1 / 25

Results Generated from the Questionnaire Disseminated Prior to the Workshop Pres. 1

Results Generated from the Questionnaire Disseminated Prior to the Workshop Pres. 1. To better understand census evaluation activities at the country level To facilitate the sharing of experience on methods of census evaluation among countries in the region. The Objective of the Questionnaire.

nonnie
Download Presentation

Results Generated from the Questionnaire Disseminated Prior to the Workshop Pres. 1

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Results Generated from the Questionnaire Disseminated Prior to the WorkshopPres. 1

  2. To better understand census evaluation activities at the country level To facilitate the sharing of experience on methods of census evaluation among countries in the region The Objective of the Questionnaire

  3. Was PES conducted for 2000 census Yes : 11 No: 6  8 countries did not return the questionnaire All countries involved National statistical/population census offices in planning and conducting PES except Egypt 7 countries where PES team different from census team (Ghana, Lesotho, Nigeria, South Africa, Tanzania, Uganda, Zambia) 4 countries used same team for planning census & PES (Kenya, Liberia, Mozambique, Namibia)

  4. Funding Planned as part of the census programmed in all countries except Egypt, Ghana and Kenya Funding adequate in 5 countries (Lesotho, Mozambique, Namibia, Nigeria and South Africa)

  5. Objectives of PES Estimate census coverage Characterize missed or over-counted individuals Estimate under-count and over-count for adjusting census results Learn quality of information collected – content analysis Obtain information for design of future censuses and surveys

  6. Recruitment and training of PES staff All countries selected enumerators and supervisors directly from those used in census except 4 (Egypt, Nigeria, South Africa and Zambia) Selection criteria includes: Academic qualification and skills Experience, competence and performance in past enumeration/supervision assignments Location e.g. worked in different areas for the census Training periods: About 2-14 days for most countries 21 days for enumerators in Nigeria

  7. PES sampling designs Stratified cluster sampling (one stage) was the most common design Stratified on regions/provinces and urban-rural dichotomy Sampling proportional to size in Mozambique and Zambia Two stage designs in South Africa and Liberia Three stage designs in Egypt, Ghana and Nigeria Systematic random sampling used in Nigeria

  8. Sample size for 1st stage units EAs were the primary sampling units in most countries Sample sizes for EAs range from 224 EAs in Lesotho to 600 EAs in South Africa Mozambique had a minimum of 30 EAs for each province and 40 for the biggest stratum (the capital city, Maputo) 352 in Liberia (5% of a total of 7026 EAs) 6 domains were sampled for 1st stage in Egypt (with 3 stage design)

  9. Domains of analysis National (all countries) Most countries also included Urban-rural and regional (provincial) analyses District level as well in Ghana Only at national level in Nigeria? Egypt had also analyses for the capital and big cities

  10. Matching methods and documentation of rules Matching method  Manual: 5 (Kenya, Namibia, Tanzania, Uganda and Zambia)  Computer: 2 (Egypt and Mozambique) Combination: 4 (Ghana, Lesotho, Nigeria and South Africa) Documentation of rules Yes:8(Egypt, Kenya, Lesotho, Mozambique, Nigeria, South Africa, Tanzania and Zambia) No: 2 (Ghana and Namibia)

  11. Field reconciliation visits Yes: 9 (Egypt, Ghana, Kenya, Lesotho, Nigeria, South Africa, Tanzania, Uganda and Zambia) No: 2 (Mozambique and Namibia) Note that the field reconciliation process was not completed in Kenya

  12. Nearly all countries used the coverage rate as coverage measure. Lesotho and South Africa also used the net coverage rate. Some of the coverage measures used

  13. For content analysis the following were considered Lesotho and South Africa calculated rate of agreement, net difference rate and index of inconsistency Lesotho also calculated gross error rate South Africa also calculated gross difference rate (and off-diagonal proportion) Zambia intended to calculate content error (but the process was not completed due to technical and funding constraints) The content analysis was also not completed in Kenya

  14. Producing report to inform lay users Planned to produce reports for lay users (5 countries – Egypt, Ghana, Lesotho, South Africa and Tanzania) Not planned to produce reports (Kenya, Nigeria, Uganda and Zambia) All countries that planned to produce reports did so except Ghana Periods of 3 to 36 months from data collection to publication of PES reports for lay users Internal reports took periods ranging from 3 weeks for Nigeria to about 30 months in case of Uganda and Zambia The process was not completed in Kenya

  15. Using PES/census evaluation results to adjust census figures Yes : 5 countries (Kenya, Mozambique, Nigeria, South Africa and Tanzania) No: 6 countries (Egypt, Ghana, Lesotho, Namibia, Uganda and Zambia) Coverage rates used to adjust total population at national and provincial domains of estimation

  16. Some details on adjusting census figures In South Africa net undercount rate was used to calculate adjustment factors within homogenous groups  Groups defined using Chi-square Automatic Interaction Detection techniques  Hierarchical model estimation In Kenya, though matching was not fully done, matched and reconciled data showed census undercount. A consultant assisted in adjusting figures.

  17. Reasons and challenges for not adjusting census figures Incomplete analysis (Ghana and Zambia)  Technical and funding constraints in Zambia Namibia only matched households and not individual information Errors within acceptable limits Coverage rate and other errors were within acceptable limits in Tanzania  Under-coverage of 5.6% in Uganda was also deemed to be within acceptable limits

  18. Different field staff (Nigeria and South Africa). Assigning field staff to areas different to where they worked during census  Enumerators and supervisors in Lesotho, Namibia and Uganda;  Supervisors in Kenya and Zambia A different organization conducted PES in Egypt and Nigeria Different team for planning, implementation and quality control of PES in Tanzania To achieve independence for PES:

  19. An independent consultant supervised PES implementation in Uganda A different technical committee was responsible for the PES in Zambia In Kenya, different starting point Within an EA for the PES to that for census; also sampling independently done from headquarters and later sent to field teams To achieve independence for PES (cont’d)

  20. Highlight of lessons learned Ghana: Planning, documentation and independent body to organize the PES are essential Lesotho: Importance of independence of PES; some knowledge and experience in planning and implementation Mozambique: Need for Piloting PES; organizing short training courses in PES methodology for PES team Namibia: Some households were omitted due to poor supervision in some areas; accessibility difficulties due to locked gates  Also could not match individual information due to shortage of staff

  21. Highlight of lessons learned (cont’d) Nigeria:Potential problems include nomadic and seasonal movements e.g. cattle keepers and farmers and the area was large South Africa:  Bias on adjusting for smaller geographical levels due to smaller samples  Different undercounts for different geographical levels  Need for new methods to account undercount at lower geographical levels  Prolonged census field operations delays those of the PES  Manual matching was required to correct implementation deficiencies of computer matching Tanzania: To plan PES with the main census e.g. documentation, design questionnaire and sampling

  22. Highlight of lessons learned (cont’d) Uganda: To plan PES with the main census  Need for piloting PES Zambia: Need for same format for census and PES questionnaires Need for computer matching Review maps not updated during census Need for appropriate time frames and good timing e.g. flow of resources Kenya: Need of independent staff for PES Plan PES as an integral part of census Need to do PES within a month after census for memory recall Plan PES funding within census budget  Need for capacity building especially computer matching

  23. Many countries (11) are planning to do PES (Gambia, Ghana, Kenya, Lesotho, Mozambique, Namibia, Seychelles, South Africa, Tanzania, Uganda and Zambia) Egypt and Nigeria are not planning to do PES Mauritius has not yet finalized plans No information on Liberia, Sudan as well as Southern Sudan PES for 2010 round

  24. Assess quality of census data i.e. determine coverage and content errors (at different geographical levels) Find sources and causes of errors; Examine characteristics of persons missed in the census Examine implications of any coverage deficiencies on census data Provide census data accuracy indicators to users (e.g. coverage and response quality indicators) Adjust census results for under-count or over-count Update population estimates and Geographic Information Systems Develop and update sampling frames and population registers (e.g. for intercensal surveys) Learn how to improve future censuses and surveys Some of the objectives mentioned for PES for 2010 round

  25. Thank You END

More Related