Learn about diversity at ASEE
ASEE would like to acknowledge the generous support of our premier corporate partners.


By Gillian M. Nicholls, Harvey Wolfe, Mary Besterfield-Sacre, and Larry J. Shuman

Laptop with People on screenUncovering STEM Talent

A math model shows students’ potential is evident by the eighth grade.

Five years ago, a blue-ribbon panel warned in “Rising Above the Gathering Storm” that the scientific and technical building blocks of U.S. economic leadership were eroding compared with other nations. One recommendation was to “enlarge the [STEM] pipeline by increasing the number of students who take AP/IB science and mathematics courses by creating opportunities and incentives for middle and high school students to pursue advanced work in science and mathematics.”

How bad is the situation? Of the 4 million first university degrees in science and engineering awarded worldwide in 2006, 21 percent were earned by students in China, compared with 11 percent by students in the United States. Science and engineering account for approximately one third of bachelor’s degrees awarded in the United States, versus 63 percent of those in Japan, 53 percent in China, and 51 percent in Singapore.

While these statistics seem bleak, our mathematical modeling shows there is potential to increase significantly the proportion of U.S. students studying science, technology, engineering, and mathematics – the STEM fields. In fact, educators can gain valuable insight into students’ interest in and capability of ultimately completing a STEM degree as early as the eighth grade.

Using a battery of sophisticated statistical techniques, we examined data from the National Education Longitudinal Study of 1988 containing over 11,000 student records from eighth grade through final educational outcomes. This data set includes a rich array of variables from school records, parents, educators, and students themselves. The variables examined included grades, mathematical skill assessments, standardized test scores, race and ethnicity, and gender. Additional variables reflected demographic characteristics; how students viewed their education; how their parents or guardians viewed the importance of education; the investment of personal time in school, work, and social activities; and student academic performance across multiple subjects.

Students’ final educational outcomes were classified as earning a four-year college degree in STEM, non-STEM, or in a STEM-related subjects, the latter being a category we created to examine students completing quantitative coursework for majors not traditionally considered STEM. Other outcomes included earning a two-year degree, completing high school, or dropping out of high school. We created a model to predict which students would earn a STEM degree as opposed to another outcome, and compared predicted and actual outcomes to measure model accuracy.

We found that approximately half of students who earned a STEM-related or non-STEM degree would have been candidates to pursue a STEM degree, thereby increasing the pool of potential STEM students by as much as a factor of four. We believe the model has strong potential to select students for a pro-STEM intervention program.

Our findings suggest three strategies to increase STEM enrollment: First, explore options to improve students’ educational preparation before junior high school. By eighth grade, students struggling in mathematics are less prepared to keep up with the requisite math courses leading eventually to a STEM degree. Second, engage the interest of students in scientific and quantitative subjects so that they consider pursuing a STEM degree. Third, work to reduce the number of capable college students who, because of academic difficulties or waning interest, switch from STEM to other majors.

Gillian M. Nicholls is an assistant professor of industrial and systems engineering and engineering management at the University of Alabama in Huntsville; at the University of Pittsburgh in the Department of Industrial Engineering, Harvey Wolfe is professor emeritus, Mary Besterfield-Sacre is an associate professor and faculty fellow and a center associate for the Learning Research and Development Center, and Larry J. Shuman is senior associate dean for academics and professor of industrial engineering. This article is based on “Predicting STEM Degree Outcomes Based on Eighth Grade Data and Standard Test Scores” in the July 2010 Journal of Engineering Education.




© Copyright 2010
American Society for Engineering Education
1818 N Street, N.W., Suite 600
Washington, DC 20036-2479
Telephone: (202) 331-3500