Democratizing Signal Processing and Machine Learning: Math Learning Equity for Elementary and Middle School Students

Namrata Vaswani, Mohamed Y. Selim, Renee Serrell GibertarXiv 2024 · DOI · arXiv

Introduction

Success in high-level technical fields like Signal Processing (SP) and Machine Learning (ML) is often decided long before a student enters a university lecture hall. While these disciplines rely on advanced concepts such as linear algebra, probability, and complex coding, the foundation for such mastery is laid in elementary school arithmetic. This paper explores the "leaky pipeline" in STEM, framing early math access as a critical equity issue and proposing university-led outreach as a viable solution to bridge the cumulative learning gap.

The Challenge of Cumulative Learning Gaps

Mathematics is inherently hierarchical. A student cannot master scalar algebra in middle school without fluency in basic arithmetic; without algebra, the path to engineering and computer science degrees is effectively blocked. Systemic barriers—ranging from socioeconomic disparities to pandemic-related learning losses—often cause students to miss the "arithmetic boat" early on. Because math learning builds upon itself, once a gap forms, it tends to widen every year, making it statistically improbable for a student to catch up without targeted intervention.

Likelihood of students catching up in math based on early performance.

As illustrated in the data above, the probability of a student meeting grade-level expectations drops precipitously if they fall behind in earlier years. A student who is "far off track" in 4th grade has only a 10% chance of meeting 8th-grade expectations. By the time they reach 12th grade, a student who was struggling in 8th grade has a mere 3% chance of meeting math expectations, highlighting the urgent need for early-stage intervention.

Programmatic Frameworks for Intervention

Rather than proposing new algorithms, this research outlines a framework for university-led math support. Two successful models are highlighted:

  • CyMath (Iowa State University): This graduate student-led program targets 3rd–8th graders using a hybrid model. It leverages ALEKS, an adaptive learning software, to provide personalized practice that addresses specific individual gaps.
  • Algebra by 7th Grade (Ab7G, Purdue University): Operated by the Minority Engineering Program, Ab7G focuses on under-represented students in grades 3–7. It pairs rigorous math mentoring with "STEM fun labs"—such as Scratch coding and Lego engineering—to build student self-efficacy and interest in technical fields.

Key Insights and Strategic Shifts

The implementation of these programs has yielded several critical insights for educators and policymakers:

  1. The Education-STEM Synergy: One of the most effective tutoring strategies involves pairing Education majors (who excel at pedagogy and emotional support) with STEM majors (who provide rigorous application and real-world context).
  2. The Hidden Gatekeeper: Many parents are unaware that placement tests taken as early as 4th or 5th grade determine the "math track" a student follows. Being placed on a standard track early often prevents a student from taking calculus in high school, which is a common prerequisite for engineering programs.
  3. Practice Over Gamification: While "math games" are popular, the authors argue they often consume valuable instructional time without building fluency. They advocate for a return to consistent, checked homework and individual practice to ensure arithmetic mastery.

Results and Future Outlook

The impact of these interventions is measurable. In a CyMath case study, a student moved from the 40th percentile to the 92nd percentile in standardized testing within 18 months. Similarly, the Ab7G program has scaled from 44 to over 200 participants, showing a direct correlation between software-based topic mastery and proficiency on state-wide (ILEARN) exams.

While university programs are effective, they cannot reach every student due to resource constraints. Long-term equity requires systemic public school policy changes, including better "summer bridge" programs and increased parent awareness. Professional organizations like the IEEE are encouraged to partner with or fund these initiatives to ensure the next generation of SP and ML experts reflects a diverse and equitable talent pool.

Summary updated

Abstract

Signal Processing (SP) and Machine Learning (ML) rely on good math and coding knowledge, in particular, linear algebra, probability, trigonometry, and complex numbers. A good grasp of these relies on scalar algebra learned in middle school. The ability to understand and use scalar algebra well, in turn, relies on a good foundation in basic arithmetic. Because of various systemic barriers, many students are not able to build a strong foundation in arithmetic in elementary school. This leads them to struggle with algebra and everything after that. Since math learning is cumulative, the gap between those without a strong early foundation and everyone else keeps increasing over the school years and becomes difficult to fill in college. In this article we discuss how SP faculty, students, and professionals can play an important role in starting, and participating in, university-run, or other, out-of-school math support programs to supplement students' learning. Two example programs run by the authors, CyMath at Iowa State and Algebra by 7th Grade (Ab7G) at Purdue, and one run by the Actuarial Foundation, are described. We conclude with providing some simple zero-cost suggestions for public schools that, if adopted, could benefit a much larger number of students than what out-of-school programs can reach.
BibTeX
@misc{vaswani2024democratizing,
  author = {Namrata Vaswani and Mohamed Y. Selim and Renee Serrell Gibert},
  title = {{Democratizing Signal Processing and Machine Learning: Math Learning Equity for Elementary and Middle School Students}},
  year = {2024},
  eprint = {2409.17304},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2409.17304},
}
Figure