2022 (Virtual) MARC Pre-Conference Workshop
Digital learning and assessment generate a large amount of data in various forms. To deal with the complex data,
psychometric researchers need to garner new techniques from data science, machine learning, and natural
language processing (NLP). To facilitate the learning, an edited volume, Computational Psychometrics (Eds. von
Davier, Mislevy, & Hao, 2022), was developed to introduce the new techniques in a way more accessible to
researchers with psychometrics backgrounds. ML and NLP are two very extensive research disciplines
themselves, and it is impossible to cover them fully in such a short workshop. As such, this workshop aims to
provide the participants with a high-level overview of ML and NLP as well as some practical knowledge of using
them in the context of educational assessment. We hope the workshop can help participants to get started with the
acquisition of these important skills.
Basic knowledge in statistics and python programming language.
• Learn the basics of machine learning and some typical algorithms
• Learn how to use machine learning in Python and Orange
• Learn the basics of natural language processing (NLP)
• Learn how to use NLP in Python
Bios of the Instructor
Dr. Jiangang Hao is a research director at the Research and Measurement Sciences Division of Educational
Testing Service and is currently leading the Psychometric and Data Science Modeling group. His research centers
on leveraging big data to measure people and their interactions, with a particular focus on behavioral data mining, game-based assessment, learning analytics, AI/machine learning, natural language processing, automated scoring, learning, and assessment of 21st-century skills such as collaboration and communication. Dr. Hao got his Ph.D. in Physics and MA in Statistics from the University of Michigan. Before joining ETS, he worked on large-scale data in astrophysics and cosmology at Fermi National Accelerator Lab. His research spanned from developing CCD camera and astronomical digital image processing pipeline to developing algorithms for galaxy clustering and cosmological parameter estimation. Dr. Hao published extensively, and most of his publications can be accessed at the Google scholar site: https://goo.gl/pgggDK.