Abstract
Since 2003, the Programme for International Student Assessment (PISA) has included students with special educational needs (SEN), identified as those with functional disabilities, those with cognitive/behavioural/emotional disabilities and those with limited test language proficiency. While the number of countries and included students has increased with each test administration, the percentage of students with SEN remains extremely low. The inclusion of these students is not an intentional PISA design parameter but rather a response to the interaction between the need to maintain strict sampling criteria and country-level educational mandates to include SEN students in standardised testing. Based on the analysis of student participation and performance across four cycles of PISA (2003–2012), this paper examines the challenges that exist in current PISA procedures related to: student sampling, eligibility and identification; assessment methodology; and reporting results. PISA practices, their limitations for scientific inferences and recommendations for design improvements are given.
Acknowledgement
The lead author (Barbara LeRoy) received funding from the OECD Thomas J. Alexander Fellowship Program for this work. The opinions and arguments herein are those of the authors and not the OECD.