Nigeria - Nigeria Living Standard Survey 2003, First round

Headquarters Training of Trainers (T0T) The first level of training at the headquarter consisted of three categories of officers, namely, the trainers at the zonal level, fieldwork monitoring officers and data processing officers who were crucial to the successful implementation of the survey.
The intensive and extensive training lasted for five days. Zonal Level Training The training took place in the six zonal FOS [now NBS] offices representing the six geo-political zones of the country. These are Ibadan (South West) Enugu (South East), Calabar (South South), Jos (North Central), Maiduguri (North East) and Kaduna (North West).
The composition of the team from each State to the six different zones were the State officer, one scrutiny officer and two field officers, making four persons per state. Two resource persons from the headquarters did the training with the zonal controllers participating and contributing during the five-day regimented and intensive training. State Level Training
The third level training was at the State level. A total of 40 officers were trained, comprising 20 enumerators, 10 editing staff and 10 supervisors.
The State Statistical Agencies, as a matter policy, contributed 5-10 enumerators. The ten-day exercise was also regimented, intensive and extensive because the enumerators were also crucial for effective implementation of data collection.

There were five levels of computer edits before analysis took place. This was critical to ensuring the quality and acceptability of the data. Level 1: Control Edits:
These were to ensure the sample integrity. The total households captured must match with master sample list. Level 2: Inter-Questionnaire Structure:
These were required in order to compute the Standard of Living (SOL), quintile distribution or compute per capita value. Mismatches and duplicates were reconciled.
Level 3: Intra-Questionnaire: This was required for sectoral analysis. Information from the roster (age and sex) was matched with respective sections in the questionnaire. Since the household roster was the primary source for computing the universe of subsequent sections, these had to be consistent. Mismatches and duplicates of household members’ identification were rectified. Level 4: Edits: These checks monitor the intra-record consistency. It was important that logical responses and skip patterns were followed. Level 5: General Edits: This checked for outliers and corrections were made through static or dynamic imputation.
Data Analysis The Staff of Computer Management and Information Services (CMIS) of the NBS carried out the data entry of the edited questionnaire and ran programmes to further detect inconsistencies and other related errors as part of the final editing. Tables were then generated from the analyses. Also at the request of the then Federal Office of Statistics, under the British Council Economic Management Capacity Building (EMCAP) Project, a DFID Consultant came to Nigeria to provide technical assistance in the evaluation of dataset.