Nigeria - Multiple Indicator Cluster Survey MICS3 (2007), Nigeria, Third round
Reference ID | NGA-NBS-MICS3 2007-v1.2 |
Year | 2007 |
Country | Nigeria |
Producer(s) | National Bureau of Statistics [nbs] - Federal Government of Nigeria |
Sponsor(s) | Fedral Government of Nigeria - FG - Funding United Nation Children Educational Fund - UNICEF - Funding National Bureau of Statistics - NBS - Funding |
Metadata | Download DDI Download RDF |
Created on | Oct 18, 2010 |
Last modified | Dec 02, 2013 |
Page views | 645869 |
Downloads | 37271 |
Sampling
Sampling Procedure
Two - stage cluster sample design was adopted in each state where Enumeration Areas (EAs) form first stage or Primary Sampling Units (PSUs) and Housing Units (HUs) form second stage or Ultimate Sampling Units (USUs)
EAs demarcated for 1991 Population Census served as first stage sampling frame
Household listing was conducted in selected first stage units to provide second stage sampling frame
Sample sizes: Within each state of the federation 750 HUs was drawn from 30 EAs.
There were 36 states and Federal Capital Territory (FCT), this makes 37, which amounts to 27,750 Housing Units drawn from 1,110 EAs.
The sample for the Nigeria MICS3 was designed to provide estimates on a large number of indicators on the
situation of children and women at the country level, for urban and rural areas; and for each of the 36 States
of the Federation and the Federal Capital Territory of Abuja. The States were the main reporting domains.
The sample design was two-stage in each state, where a systematic sample of 30 census enumeration areas
(EAs) was selected with equal probability to form the first stage or primary sampling units (PSUs). The
updated 1991 Population Census Enumeration Area demarcation was used because the latest demarcation
was not available for use at the time MICS3 sample was designed. Also, information about the household
composition of enumeration areas was not available to permit selection of EAs with probability proportional
to number of households in the enumeration area.
Household listing was conducted in each of the selected EAs to provide an adequate, up-to-date frame of
housing units as the secondary sampling units (SSUs). A systematic sample of 25 housing units was
subsequently drawn with equal probability within each of the selected EAs and all the households in each of
the selected HUs were canvassed. Thus, at state level, 750 HUs were drawn from 30 EAs which meant
27,750 HUs from 1,110 EAs at the national level. The sample was stratified by states and was hardly self
weighting at either state or national level. Hence, sample weights were used for reporting state or national
results.
All the selected enumeration areas were successfully canvassed. Table HH.1 presents a summary of results
of interviews of households, individual women aged 15 – 49 years and children aged less than five years. A
total of 28,603 households (20,825 rural and 7,778 in the urban sectors) were sampled. The total number of
occupied sampled households was 28,431 including 20,735 rural and 7,696 urban households. The total
number of interviewed households was 26,735 including 19,569 rural and 7,166 urban households. These
figures translated into 94.0 percent response rates for the total, 94.4 percent for the rural and 93.1 percent for
the urban. The total number of eligible women was 27,093 with 19,674 and 7,419 for rural and urban sectors,
respectively. The corresponding figures of interviewed women were 24,565, 17,928, and 6,637 respectively;
these figures amounted to 85.3, 86.0 and 83.3 percent effective response rates respectively for the total, rural
and urban sectors. Eligible children under-five years of age were 17,093, (12,898 rural and 4,195 urban) and
interviews were achieved for 16,549, 12,494 and 4,055 respectively; again the corresponding effective
response rates were 91.0, 91.4 and 90.0 percent respectively.
Deviations from Sample Design
There were no deviation from sample Designed
Response Rate
We had 96% Response Rate
Table HH.1 presents a summary of results of interviews of households; individual women aged 15 –
49 years and in respect of children aged under-five years. A total of 28,603 households including
20,825 and 7,778 in the rural and urban sectors respectively were sampled; total number of
occupied sampled households was 28,431 including 20,735 rural and 7,696 urban households. Total
number of interviewed households was 26,735 including 19,569 rural and 7,166 urban households.
These figures translated into 94.0 percent response rates for the total, 94.4 percent for the rural and
93.1 percent for the urban. Total figure of eligible women was 27,093 including 19,674 and 7,419
for rural and urban sectors respectively while corresponding figures of interviewed women were
24,565, 17,928, and 6,637 respectively; these figures translated into 85.3, 86.0 and 83.3 effective
response rates respectively. Numbers of eligible under-five children were 17,093, 12,898 and 4,195
and interview was completed for 16,549, 12,494 and 4,055 respectively; again the corresponding
overall response rates were 91.0, 91.4 and 90.0 percent respectively. Urban-rural disparities in
response rates were quite marginal.
Table HH.1: Results of household and individual interviews
Numbers of households, women and children under 5 by results of the household, women's and under-five's interviews, and household,
women's and under-five's response rates, Nigeria, 2007
Households’ response rates varied from 81 percent in Osun State to 100 percent in Katsina State;
but the variations have been bridged across geopolitical zonal aggregates although the northern
zones show greater household response rates. This pattern of variation is true also of women and
under-five children response rates respectively. No immediate explanations could be adduced for
these differentials beyond the fact that the less educated North is ever more prepared to cooperate
with the interviewer and that the terrain in the North is friendlier for purposes of interviewing.
Detailed information attached as external document
Weighting
Sample weights were calculated for each of the data files.. Sample weights for the household data were computed as the probability of selection of the household, computed at the sampling domain level (urban/rural within each state). The household weights were adjusted for non-response at the domain level, and then nomalised by a constant factor so that the total weighted number of households equals the total unweighted number of households. The hosehold weight variable is called HHWEIGHT and is used with the HH data and the HL data
Sample weights for the women's data used the un-nomalized household weights, adjusted for non-response for the women's questionnaire, and were then normalized by a constant factor so that the total weighted number of women's cases equals the total unweighted number of women's cases.
Sample weights for the children's data followed the same approach as the women's and used the un-nomalized household weights, adjusted for non-response fr the children's questionnaire, and were then normalized by a constant factor so that the total weighted number of children's cases equals the total unweighted number of children's cases
Estimation Procedures:
Let the probability of selecting the EA be fj and the probability of selecting the housing unit be fk. Then the product f = fjfk = 1 where fj = n and fk = h
Ys = Estimate for states
N = Total Number of EAs in states
n = Selected number of EAs in states
H = Total number of Housing Units listed in the jth EA
h = Selected number of Housing Units in the jth EA.
Xsj k = Value of the element in the kth housing unit of jth EA in states.
Wsjk = Weight of the element in kth housing unit of the jth EA in states.