300 households were sampled from lists provided by organizations that provide development services to these communities. Communities were sampled as those where Farmer Field Schools could be organized. From each selected community, 2 households were randomly sampled for the RHoMIS survey, balanced across communities.
Uganda Wageningen University 2017 130 In an earlier household survey application in 2012 200 households were randomly sampled from a 10 by 10 km block. In this RHoMIS application 130 households were randomly selected from these original 200.
Vietnam International Center for Tropical Agriculture (CIAT) 2016 310 Respondents were randomly selected from village and commune lists obtained from local authorities by national partners.
Zambia World Agroforestry (ICRAF) 2017 600 Households were interviewed in the regions of Petauke, Katete and Chipata. In all three sites villages and households were randomly sampled based on village lists obtained from NGO collaborators.