Nigeria - National Nutrition and Health Survey 2014, Second round
Reference ID | NGA-NBS-NNHS-2014-v1.0 |
Year | 2014 |
Country | Nigeria |
Producer(s) | National Bureau of Statistics (NBS) - Federal Government of Nigeria (FGN) |
Sponsor(s) | United Nations Children’s Fund - UNICEF - Funding Micronutrient Initiative - MI - Funding Nigeria's Saving One Million Lives Initiative - - Funding United State Agency International Development - USAID - Funding UKAID - - Funding |
Metadata | Documentation in PDF Download DDI Download RDF |
Created on | Dec 21, 2016 |
Last modified | Dec 21, 2016 |
Page views | 183721 |
Downloads | 13926 |
Sampling
Sampling Procedure
The National nutrition and health survey used Standardised Monitoring and Assessment of Relief and Transitions (SMART) methods.
Data were collected from a total of 25,567 households, 20,939 children under-five years of age and 23,942 women of reproductive age.
The 36 states and Federal Capital Territory (FCT) constitute the domains of the survey. The domains used by MICS and DHS are similar, which allows comparison of results, the only exception being the state of Borno, where 9 Local Governmental Areas (LGA) were excluded for security reasons. Therefore, results for Borno are not representative of the whole state.
It is a cross-sectional household survey using a two stage cluster sampling representative at the state level.
At first stage, clusters were drawn randomly and independently for each survey domain from the national master sample frame with the support from National Population Commission according to the probability proportional to size (PPS) method.
The second stage of sampling consists of selecting households within each cluster by using systematic random selection. The team leader verified the population and/or number of households in the cluster by updating the cluster household listing form through detailed enumeration with a support from the village chief or community leader. With total number of households, the team leader calculated the sampling interval and drew a random start number using random number table. Within each selected household, the head of household or next adult was interviewed and all women and children were measured.
In clusters with more than 250 households, segmentation was used to divide the cluster into areas of equal number of households. One segment was randomly chosen, the second stage of sampling was completed for the segment and all selected households were interviewed.
In order to be able to estimate most of the indicators with reasonable precision, the sample size for the survey is calculated using a prevalence of Global Acute Malnutrition (GAM), based on children age 6-59 months. Indicators with narrow age range; 0-23, 6-23 and 12-23 months will be estimated with reasonable precision for each state. However, indicators with narrower age group such as 0-5, 12-15, 20-23 months and very low prevalence, such as treatment of children with ARI and Malaria, will be estimated at zonal level by pooling the data from the survey domain within each zone.
The sample size for the survey was based on sample calculation for the prevalence of Global Acute Malnutrition (GAM) in children of age 6-59 months. The indicators with age ranges of one year or more; 0-23, 6-23 and 12-23 months were found to have reasonable precision for state level estimates.
Those indicators with narrower age ranges such as 0-5, 12-15, 20-23 months and very low prevalence such as treatment of children with ARI and malaria are estimated only at zonal level by aggregating the state level data within each zone.
Significantly different health and demographic conditions are found across Nigeria. In general, the southern half of the country has smaller family sizes and better health and nutrition conditions. These differences were accounted for in two separate sample calculations (for Northern and Southern states), thus two different sample sizes were used to achieve similar level of precision at a national level.
Deviations from Sample Design
No Deviation
Response Rate
Overall 23,942 women and 20,939 children were interviewed. The response rate was 100%.
Weighting
Survey weights were calculated based on populations provided from the master sample frame and number of valid cases. The state level results were self-weighted as per the sample design. The national results were weighted by the survey weights. Three sets of survey weights were used for household, woman level, and child level results, respectively.