Xueying Tang
University of Arizona, AZ·Mathematics & Statistics
When students take tests on a computer, the software can record every click, keystroke and action they make on the way to an answer. These records, called response process data, show how people solve problems, something a final right-or-wrong answer cannot capture.
- Fields
Mathematics & Statistics
Education Research
- Topics
response process data
educational assessment
computer-based testing
- Grants
- 1 known grant, through Jul 2027
What they work on
- Develop a data-driven method to extract informative features from response process data, the logs of test-takers' actions recorded during computer-based assessment items.
- Build a latent variable model that explains how the moment-to-moment dynamics of a response are driven by a respondent's hidden traits, with new computing algorithms for statistical inference.
- Create a scalar-on-process regression model that describes statistical relationships between response processes and other observed variables, and release it in user-friendly software.
More about the research
But the data are messy, vary in length and format, and reflect very different human behaviors. That makes them hard to analyze systematically. Machine-learning tools can find patterns in them but often work as black boxes that are hard to interpret or validate.
Why it matters
Statistical tools that reveal how individual students approach problems could help educators evaluate students more fully and design individualized instruction or interventions. Methods that are easy to interpret, released as user-friendly software, could let assessment practitioners check and trust results from rich test-taking data, supporting student success and fairness in education.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Peter Halpin
University of North Carolina at Chapel Hill, NC·Education Research
Develop estimators of treatment effects that are robust to item-level heterogeneity, by combining robust statistics with item response theory, the framework linking student ability to the probability of answering each test question correctly.
Jonathan Stewart
Florida State University, FL·Mathematics & Statistics
Develop a new class of generalized latent position models for network data that drop the usual conditional independence assumption, so that dependence among connections, not just node positions, shapes the modeled network.
Matias Cattaneo
Princeton University, NJ·Mathematics & Statistics
Demonstrate the empirical and theoretical limitations of binned regression models applied to mixed-frequency panel data in climate economics, showing when they can lead to incorrect conclusions.
The school
University of Arizona: 36 labs on PI Pages, $1.0B a year on research, 487 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.