Recipient of the Bruce H. Choppin Memorial Award, International Association for the Evaluation of Educational Achievement (IEA) 2025

• The award recognizes outstanding master's theses or doctoral dissertations that employ empirical research methods and use IEA data

• Awarded to one at the doctoral level annually, see IEA Insider 2025 for details

Recipient of the George Madaus Graduate Student Research Award, Lynch School of Education and Human Development, Boston College 2024

• The award honors a founder of the MESA department and a major figure in educational assessment and evaluation

• Awarded to one individual annually, "in recognition of outstanding research that contributes to one or more of the following topical areas: advancing the quality of educational and psychological assessments, enhancing the practice of program evaluation, or contributing to reducing the negative impact of test use on students, teachers, and society-at-large"

Recipient of the Doctoral Dissertation Fellowship, Lynch School of Education and Human Development, Boston College 2023

• "Receipt of a doctoral dissertation award is a major testimony to your talent and the contribution we anticipate that you will make to research in your field" 

• The fellowship carries a non-service stipend of $30,000, 2 credits of tuition remission, and 100% coverage of Boston College Student Medical Insurance

• Only 2 fellowships are awarded each year

Research Assistant at TIMSS & PIRLS International Study Center, Boston College 2019.9-2023.8

• For psychometrics and data analysis team: I used R, Stan, and Mplus to perform analysis on process data. 

Supervisors: Drs. Matthias von Davier and Lale Khorramdel

Example projects:

• • I examined whether process data and machine learning models can help improve the estimation of achievement in latent regression models on TIMSS 2019 and PIRLS 2021 data. I used IRT models for deriving EAP estimates of student achievement and log-normal models for speed estimates.

  • I conducted a series of analyses to validate the process data and achievement data for PIRLS 2021.

• For assessment development and reporting team: I conducted research on innovative technology-enhanced item formats and developed new mathematics items (especially Problem Solving and Inquiry items) for TIMSS 2023 assessments.

Supervisor: Dr. Ina Mullis

Using LLMs for Generating Mathematics Items 2023.6-2023.8

GPT has demonstrated its power in automatic passage and item generation in language assessments like the Duolingo English Test, but it has been less commonly used for mathematics item generation. This project explores the use of GPT for generating future TIMSS mathematics assessment items at the fourth and eighth grades. I developed a proposal, which includes evaluating the grade appropriateness, accuracy, and relevance of the generated items, and experimented with GPT 3.5 and 4.0 with the Wolfram Plugin.

Supervisor: Dr. Matthias von Davier

Automated Scoring of Mathematics Expressions 2022.5-2022.9

I completed an independent study project, in which I explored algorithms for automatically scoring constructed response items in the mathematics expressions category of TIMSS 2023 eighth-grade Field Test data. I used longest common subsequence (LCS) and Levenshtein distance and similarity measures in the automatic scoring algorithms. The resulting machine scores achieved a high agreement with human scores (over 93%) for all five items and provided evidence that the proposed methods may be used in TIMSS 2023 operations and future TIMSS cycles for reliably automatically scoring items of similar types.

Supervisor: Dr. Matthias von Davier

A Critical Examination of China's College Admissions 2022.2-2022.5

I critically examined the implementation, outcomes, and concerns related to China’s Comprehensive Assessment policy, a college admissions reform that mandates high schools to establish portfolios for students to document their performances across multiple areas and universities to base enrollment decisions on more than just college entrance examination scores but also the comprehensive assessment results. I found that although the policy aimed to promote non-academic participation among students, it ultimately failed to achieve its intended objectives; the variations in the properties it aimed to assess did not cause the variations in the outcomes (scores, grades, or reports) among students.

Research assistant, “How to improve low-quality courses? Development of a visualization system for assessing general education curriculum,” Project funded by the Ministry of Education (Supervised by Associate Professor Lu Yi)   2016-2018

• In order to evaluate and improve the quality of general education core curriculum, I used MySQL to process data from 10,000+ samples of students from China’s top universities and used Baidu E-charts to generate charts from the MySQL database.

• The system has been adopted by Fudan University and Peking University, and has aided teachers in identifying the strengths and weaknesses of the courses they taught.

Research assistant, “Comparing training programs for top-notch students within different university environments,” Project funded by the Ministry of Education (Supervised by Associate Professor Lu Yi) 2016-2018

Researcher, “An evaluation of Shanghai’s college entrance examination reform policy” 2017.6-2018.3

I analyzed data from interviews with principals from 9 high schools in Shanghai and staff of the Shanghai Education Commission. I found that instead of offering students more freedom to grow academically, the policy incentivized students to choose easier subjects in order to get higher scores.

Researcher, “Investigating the learning behavior and learning gains of postgraduate students of Fudan University’s Research Institute for Higher Education” 2016.10-2017.2

In this study, I used the NSSE-China questionnaire and collected data on the learning behavior and gains of postgraduate students at Fudan University. Specifically, I used SPSS software to describe, correlate, and perform a linear regression analysis. The study concluded that second-year students devoted the least amount of time to their studies, resulting in the lowest level of learning gains.