National Population Censuses (1991/2006)
1.1 1991 Census and PES Data Processing:- The data capturing and the initial processing of the 1991 Census data took place at the seven (7) Data Processing Centres (DPCs) of the Commission, located in Kano, Yola, Port-Harcourt, Enugu, Lagos, Ibadan and Kaduna. The final processing however took place at the Lagos Data Processing Centre. The Commission Headquarters during the 1991 Census was in Lagos and the DPC, at that period, also serves as the Data Processing Headquarters.
The final processing of census ’91 entailed the followings:-
- Data were sent to the Headquarters from each of the DPCs in Tape Streamers.
- Data from each of the DPCs were collated at the Headquarters.
- The collated data were aggregated into Localities, LGAs, States and National.
- The collated data are further edited and validated.
- Tables were then produced from the aggregated data.
The production of the Analytical Report of Census ’91 was carried out by the Data Analysis Team (DAT). The team, comprising of staff from IT and Census Departments, was headed by a United Nations Chief Technical Adviser.
The Provisional Result was released based on 593 LGAs, however with creation of additional LGAs, and adequate input from the Cartography Department, the affected LGAs have to be broken up by their corresponding EAs accordingly into 774 LGAs for the release of the Final Result, thereby incorporating the newly created States and LGAs.
The results on the above activities were thereafter integrated into a Central System which, at that time, served as the Central Database for the Census ’91 data for security reasons. The Back-ups were done on Tape Streamers and Optical Disks for the data, and on Floopy Disks (3.5/5.25)/Optical Disks for the Statistical Tables. Copies of these backups are presently available in the ITD Systems Library in Lagos.
1.2 The Post Enumeration Survey (PES) Data:- The final processing of the PES data was exclusively carried out by the Data Analysis Team. Though the initial processing, including the machine data capturing were done at the DPCs.
1.3 Census 2006 Processing: The Technology and Methodology:-
Unlike the data capture method used for the country’s previous censuses, where information from the census forms are typed into the computer system, data capture for census 2006 was carried out by OMR/OCR/ICR systems where questionnaires are scanned through high speed optical scanners. The choice of the scanning system was because it is faster and more accurate than the data keying method.
Definition of terms
OMR (Optical Mark Recognition) - This means the ability of the scanning machine to detect pencil marks made on the questionnaires by the Enumerators in accordance with the responses given by the respondents.
OCR (Optical Character Recognition) - This means the ability of the scanning machine to recognize machine printed characters on the questionnaires.
ICR (Intelligent Character Recognition) - This means the ability of the scanner to recognize characters hand written by the Enumerators in accordance with the responses given by the respondents.
1.4 Processing Procedures of Census 2006 at the DPCs:- Data processing took place in the Commission’s seven (7) Data Processing Centres located in different geographical zones in the country. There was absolute uniformity in the processing procedures in the seven DPCs.
1.5(a) Questionnaire Retrieval/Archiving
Questionnaires from the fields were taken directly from the Local Government Areas to designated DPCs. The forms on arrival at the DPCs were counted, archived and labeled.
Retrieval of the questionnaires at the DPCs were carried out based on the EA frame received from the Cartography Department. Necessary Transmittal Forms are completed on receipt of the Forms at the DPCs. The Transmittal Forms are also used to keep track of questionnaires movement within the DPC.
1.5(b) Forms Preparation
The scanning machine has been designed to handle A4 size paper. And the Census form being twice that size has to be split into two through the dotted lines at the middle of the form. This forms preparation procedure is to get the questionnaires, for each Enumeration Areas (EAs), ready for scanning. There is a Batch Header to identify each batch.
Each Batch on getting to the Scanning Room was placed on joggers (a vibrating machine)to properly align the forms, and get rid of dust or particles that might be on the forms.
The forms are thereafter fed into the scanner. There were security codes in form of bar codes on each questionnaire to identify its genuineness. There was electronic editing and coding for badly coded or poorly shaded questionnaires by the Data Editors. Torn, stained or mutilated forms are rejected by the scanner. These categories of forms were later manually keyed into the system.
1.6 Re-archiving of Scanned Forms:- Scanned forms were placed in their appropriate marked envelopes in batches, and thereafter returned to the Archiving Section for re-archiving.
1.7 Data Output from the Scanning Machine:- The OMR/OCR Software interprets the output from the scanner and translates it into an XML file from where it is further translated into the desired ASCII output that is compatible for use by the CSPro Package for further processing and tabulation.
1.8 Data back-up and transfer:- After being sure that the data are edited for each EA batch in an LGA, data then was exported to the SAN (Storage Area Network) of the Server. Two copies of images of the questionnaires for each EA copied to the LTO tapes as backup and then transferred to the Headquarters. The ASCII data files for each LGA are zipped and encrypted, and thereafter transfer to the Data Validation Unit (DVU) at the Headquarters in Abuja.
1.9 Data collation and validation:- The Data Validation Unit at the Headquarters was responsible for collating these data into EAs, LGAs, States and National levels. The data are edited/validated for consistency errors and invalid entries. The Census and Survey Processing (CSPro) software is used for this process. The edited, and error free data are thereafter processed into desired tables.
1.10 Activities of the Data Validation unit (DVU):-
- Decryption of each LGA Data File
- Concatenation/merging of Data Files
- Check each EA batch file for EA completeness within an LGA and State
- Check for File/Data Structure
- Check for Range and Invalid Data items
- Check for Blank and empty questionnaire
- Check for inter and intra record consistency
- Check for Skip Patterns
- Perform Data Validation and Imputation
- Generate Statistics Report of each function/activity
- Generate Statistical Tables on LGA, State and National levels.