Graduate Students in Research on International Large Scale
Assessments – TIMSS & PIRLS Lab

Meng Lyu

Graduate Research Assistant, TIMSS

Meng is a PhD student in the Measurement, Evaluation, Statistics, and Assessment program, focusing on psychometrics and the use of large language models (LLMs) in educational assessment. At TIMSS & PIRLS, she works on Automated Item Generation, investigating how different LLMs perform in generating mathematics items for large-scale assessments and how item quality can be evaluated through post-generation review and validation. In addition to her work on AIG, Meng is broadly interested in the full cycle of large-scale assessment, including item review and automated scoring. Her research explores how AI can be integrated into assessment systems in ways that enhance efficiency while maintaining measurement validity and fairness.

Jeneve Swaby

Graduate Research Assistant, PIRLS

Jeneve is a second-year PhD student in Measurement, Evaluation, Statistics, and Assessment. She works on the development of reading assessment instruments for PIRLS: reviewing feedback from experts and countries, creating text maps for reading passages, and ensuring the passages are sensitive to cultural differences and suitable for fourth-grade students. Jeneve also contributes to an exploratory project for generating items using artificial intelligence, reviewing the output according to the four reading comprehension processes specified by the PIRLS framework.

After obtaining her PhD, Jeneve hopes to work as a psychometri­cian/quantitative methodologist on large-scale assessments that influence policy decisions. She is interested in using process data to identify sources of variation in test-taker performance and to support the valid interpretation of test scores.

Tianzheng (Mark) Mao

Research Associate, Measurement and Data Analysis; PhD student in MESA

Tianzheng (Mark) Mao is a PhD student and Research Associate. His academic and professional endeavors are rooted in the field of educational measurement and data analysis, with a particular focus on Item Response Theory (IRT) modeling.

In his current role, Mark is engaged in a comprehensive study to evaluate the efficacy of various linking methods for IRT modeling. Specifically, he is exploring techniques that facilitate the linking of a large number of independent calibrations. This includes an in-depth analysis of the simultaneous linking method proposed by Haberman and the Alignment method developed by Muthén et al. He is also investigating the potential of machine learning techniques to achieve similar linking objectives.

Mark’s goal is to enhance the precision and scalability of linking processes in IRT modeling by bridging the gap between classical statistical methods and modern machine learning techniques in educational measurement. Through his research, he aspires to contribute to the advancement of assessment methodologies, ensuring that they are both scientifically rigorous and practically applicable.