International large-scale assessments have been increasingly used by countries all over the world to reveal insights into education systems through the lens of student performance. However, there is growing evidence to suggest that these assessments fall short when attempting to accurately and reliably assess students from lower-performing countries, particularly those with challenging and diverse educational contexts. This seminar critically examines the methodological challenges of measuring foundational skills in diverse educational environments.
Drawing on psychometric analysis of data from PIRLS and PISA for Development, the presentation explores how current assessment approaches struggle to provide meaningful information about student learning in contexts with low learning outcomes and significant educational heterogeneity. By exploring the challenges that emerge when these tests are administered across vastly different educational environments, the presentation unpacks the complex interactions of widening participation and measurement precision. Specific attention is given to how flooring effects – where large proportions of students score at very low levels – have the potential to distort our understanding of learning outcomes and potentially misrepresent educational progress in low- and middle-income countries.
