IEA INSIDER 2025
PIRLS: Helping South Africa Move from Crisis to Correction
BY MARK CHETTY
An extraordinary amount of critical commentary on South African children’s reading ability followed the release of the PIRLS 2021 results. Most of the criticism pointed to weak educational strategies around foundational learning and language competence. The headlines gravitated towards a crisis of educational planning but offered little or no tangible solutions. Neither did they show an appreciation of the uniqueness of the PIRLS South Africa data and the depth of diagnosis needed.
PROBLEMATIZING THE PIRLS SCORES The South African PIRLS 2021 results indicated that more than 25 percent of students have proficiency levels that are at or below the level of guessing on the assessment. A PIRLS Technical Advisory Group (TAG) composed of international and local experts in the field of large-scale assessment and measurement was commissioned by the Department of Basic Education (DBE) to conduct an intensive review of the PIRLS data and its implications for policy making and reporting in South Africa. A deeper conversation on PIRLS data in context of the rich linguistic diversity of South Africa emerged. IMPLICATIONS FOR REPORTING PIRLS SCORES The TAG argued that the estimated achievement distribution of grade 4 South African students on the 2021 cycle of PIRLS depends more on the statistical assumptions of the model than on actual item responses from the students. Importantly, for many PIRLS items, the probability of a correct response was low for a typical South African student and a typical grade 4 student in South Africa had just a 32 percent chance of correctly answering a typical PIRLS item. The consensus of the TAG was that South African children do not have reading skills that would allow them to do well on PIRLS, in its current design. The significance of this was that solving the reading crisis needed to go beyond item statistics, to a deeper root of the problem. MAKING A DRASTIC CORRECTION While there were clear implications for IEA in making the PIRLS test more suitable to assess proficiency in economically developing countries and introducing a lot
more easier items in line with the existing framework, the PIRLS data invigorated a substantive discussion and policy dialogue on aligning language of instruction to home language teaching, learning, and assessing that reflects the rich linguistic diversity of South African students. What could clearly be seen from the PIRLS data is that students tested in English and Afrikaans (2 of 11 official school languages) substantially outperform their peers tested in African home languages. To that end, understanding and/or articulating plausible reasons for these differences is an important first step to improving reading literacy in South Africa. The dialogue on PIRLS propelled actions underpinning a constitutional right of South African children to receive quality education in 1 of the 11 official home languages and to affirm the linguistic equality of these languages. This fundamental right to learning has been restrictively applied with the promotion of English and Afrikaans as historical languages of instruction at the expense of extending children’s linguistic ability in their mother-tongue or home language beyond foundational learning. However, it is important to recognize that although English- and Afrikaans-tested students do better than their African language-tested counterparts, their achievement is still substantially below the international average. To contextualize this relatively low performance, one can again refer to the PIRLS benchmarks, where 400 is set as the low benchmark. On average, English- and Afrikaans- tested students are below this low benchmark, indicating
34
Made with FlippingBook - PDF hosting