Nigeria - National Nutrition and Health Survey 2015, Third Round

The National Nutrition and Health Survey using SMART methods is designed as a cross-sectional household survey using a two stage cluster sampling to provide results representative at the state level.
Data were collected from a total of 25,210 households, 20,060 children under-five years of age and 23,688 women of reproductive age.
The sample for the 2015 NNHS is nationally representative and covers the entire population residing in non-institutional dwelling units in the country. The survey uses the national sample frame, which is a list of Enumeration Areas (EAs) prepared for the 2006 Population Census. Administratively Nigeria is divided into states, Local Government Areas (LGAs), and localities. In addition to these administrative units, during the 2006 population census, each locality was subdivided into census Enumeration Areas (EAs). The primary sampling unit (PSU), referred to as a cluster in this survey, is defined on the basis of EAs from the 2006 EA census frame.

The 2015 NNHS sample has been selected using a two-stage cluster design as described below.

-First stage sampling procedure: cluster selection
The PSU (clusters) for each state were randomly selected from the national master sample frame according to the probability proportional to size (PPS) method with the support from National Population Commission.

-Second stage sampling procedure: household selection
The second stage of sampling consisted of selecting households within each cluster by using systematic random selection. The team leader was responsible for the determination of the total number of households in the cluster by conducting a household listing through detailed enumeration of the selected cluster with a support from the community leader. This served as the sampling frame for the selection of households. The team leader then entered the total number of households onto the tablet and sampling interval was automatically calculated and displayed on the tablet. The sampling interval is calculated by dividing the total number of households in the cluster by the number of households to be interviewed. A random number table was used to randomly select a start number, between 1 and the sampling interval, to identify the first household. The sampling interval was used to identify all subsequent households to be included in the survey.
been estimated at zonal level, by pooling the data from the survey domain within each zone.


No Deviation.


The target was to interview 26,048 households across the country. The set target was notreached and overall 25,210 households were interviewed (97 percent).


The data was weighted using sampwgt.