Using Representative Opinion Surveys in the African Peer Review Mechanism Process

The document examines the methodological requirements for conducting representative public opinion surveys to support the African Peer Review Mechanism (APRM). It argues that while participatory consultations are useful for public ownership, they often produce biased results. Instead, the author advocates for random probability sampling, noting that relatively small samples (e.g., 1,200 respondents) can provide accurate national estimates. The paper outlines essential prerequisites—such as recent census data and freedom of travel—and warns against the common error of polling only heads of households, which biases results toward older, employed males. It also provides a practical timeline and cost estimate for implementing such surveys in Africa.

• Broad participatory consultations often fail to reach a representative cross-section of the public because participation requires individual initiative, which biases results toward the better educated and those more active in civil society.
• Representative accuracy depends more on the method of selection than the number of people contacted; random probability samples of 1,000 to 2,500 people are generally sufficient for socio-political surveys due to the law of diminishing returns.
• A credible representative survey requires three essential conditions: the absence of widespread civil conflict to allow fieldworker travel, the availability of recent and detailed census data for 'probability proportionate to size' (PPPS) sampling, and accurate translations into all relevant home languages.
• Using the 'head of household' as the sole informant is an inappropriate methodology for governance and democracy surveys, as it produces biased results favoring older, employed males and ignores the individual citizen as the proper unit of analysis.
• Public opinion questionnaires for the APRM should not attempt to mirror the official APRM instrument exactly, as ordinary citizens lack the specific experience to answer questions on certain technical areas, such as corporate governance.
• Implementing a nationally representative survey of 1,200 respondents in Africa typically requires a timeline of five to six months and a budget between US$85,000 and US$125,000.

APA

Mattes, R. (2008). Using Representative Opinion Surveys in the African Peer Review Mechanism Process. South African Institute of International Affairs. https://saiia.org.za/wp-content/uploads/2008/10/Occasional-Paper-13.pdf

Chicago

Mattes, Robert. Using Representative Opinion Surveys in the African Peer Review Mechanism Process. South African Institute of International Affairs, 2008. https://saiia.org.za/wp-content/uploads/2008/10/Occasional-Paper-13.pdf.

Wikipedia

{{cite report |last1=Mattes |first1=Robert |title=Using Representative Opinion Surveys in the African Peer Review Mechanism Process |publisher=South African Institute of International Affairs |date=November 2008 |url=https://saiia.org.za/wp-content/uploads/2008/10/Occasional-Paper-13.pdf |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{mattes2008using, author = {Mattes, Robert}, title = {{Using Representative Opinion Surveys in the African Peer Review Mechanism Process}}, institution = {South African Institute of International Affairs}, year = {2008}, month = nov, url = {https://saiia.org.za/wp-content/uploads/2008/10/Occasional-Paper-13.pdf}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }