TItle: From Raw Responses to Statistical Inference: Preserving Data Integrity in Assessment Modeling

Abstract: Modern assessments generate large and complex datasets, but data-quality issues can have important consequences for statistical modeling and inference. In this talk, I will discuss data integrity across the assessment data lifecycle, focusing on how atypical response patterns and other sources of data-quality issues can affect statistical estimates and assessment outcomes. I will highlight statistical and machine learning approaches for detecting and addressing these issues and discuss how data-integrity checks can be incorporated into data-processing pipelines to support more robust statistical analyses and assessment decisions.

Bio: Guher Gorgun is an Assistant Professor of Measurement, Data Science, and Statistics in the Department of Educational Psychology at the University of Georgia. She received her Ph.D. from the University of Alberta, Canada in 2024. Her research lies at the intersection of psychometrics, learning sciences, and artificial intelligence, with a focus on computational approaches to learner modeling and adaptive assessment. Drawing on her experience across academia and industry, her work aims to translate methodological advances into scalable and practical educational and assessment systems. Her current research focuses on automated assessment generation and evaluation, with the goal of advancing more personalized assessments.