Total: 1
A central objective of international large-scale assessment (ILSA) studies is to generate knowledge about the probability distribution of student achievement in each education system participating in the assessment. In this article, we study one of the most fundamental threats that these studies face when justifying the conclusions reached about these distributions: the problem that arises from student non-participation during data collection. ILSA studies have traditionally employed a narrow range of strategies to address non-participation. We examine this problem using tools developed within the framework of partial identification that we tailor to the problem at hand. We demonstrate this approach with application to the International Computer and Information Literacy Study in 2018. By doing so, we bring to the field of ILSA an alternative strategy for identification and estimation of population parameters of interest.