RESEARCH

RESEARCH INTERESTS

My research focuses on technology-enhanced assessment of 21st-century skills, particularly problem-solving, collaboration, and self-regulated learning. I investigate how process data generated in authentic digital learning and assessment environments—including log data, temporal interaction traces, chat, and other multimodal data—can reveal how students approach tasks, interact with others, and regulate their learning over time. Drawing on computational psychometrics, learning analytics, educational data mining, and AI-based methods, I develop and apply analytical approaches to identify meaningful patterns in students’ cognitive, behavioral, and collaborative processes. Through this work, I aim to advance more authentic and informative approaches to assessment by complementing traditional outcome measures with process-based evidence and providing a more fine-grained understanding of students’ 21st-century skills.

Keywords: Educational Assessment; 21st-Century Skills; Process Data; Authentic Digital Learning and Assessment Environments

Image generated using ChatGPT

RESEARCH PROJECTS

2026.04.01 – 2026.10.31 Identifying Adaptive Problem-Solving Profiles to Mitigate Systemic Barriers in Canada’s Labour Market (Co-PI; PI: Dr. Okan Bulut)

Funding: CA $88,480 from Future Skills Centre
Summary: This project investigates adaptive problem-solving (APS) skills among Canadian adults using data from the international large-scale assessment PIAAC. It examined the impact of APS on outcomes such as health, wages, and job security, compared Canada’s performance with that of the United Kingdom, Germany, Japan, and Finland, and explored effective training approaches and learning supports to strengthen these skills.

2025.07.16 – 2025.11.15 Response Time-Informed Deep Learning-Based Multiple Imputation (PI)

Funding: CA $17,500 from Mitacs Accelerate and the Educational Testing Service (ETS)
Summary: This project leverages deep learning models and item response times to enhance the imputation of planned missing responses in the large-scale assessment PISA.
Deliverables: Findings have been presented at AERA 2026 and NCME 2026.

2023.09.01 – 2025.08.31 Assessment of Collaborative Problem-Solving Ability Using Machine Learning Techniques (PI)

Funding: CA $62,000 from Alberta Innovates
Summary: This project proposes a process-based analytical framework for assessing collaborative problem-solving skills by modeling both dynamic collaboration and problem-solving processes. Leveraging generative AI models, social network analysis, and sequential response modeling, the project analyzes group conversations and action-sequence data and integrates these into a psychometric framework to estimate latent skills at both the individual and group levels.
Deliverables: Findings have been published as a PhD Dissertation.

2023.09.01 – 2024.03.31 Examining Preschool Children’s Problem-Solving Strategies Using Log-Based Process Data (PI)

Funding: CA $12,000 from the University of Alberta
Summary: This project investigates preschool children’s planning and problem-solving behaviors using mobile log data collected from an intelligent tutoring system.
Deliverables: Findings have been published in Interactive Learning Environments.

2022.08.01 – 2023.04.28 A Machine Learning Assisted Systematic Review of the Use of Process Data in Computer-Based Assessments (PI)

Funding: US $16,000 from the Achievement & Assessment Institute, University of Kansas
Summary: This review examines the use of process data in computer-based assessments, including indicator construction, developmental trends, and applications in learning contexts.
Deliverables: Findings have been published in Computers & Education.
Publicly Available Resources: Extracted information from the systematic review.

Page updated

Google Sites

Report abuse