By Lina Sorg
The U.S Department of Labor is forecasting a higher-than-average percent change in workforce size over the next 10 years for occupations in science, technology, engineering, and mathematics (STEM). As such, the prospects are promising for upcoming and recent graduates seeking to find employment in these fields — so long as they know what types of skillsets and experience employers expect. During a minisymposium presentation at the 2020 SIAM Annual Meeting, which took place virtually this week, Ariel Cintron-Arias of East Tennessee State University spoke about retention, graduation, and employment rates in quantitative STEM fields. His talk reflected research from the last several years and offered both relevant statistics and guidance on the ways in which early-career mathematicians can best prepare for the workforce.
Cintron-Arias opened his presentation with a discussion of STEM workforce production from 1970 to 2018 in five distinct fields: biological and biomedical sciences, mathematics and statistics, computer and information sciences, physical sciences, and engineering (see Figure 1). “We see very promising features over the last 50 years, in particular from the year 2000 to 2018,” he said. “Every single one of these curves is showing some signs of growth.” The strongest growth occurred in biological and biomedical sciences, engineering, and computer and information sciences.
Figure 1. Longitudinal number of awarded bachelor’s degrees in science, technology, mathematics, and engineering (STEM) fields from 1970 to 2018. Data courtesy of the U.S. Department of Education.
To continue this positive trend, universities must first focus on recruitment efforts. Immediately after students enroll, they should shift their focus to retention. Cintron-Arias shared a schematic that showcases the relative percent change of U.S. enrollment in 12-month cycles for the same five STEM fields from 2016 to 2019 (see Figure 2). As evidenced by the chart, at least one area of STEM was in decline per year. Cintron-Arias drew listeners’ attention to 2019. “Last year was worrisome,” he said. “At the end of the last decade, most STEM areas had negative percent change, which means that there were declines in enrollment in physical sciences, mathematics and statistics, engineering, and biological sciences.”
Figure 2. Annual relative percent change in STEM enrollment at four-year institutions of higher education across the U.S. Any measure to the right of 0 indicates positive change, while any measure to the left is negative and implies decline rather than growth. Data courtesy of the National Student Clearinghouse Research Center.
In 2017, the U.S. Department of Commerce published an update report about STEM jobs that compared job openings with qualified applicants (see Figure 3). In every aforementioned area—with the exception of the biological sciences, whose market is at capacity—there are more job openings than qualified personnel to fill them. This is good news for students and early-career scientists, who should have ample employment opportunities after graduation.
Figure 3. Stacked bar plots with the national expected number of annual job openings and qualified applicants in STEM areas. If the bar is split half-and-half, there are as many job openings as applicants and the market is at capacity. Figure created by Ariel Cintron-Arias based on data from the U.S. Department of Commerce's "STEM Jobs: 2017 Update" report.
Cintron-Arias moved to a brief exploration of the required and desired skillsets for various STEM positions. He used O*NET OnLine to conduct a search for the types of technical skills and “hot technologies”—technology requirements frequently included in employer job postings—that employers seek in physicists, mathematicians, and statisticians. For example, mathematicians should possess a strong programming background with experience in Python, R, Linux, and MATLAB, among other languages. The database functions somewhat like a background check for an occupation, and users can filter their searches to customize results for any field.
Last year, computer architecture researcher Anant Agarwal published an article entitled “Three Skillsets Every Employee Needs In 2019's Digital Economy.” Agarwal identifies these skillsets as human skills, business enabler skills, and digital building block skills. The first category encompasses traditional soft skills, such as the ability to work cohesively in a team and apply social, creative, and critical thinking to work projects. Business enabler skills refers to the use of pragmatic approaches to solve practical problems, manage projects, communicate data and digital design, and connect digital technologies with broader business goals. The last division—digital building block skills—is most relevant to mathematicians and STEM-based researchers. Qualified workers should be able to leverage technology to add value and align with fundamental domains, engage in computational thinking, and embrace data science and machine learning techniques. This class is especially useful to current or aspiring data-driven decision-makers.
Cintron-Arias then transitioned to cloud computing, which underlies the growing fields of data science and artificial intelligence. Cloud computing refers to the on-demand availability of computer system resources, especially data storage (cloud storage) and computing power, without direct active management by the user; it is thus particularly relevant in the time of COVID-19 lockdowns. Many cloud computing service providers—such as Amazon Web Services (AWS), Google Cloud, Microsoft Azure, and IBM Cloud—are breaking the expectation for “normal training,” as a formal degree is not always necessary to acquire digital building block skills. “It’s unrealistic that we’re going to have a chemist or biologist become specialized with a set of skills, pause their employment, go back to school for a few years, then come back to the workforce,” Cintron-Arias said. This is where the idea of “micro-credentials” comes into play, which honor participants with merit badges or continuing education units in a more flexible format than traditional schooling.
AWS Educate is one example of an initiative that offers micro-credentials. It is based on a system of cloud career pathways that users can explore, earning “badges” for the training sequences they complete. Google’s Kaggle houses micro-courses in practical skills through Kaggle Learn, which awards users with certificates upon course completion. LinkedIn Learning, edX, and Coursera also have similar courses in data science, meant for participants to learn relevant skills at an accelerated pace. Universities would likely benefit from structuring their programs for a mathematical modeling minor—with either a focus on statistical modeling or applied and computational mathematics—based on existing micro-credential schemes.
Ultimately, Cintron-Arias emphasized the value of equipping students, researchers, and employees with a robust toolkit comprised of database knowledge, object-oriented programming, data analysis skills, and cloud computing. Data science continues to grow in importance and relevance, and STEM students must be prepared to encounter it in the workforce. “When you integrate mathematics, statistics, and computation, that’s one way of defining what data science is,” Cintron-Arias said. “These types of skills are crucial to economic growth, national competitiveness, and national security.”
 |
Lina Sorg is the managing editor of SIAM News. |