Nigeria - National Literacy Survey Nigeria-2009, First Round

Nigeria National Literacy Survey Nigeria-2009

First Round

Other Household Survey [hh/oth]

Though the National Bureau of Statistics generates youth and adult literacy data regularly on annual basis, the survey was conducted with a wider scope to complement the existing data on literacy in Nigeria.

The main purpose of the survey was to determine the magnitude, levels and distribution of adult literacy and obtain comprehensive data and information with a view identifying issues of concern, which need to be addressed in the promotion of adult literacy in Nigeria.

Underlying this is the fact that literacy is fundamental to information dissemination, socio-economic development and poverty alleviation among others.

It was the first attempt to carry out a stand alone survey on Literacy Survey Nigeria.

ID Number NGA-NBS-LIT-2009-v1.0

Household Socio Demographic Background:

For all members of household

Educational Attainment:

For persons currently attending school and

For persons that attended school in the past

Literacy in English:

For persons ever attended school

For persons who had never attended school

Literacy in any other language:

For persons ever attended school

For persons who had never attended school

Knowledge and Accessibility of Literacy Programme:

For persons aged 15 years and above

Use & application of reading skills

Topic	Vocabulary	URI
consumption/consumer behaviour [1.1]	CESSDA	http://www.nesstar.org/rdf/common
economic conditions and indicators [1.2]	CESSDA	http://www.nesstar.org/rdf/common
income, property and investment/saving [1.5]	CESSDA	http://www.nesstar.org/rdf/common
rural economics [1.6]	CESSDA	http://www.nesstar.org/rdf/common
employment [3.1]	CESSDA	http://www.nesstar.org/rdf/common
basic skills education [6.1]	CESSDA	http://www.nesstar.org/rdf/common
compulsory and pre-school education [6.2]	CESSDA	http://www.nesstar.org/rdf/common
post-compulsory education [6.5]	CESSDA	http://www.nesstar.org/rdf/common
general health [8.4]	CESSDA	http://www.nesstar.org/rdf/common
health care and medical treatment [8.5]	CESSDA	http://www.nesstar.org/rdf/common
specific diseases and medical conditions [8.9]	CESSDA	http://www.nesstar.org/rdf/common
housing [10.1]	CESSDA	http://www.nesstar.org/rdf/common
children [12.1]	CESSDA	http://www.nesstar.org/rdf/common
elderly [12.2]	CESSDA	http://www.nesstar.org/rdf/common
family life and marriage [12.5]	CESSDA	http://www.nesstar.org/rdf/common
gender and gender roles [12.6]	CESSDA	http://www.nesstar.org/rdf/common
community, urban and rural life [13.1]	CESSDA	http://www.nesstar.org/rdf/common
social behaviour and attitudes [13.6]	CESSDA	http://www.nesstar.org/rdf/common
social conditions and indicators [13.8]	CESSDA	http://www.nesstar.org/rdf/common
fertility [14.2]	CESSDA	http://www.nesstar.org/rdf/common
morbidity and mortality [14.4]	CESSDA	http://www.nesstar.org/rdf/common

The NISH is a multi purpose on – going programme of household surveys

NISH Master Sample (MS) is derived from frame of EAs demarcated by the National Population Commission (NPopC) for the 2006 Housing and Population Census

The NISH MS is made up of 200 EAs drawn in 20 independent Replicates

Replicate made up of 10 EAs Replicates 7 to 10 will be used for the survey

The survey will cover all the 36 states and Federal Capital Territory (FCT). Both urban and rural areas will be canvassed

2.1 Sample Design

2.1.1 Introduction of NISH Design 1993/99

The Multiple Indicator Cluster Survey (MICS) 1999 was run as a module of the National Integrated Survey of Households (NISH) design. NISH is the Nigerian version of the United Nations National Household Survey Capability Programme and is a multi-subject household based survey system. It is an ongoing programme of household based surveys enquiring into various aspects of households, including housing, health, education and employment. The programme started in 1981 after a pilot study in 1980. The design utilizes a probability sample drawn using a random sampling method at the national and sub-national levels.

The main features of the NISH design are:

Multi-Phase Sampling: In each state 800 EAs were selected with equal probability as first phase samples. A second phase sample of 200 EAs was selected with probability proportional to size.

Multi-Stage Sampling Design: A two-stage design was used. Enumeration Areas were used as the first stage sampling units and Housing Units (HUs) as the second stage sampling units.

Replicated Rotatable Design: Two hundred EAs were selected in each state in 10 independent replicates of 20 EAs per replicate. A rotation was imposed which ensured 6 replicates to be studied each survey year but in subsequent year a replicate is dropped for a new one, that is, a rotation of 1/6 was applied. This means in a survey year, 120 EAs will be covered in each state. In the Federal Capital Territory (Abuja), 60 EAs are covered.

Master Sample: The EAs and HUs selected constitute the Master Sample and subsets were taken for various surveys depending on the nature of the survey and the sample size desired. In any one-year, the 120 EAs are randomly allocated to the 12 months of the year for the survey. The General Household Survey (GHS) is the core module of NISH. Thus, every month 10 EAs are covered for the GHS. For other supplemental modules of NISH, subsets of the master sample are used. The MICS 1999 was run as a module of NISH.

2.1.2 Sample Size

The global MICS design anticipated a sample of 300-500 households per district (domain). This was based on the assumption of a cluster design with design effect of about 2, an average household size of 6, children below the age of 5 years constituting 15 percent of the population and a diarrhoea prevalence of 25 percent. Such a sample would give estimates with an error margin of about 0.1 at the district level. Such a sample would usually come from about 10 clusters of 40 to 50 households per cluster.

In Nigeria, the parameters are similar to the scenario described above. Average household size varied from 3.0 to 5.6 among the states, with a national average of about 5.5. Similarly, children below 5 years constituted between 15-16 percent of total population. Diarrhoea prevalence had been estimated at about 15 percent. These figures have led to sample sizes of between 450 and 660 for each state.

It was decided that a uniform sample of 600 households per state be chosen for the survey. Although non-response, estimated at about 5 percent from previous surveys reduced the sample further, most states had 550 or more households. The MICS sample was drawn from the National Master Sample for the 1998/99 NISH programme implemented by the Federal Office of Statistics (FOS).

The sample was drawn from 30 EAs in each state with a sub-sample of 20 households selected per EA. The design was more efficient than the global MICS design which anticipated a cluster sub-sample size of 40-50 households per cluster. Usually, when the sub-sample size was reduced by half and the number of clusters doubled, a reduction of at least 20 percent in the design effect was achieved. This was derived from DEFF = 1 + (m-1) rho where m is sub-sample size and rho is intra-class correlation. Therefore, the design effect for the Nigerian MICS was about 1.6 instead of 2. This means that for the same size of 600 households, the error margin was reduced by about 10 percent, but where the sample was less than 600 the expected error margin would be achieved.

It should be noted that sampling was based on the former 30 states plus a Federal Capital Territory administrative structure [there are now 36 states and a Federal Capital Territory].

2.1.3 Selection of Households

The global design anticipated either the segmenting of clusters into small areas of approximate 40-45 households and randomly selecting one so that all households within such area was covered or using the random walk procedure in the cluster to select the 40-45 households. Neither of the two procedures was employed. For the segmentation method, it was not difficult to see that the clustering effect could be increased, since, in general, the smaller the cluster the greater the design effect. With such a system, DEFF would be higher than 2, even if minimally. The random walk method, on the other hand, could be affected by enumerator bias, which would be difficult to control and not easily measurable.

For NISH surveys, the listing of all housing units in the selected EAs was first carried out to provide a frame for the sub-sampling. Systematic random sampling was thereafter used to select the sample of housing units. The GHS used a sub-sample of 10 housing units but since the MICS required 20 households, another supplementary sample of 10 housing units was selected and added to the GHS sample. All households in the sample housing units were interviewed, as previous surveys have shown that a housing unit generally contained one household.

Name	Abbreviation	Affiliation	Role
National Bureau of Statistics	NBS	Federal Government of Nigeria (FGN)	Metadata Producer

2010-03-08 Version 1.0 (March 2010)

DDI-NGA-NBS-LIT-2009-v1.0