The fields of science, technology, engineering, and math have made strides toward greater inclusivity in recent decades, but much work still remains before the STEM workforce truly mirrors the diversity found in the general population. For example, a recent Pew Research study found that only nine percent and eight percent of STEM jobs are held by Black and Hispanic workers, respectively. The study also found that women are under-represented in key STEM job categories—making up 40 percent of physical sciences jobs, 25 percent of those in computing, and just 15 percent of those in engineering.

So, what can be done to close these and other diversity gaps? One idea, currently being studied by a Harvard postdoctoral fellow, is to use machine learning to understand the factors that contribute to disparities in the STEM fields.

The initiative will use machine learning (ML) tools to identify the factors that feed into implicit bias about science and math in middle school students. The hope is that ML may be able to detect students who are likely to fall prey to implicit bias and may be at risk for giving up on more advanced math classes. If successful, the project would enable administrators, teachers, parents, or other stakeholders to intervene and hopefully encourage the students to continue their studies and retain their interest in STEM.

It’s an interesting concept because middle school is recognized as a critical time to focus on STEM education. Students are typically developing stronger critical-thinking and problem-solving skills and also beginning to think more realistically about future career paths. But it’s also important to continue to focus on diversity within the STEM community at the collegiate level.

According to Katharine H. Cole, vice provost and dean of academic affairs at the University of Maryland, Baltimore County (UMBC), success in STEM is dependent on three key factors:

  • An inclusive community that helps support the student
  • The student’s innate resiliency—for example, being able to receive a D in a statistics exam and not become discouraged about STEM
  • Learning strategies

UMBC has developed a number of programs to help students identify new learning strategies and nurture the resiliency required for long-term STEM success. As more universities follow suit and as researchers evaluate the role of ML and other technologies, hopefully we’ll see a more diverse STEM workforce.