How can we support introductory science students in building necessary quantitative and other problem solving skills?

Many introductory science courses require basic understanding of calculus and other quantitative skills. Due to varying levels of prior math preparation among students, there are large grade disparities in these courses. In this project, we will build on promising results from a project funded by UCLA’s Teaching and Learnind Center (TLC), described in more detail in our 2025 Physical Review Physics Education Research publication.

The interventions are:

  1. Supplemental Math assignments designed to address gaps in students’ mathematical knowledge;
  2. Scaled incentives, encouraging less prepared students to complete supplemental assignments without introducing the “gifts for the gifted” problem;
  3. State-of-the-art AI tools in a safe and equitable environment, without any extra cost to the student, helping students with the supplemental assignments and enriching the overall learning experience in the course.

Together, these components form the AI and Math Skills (AIMS) program. Our prior research shows that these interventions support student learning and reduce inequities in exam performance.

Supplemental math associated with higher exam scores

Set of graphs indicating correlation between students who complete supplemental math and exam scores.

From our publication – Lu et al. 2025 – this figure depicts the correlation of student exam performance with Force Concept Inventory score (an indication of their conceptual physics understanding upon entering the course), separated by supplemental math material 1 completion. The colored bands mark regions with 95% confidence intervals (after taking into account other variables in the model); “RI” indicates the “Relevance Index” of each exam to the supplemental math assignment.  Controlling for incoming physics understanding and other factors, we see that those who complete the supplemental math tend to have higher exam scores, and that this effect is largest on the third exam, which aligns most with the supplemental math material.

 

EXPANDING TO OTHER COURSES

Thanks to additional funding from the TLC, we are implementing these promising interventions across a wider array of Physical and Life Sciences courses, including the Physics 1 series, Chemistry 14 and 20 series, and LS 30A over the next two academic years, making iterative improvements in our implementation plan to increase the use of the AIMS program by students that need it most. This grant will support the adaptation of supplemental math materials for different course contexts and provide support for initial implementation to instructors. Once instructors have adapted supplemental math for their course and implemented homework assignments in Kudu, using these tools in subsequent iterations of the course will require little, if any, labor on their part. If successful, we anticipate increased adoption of these interventions across the Physical Sciences and Life Sciences, resulting in broad use of supplemental math and AI-generated hints, especially by students who could most benefit from them, leading to lower DFW rates and grade disparities.

WHO IS INVOLVED?

This collaborative project involves folks across UCLA and UCSD, including (in alphabetical order):

  • Justin Caram (Associate Professor and Vice Chair, Chemistry & Biochemistry)
  • Dory Deweese (Assistant Teaching Professor, Chemistry and Biochemistry)
  • Alexander Kusenko (Professor, Physics and Astronomy)
  • Yifan Lu (Graduate Student Researcher, Physics and Astronomy)
  • Shanna Shaked (Senior Associate Director, CEILS)
  • Elizabeth Simmons (Executive Vice Chancellor, UC San Diego)
  • K. Supriya (Associate Director, CEILS)
Screenshot of Kudu AI Hint

Screenshot of AI Hint

Note that AI hints do not provide solutions.

FAQs

In this study, we tried a unique scaled incentive structure for free online supplemental math assignments to increase the completion rate, especially among students that need more support with mathematical skills. We adopted a formula for applying extra credit so that students who perform better in the exam would gain less by completing the materials. With this incentive structure, we were able to achieve higher completion rates among students who might need more support in the course compared to students who are well equipped to perform well in the course, as indicated by students’ incoming FCI scores. In other words, when incentivized with the scaled extra credit, about 54% of students that scored below the class mean on the FCI completed the first supplemental math assignment compared to only 37% of students that scored at or above the class mean. By contrast, without the incentive, only 19% of students that scored below the class mean on FCI completed the first supplemental math assignment, compared to 24% of students that scored at or above the mean.