SAMPLE DESIGN & METHODOLOGY

       The basic parameters of the survey design are listed below.

• The scope of the research was national, covering the 36 States and the federal capital territory (FCT) Abuja. A robust sample was designed and executed with a goal of achieving a national representative survey.
• The sample design was supplied by the National Bureau of Statistics (NBS) and was based on the National Integrated Survey of Households (NISH) Master Sample designed for a five year period (2007/2012).
• The survey utilized a multistage random sampling design:
  
  1st stage: The Enumeration Areas (EAs) as demarcated for the 2006 population census by the National Population Commission (NPopC,) provided the first stage sampling frame for the survey.  The total number of Enumeration Areas (EAs) in the whole country including FCT, Abuja is 662,318.  Thirty EAs were scientifically selected with equal probability from each state of the Federation and FCT, Abuja; making a total of 1,110 EAs.
  
  2nd stage: After the selection of the enumeration areas, NBS embarked on a household listing exercise of all the EAs selected in each state. The key objectives of the listing exercise were to: identify and list all housing units and households within the sample EAs; ascertain the correctness of the boundary descriptions of the EAs; and provide useful information in calculating weights/raising factors.  After the completion of the listing exercise, 20 housing units were systematically selected in each EA.  If there were less than 20 HUs listed in any EA, then all were covered for the survey.
  
  3rd stage: Out of 20 households selected in each EA, information on names and ages of all (a) qualifying male 18 years and above, (b) qualifying females 18 years and above and (c) thirdly non qualifying male and females were collected.  For a household member to qualify for the survey, he/she must be aged at least 18 years old and must be available for the interview. A person is considered ineligible if he/she is mentally disabled, is aged less than 18 years or is reported to be unavailable for a period of three days from the date of first visit. The final respondent was selected with the help of a KISH grid selection chart. This aided in minimising interviewer bias in selection of respondents.
  
  
• Research was conducted by Research & Marketing Services Ltd (RMS).
• Target sample size canvassed was 22,200 respondents.
• A total sample size of 21,110 valid interviews was achieved which is equivalent to a response rate of 95%.
• The questionnaire was translated into Hausa, Yoruba, Igbo, and Pidgin English.
• The average interview duration across all the regions was 1 hour.
• Fieldwork was conducted from May to August 2008.
• A number of measures such as the recruitment of external supervisors to monitor fieldwork operations were implemented to ensure the integrity of the data collection.
• Weighting of the data: The weighting was carried out by the National Bureau of Statistics (NBS) and was based on the 2006 National Population Figures.  The data was first weighted by using the inverse probability of selection weights.  These weights were calculated at enumeration area, household and individual levels; and were then adjusted for the probability of selection and non-response.  The final data set includes three levels of weighting – total population, household and adult population (18 years+). 
• The selected sample size is statistically robust with a margin of error of ± 5% at a 95% confidence level.

| UNIQUENESS |SAMPLE DESIGN & METHODOLOGY | QUESTIONNAIRE | RESULTS | DATA REQUEST | |SEGMENTATION ANALYSES|