Paper: National primary-level learning proficiency statistics from Africa
Abstract
This paper responds to three needs. First is a need for updated and more comprehensive and comparable country-level proficiency statistics for the primary schooling level in Africa. Secondly, a systematic and updated review of the relevant data and statistics available for Africa, and their levels of reliability, seemed necessary. Third is the need to explore new ways of harmonising learning proficiency statistics across countries. With regard to the first and second needs, nine data sources are critically discussed, each covering between 2 and 18 countries on the African continent. Seven of the nine programmes involve testing in schools and two in households. By comparing microdata across programmes, where microdata were available, but also comparing national statistics in the absence of microdata, harmonised statistics were produced for 47 of 55 African countries, representing 97% of the continent’s children. The analysis, which attempts to balance coverage with rigour, confirms that both data quality problems and methodological issues make harmonisation difficult. The more coverage achieved, the greater the reliability concerns. Arguably the value of the paper lies at least as much in the evaluation of the data from the programmes, programmes which are continually evolving, as from the final harmonised statistics. Regarding the need to explore new methods, the paper pays special attention to one issue which has probably received too little attention in past harmonisations: adjustments to take into account the fact that even within the same programme different grades might be tested. Moreover, the paper suggests that even published statistics where the underlying microdata are not available for analysis are worth considering in a harmonisation exercise.