Selected Country Experiences
In Kenya, the 2019 census was monitored by a diverse team of independent observers from UN agencies, civil society organisations, African NSOs, academic institutions, and the National Gender and Equality Commission. These observers were trained using standardised tools and methodologies and were deployed across all regions. Their responsibilities included assessing storage facilities, evaluating the clarity of Enumeration Area (EA) boundaries, reviewing pre-enumeration household listings, and observing the use of materials and the proficiency of enumerators. Their real-time feedback was submitted to the Kenya National Bureau of Statistics (KNBS). Additionally, Kenya integrated a supervisor module into its CAPI system, allowing supervisors to randomly re-interview households and address any inconsistencies. This was complemented by a team of coordinators from various census committees who conducted further monitoring and oversight.
The digital census in Kenya used over 164,000 mobile devices and while successful, it encountered issues with device and network reliability in remote areas and with public trust. Kenya’s experience underscores the importance of proactive risk management and transparency.
The data quality assurance adopted by The Gambia and several other countries in Africa was multi-faceted and included: digitisation of maps and compound imageries to be used to track progress of work and performance of enumerators in EAs, use of supervisors for CAPI validation of collected and shared data information, use of inbuilt consistency checks (filtering questions, skip patterns), use of data quality monitors, use of regional field monitors/regional census officers and their deputies, use of enumeration coverage dashboard, use of a Call Centre.
To ensure data quality, Ghana implemented a multi-faceted approach that included the use of CAPI with built-in validation, real-time data monitoring, and independent monitoring by international experts. Ghana also established District Data Quality Management Teams to monitor data collection in real-time and resolve issues as they arise. These teams were crucial for ensuring data accuracy and completeness at the local level.
Zambia implemented a multi-tiered supervision system during its census. Supervisors managed about six enumerators each, conducted re-interviews, and reviewed performance using a digital dashboard. This dashboard, developed with UNECA support, provided real-time tracking of key indicators like interview duration and population coverage. National Data Quality Monitors used this data to guide field visits and provide feedback. Geographical Information System (GIS) tools were employed to verify EA coverage before certifying completion, ensuring thoroughness and accuracy.
Tanzania emphasised quality control throughout its census process. Supervisors and coordinators were trained extensively, and pre-enumeration planning included well-designed questionnaires and GPS-supported mapping. The census application was developed using detailed specifications and underwent rigorous testing in three phases: in-house, pre-testing, and pilot. During enumeration, supervisors conducted spot checks and used tablet-based quality check programs. Data sent to headquarters was validated using edit specifications, and hard checks in the CAPI system helped minimise content errors.
Namibia focused on layered quality assurance through regional and national supervisors. Monitoring teams were deployed at the start of fieldwork to ensure adherence to procedures. They observed enumerators, especially in translating questions into local languages, and took corrective actions when needed. A real-time monitoring dashboard enabled supervisors to track data quality. If errors were found, records were returned to mappers for correction, ensuring high data integrity.
Nigeria, in its first planned digital census feared the immense challenge of managing data security and logistics for Africa's most populous nation. Mitigating risks requires a robust cloud infrastructure and extensive training for a massive number of enumerators, all within a complex security environment.
Ghana, in its 2021 digital census, like several other countries had to contend with internet connectivity issues, forcing the use of offline data collection methods in certain areas. This created additional logistical challenges and data synchronisation risks that had to be carefully managed.
South Africa experienced a census undercount, even with the use of digital systems. This points to the fact that technology alone cannot solve underlying challenges like respondent fatigue and reaching hard-to-count populations, indicating that risk mitigation must also focus on these persistent social issues.