Data Collection Notes
Census was started with Household listing which counts all the Houses and household Units and also was captured by GPS (Geographical Positioning System) as to locate their localities as in waypoints and areas of living.
Census Household listing have about 105 personals working as Enumerators, Area Managers and Census Districts Superintendent.
After the Household listing, Census office conducted Pilot Census as to Pre-Test the Census draft Questionnaires with about 70 Enumerators.
After all these, thus finalise all questionnaires, logistics and etc for the actual census on November 2009.
1 PHASE I: PRE-ENUMERATION PERIOD
The period culminating in the actual enumeration in November 2009 started with the establishment of a Census Office in Honiara in October/November 2008. The Census Commissioner seconded from the Statistics Office supported by a Secretary, a high-graded Post Finance (Accounts) Officer seconded from SIGAS, Census Statistician, Cartographer and Data Processing Officer - all Solomon Islanders - are to prepare for the conduct of a household listing of each village, a pilot census, the production of maps and village lists of each Enumeration Area (EA) and, the design and printing of questionnaire, manuals, record books, the purchase of equipment for the office and field staff; the recruitment of staff; the establishment of short term training centres; the preparation of training manuals and aids; the planning and execution of a transport schedule to gather supervisors and enumerators in the training centres, transport them to their EA's and later return them home; the publicity needed to make the whole population aware of the census and of the information they are asked to supply. All these tasks have to be done in a particular sequence and be completed strictly before certain dates. The census effort therefore requires utmost co-operation from all parties involved, authorities at various levels, private organisations and the village and urban populations. To foster the support needed for this purpose the Census Commissioner should report to the Census Users Committee at various critical stages of the project and seek concurrence for his planned activities.
In the 1976, 1986 and 1999 censuses the enumeration was done following a two-visit system. Each enumerator was required to visit all households (on the average about 60 households) in his/her Enumeration Area (EA) in the two weeks before census night and enumerate the people found there. The enumerator was then to return directly after census night to cross out any previously enumerated people who did not sleep in the household that night and add to the household any persons, who were not previously interviewed, but had slept in the household on the Census night. It is assumed in this project proposal that this system be continued in the 2009 census.
The Census Office should have the staff to carry out the tasks briefly described above. It is envisaged to divide the nine provinces and one main urban area (Honiara) into seven census districts (CD). In each of these a census districts, a census districts superintendent (CDS) should be made responsible for the proper execution of all preparatory work and of the actual conduct of the enumeration, in full consultation with the Census Commissioner. The CDS's, most or all of whom should be Solomon Islanders, would have to work closely with the provincial authorities and need the assistance of Chief Area Managers (CAM) and or the Area Manager (AM) depending on the level of work and where these personnel's are deployed. The CAM and AM will check in the field, the EA maps drafted in the Census Office on the basis of information of the Village Resource Survey (conducted by the Statistics Office in November 2007 to April 2008); supervise the conduct of a full household listing; explain the census in village meetings; prepare a transportation plan for his area; help in recruiting supervisors and enumerators; play active roles in their training and in co-ordinating their transport and work during the actual census enumeration.
There will be approximately 1611 EA's and therefore a need for 1611 trained enumerators and for 266 supervisors. All of these personnel must be trained (taking into account 20 per cent reserves for 'no-shows' and 'dropouts'), the supervisors in six (6) centres and the enumerators in 14 centres. Temporary assistance will be needed from the authorities to provide professional trainers as well as a few higher level staff to act as Directors of some of the training centres, which will last for eight and five working days for the supervisors and enumerators respectively. Obviously, the Trainers, Directors and the Quality Controllers (who are to help screen completed forms before they will be returned to Honiara for data processing) should also be trained and this will equally apply for longer term staff, i.e. Census Office staff, CDS's, and AM's who will have their own courses earlier in the year. The high level staff can participate in special courses if the need arises, however, these types of training is expected to be provided by the CTA, as part of an on-the-job training activity of the Census.
2. PHASE II: POST-ENUMERATION PERIOD
In this phase the census returns will have to be processed, tabulated, analysed and disseminated in the form of published reports. It is argued, that while in previous censuses data processing was largely done overseas the time is now ripe to carry out all the required work locally. This includes the following steps:
i. Preparation of the code-scheme, tabulation lay-outs, creation of a test file, system design for the various processing steps and testing of these systems.
ii. Arrangements of office space, furniture and equipment.
iii. manual checking of the returned forms, transportation of the information into numeric codes on the forms, manual compilation of provisional small area statistics for early release as per 2010 National General Election requirements and correction of EA maps and village listings.
iv. Scanning of the questionnaires on scanners to create a complete computer data file of the census information; this will then be transferred into defined data dictionary using CSPro. With two scanners and 24 staff, four groups of 6 persons each will operate the scanners in two shifts per day for two months.
v. Remaining erroneous codes are to be checked by a checking section of five persons verifiers working with two PC's, who are to detect the underlying reasons for errors, devise ways to correct them and, for a limited number of variables, change wrong or not stated codes by imputation.
vi. The tabulation section is to produce various types of tables (e.g. 'ad hoc' tables, provincial tables, publication tables, and work tables, tables on request of users and cartograms, which all should be carefully scrutinised for consistency between them.
vii. The specialised member of the census project will subsequently analyse the quality of the reported census data and by comparison with previous census data, establish past trends, which can then serve as indicators for probable future developments. Specialised computer packages will be used for this purpose and much attention will have to be given to presenting the conclusions in comprehensive, but readable texts.