Advancements in technology have transformed a range of fields throughout the 21st century.
To continue this momentum and support middle school science teachers, the University of Georgia will use a $3 million grant from the National Science Foundation to enhance artificial intelligence (AI) and machine learning-based assessments in classrooms.
Specifically, the project will study a machine learning-based assessment system to help science teachers make instructional decisions based on automatically generated student reports (AutoRs) based on students’ performance on assessments.
Xiaoming Zhai, the principal investigator of the study and an assistant professor in the Mary Frances Early College of Education’s department of mathematics, science and social studies education, will collaborate with Joseph Krajcik, CREATE for STEM Institute director and professor at Michigan State University; Gary Weiser, a research associate at WestEd; and Yue Yin, a professor at the University of Illinois at Chicago.
“The goal for this project is to forward machine learning and AI in the classroom to allow teachers to use the technology and increment assessment work in their classrooms,” Zhai said. “Using machine learning, we are trying to develop algorithms to allow timely feedback to teachers.”
With machine learning-based assessment systems, computers learn from human experts to develop algorithmic models to automatically score students’ complex performance on assessments.
The project team will develop computer algorithms, a suite of AutoRs and an array of pedagogical content knowledge supports. These products will assist middle school science teachers in the use of Next Generation Science Assessments (also called 3D assessments), making informative instructional changes and improving students’ knowledge-in-use learning.
Additionally, the project will generate knowledge about teachers’ uses of 3D assessments and examine the potential of automatically scored 3D assessments.
“The benefit of this project is that it is cross-institution work,” Zhai said. “This project will help thousands of students with diverse backgrounds across Georgia, Michigan and California and provide professional development to teachers to improve their assessment proficiency.”
AI and machine learning have the potential to revolutionize science assessment practices by significantly improving the functionality and automaticity of scoring constructed responses by targeting complex constructs. According to Zhai, previous research has shown the potential of AI to make inferences based on large-scale and complex data.
“Studies like these suggest AI’s great potential to improve the functionality of assessments—making accurate decisions based on evidentiary data and rigorous inference,” Zhai said. “Due to the improvement of the assessment functionality, AI could potentially assess complex constructs with developmental features such as students’ learning progression using written responses immediately, which are difficult to achieve in traditional classes.”
To further generate knowledge of integrating AI into science assessment, Zhai and Krajcik will host a two-day conference at the University of Georgia. Scholars from across the world and nation will attend to discuss how to further improve AI and machine learning in STEM education.
"Machine learning is increasingly impacting every aspect of our lives, including education," Zhai said. "It is anticipated that the cutting-edge technology may redefine science assessment practices and significantly change education in the future."