Anastasi Lecture 2022
TIMSS, PIRLS, and Friends: Driving Innovation in International Educational Assessments
Matthias von Davier, PhD.
Monan Professor in Education
Executive Director, TIMSS & PIRLS International Study Center
International Large Scale Assessments (ILSAs) have been conducted since TIMSS 1955 started reporting trends in mathematics and science achievement of students around the world. PISA and other, regional assessment such as PASEC, ERCE, and SACMEQ followed. This presentation describes current activities around TIMSS 2023, and preparations for PIRLS 2026, both setting the standard for fully digital assessments of student achievement with a focus on education, not economics. The talk will introduce the assessments, and discuss how TIMSS and PIRLS benefitted from a careful transition to computer-based assessment, by implementation of innovations that serve the needs of countries, while reducing the burden of participation.
The talk closes with notes on how adopting modern assessment technologies will help to further improve international comparisons of student achievement.
Dr. Matthias von Davier is the Executive Director at TIMSS & PIRLS International Study Center and Monan Professor in Education at Boston College. His research focuses on developing psychometric models for analyzing data from complex item and respondent samples and on integrating diagnostic procedures into these methods. His areas of expertise include topics such as item response theory, latent class analysis, classification and mixture distribution models, diagnostic models, computational statistics, person-fit, item-fit, and model checking, as well as hierarchical extension of models for categorical data analysis, and the analytical methodologies used in large scale educational surveys. Dr. von Davier's applied research uses these methodologies to analyze data from educational testing, large-scale survey assessments of student skills and adult literacy, to computer-based assessment of skills, and to the analysis of questionnaire data.