Do students who choose to major in different fields have different academic aptitudes? This question is worth investigating for many reasons, including an understanding of what fields top students choose to pursue, the diversity of talent across various fields, and how this might reflect upon the majors and occupations a culture values.
In order to explore this, I used five different measures of US students’ academic aptitude, which span 1946 to 2014, and discovered that the rank order of cognitive skills of various majors and degree holders has remained remarkably constant for the last seven decades.
An important caveat: The data presented looks only at group averages and does not speak to the aptitude of specific individuals. Obviously there are people with high academic aptitude in every major and there can be larger aptitude differences between entire schools—for example the University of Chicago and a local community college—than between majors within a school. Also interests, which are not directly assessed here, likely play an important role in which major someone selects. One could argue that any one specific test and sample may not be an accurate reflection of the aptitude of specific majors, and this would be a valid point. However, this analysis uses five independent measures and samples of academic aptitude at different points in time—which include everything from tests of cognitive abilities to tests of academic achievement—showing these findings replicate and are quite robust.
In 1952, a study by Dael Wolfle and Toby Oxtoby published in Science examined the academic aptitudes of college seniors and recent graduates by discipline. The first sample used to investigate this question was standardized test scores on the Army General Classification Test (AGCT) scale from a sample of 10,000 US college graduates from 40 universities in 1946. The AGCT was originally used as a selection test of general learning ability in the military, and its modern equivalent is the Armed Services Vocational Aptitude Battery (ASVAB), which is still in use today.
The second sample was scores from 38,420 US college seniors who took the Selective Service College Qualification Test (SSCQT) put on the AGCT scale in 1951. This was a 150-item test measuring students’ mathematical and verbal ability that does not appear to be in use today.
In both samples, the pattern was nearly identical. Students who had chosen to major in education and agriculture had the lowest average academic aptitude, whereas the opposite was found for engineering and physical sciences.
The next source of data comes from a research paper I published with colleagues David Lubinski and Camilla Benbow in the Journal of Educational Psychology (pdf). Project Talent is a stratified random sample of the US population of about 400,000 students who were tested in high school on math, verbal, and spatial aptitude and graduated in the early 1970s. They were followed up 11 years after high school graduation to assess their educational, occupational, and broader life outcomes. The following chart shows general average academic aptitude by major, along with the pattern of average math, verbal, and spatial aptitude within each major as well as between students who earned bachelor’s, master’s, and PhD degrees.
Project talent: General, math, verbal, and spatial aptitude
The pattern across majors was, again, nearly identical to the independent samples in 1946 and 1951, with education at the bottom and math/computer science, physical science, and engineering at the top.
The next sample comes from over 1.2 million students who took the Graduate Record Examination (GRE) between 2002 and 2005 and indicated their intended graduate major. The data were adapted from the earlier study (pdf), which also used Project Talent.
Even among select GRE test takers, the pattern of education at the bottom and math/computer science, physical science, and engineering at the top remained the same.
The final sample was based on the Scholastic Assessment Test (SAT) and comes from “The 2014 SAT Report on College & Career Readiness.” The average math and verbal aptitude was taken for college bound seniors who indicated their area of study based on a total sample of about 1.6 million.
Reflecting back on the graphs from 1946 and 1951, both agriculture and education were also at the bottom. And again, the traditional science, technology, engineering, and mathematics (STEM) fields such as engineering, physical sciences, and mathematics/statistics tended to be at the top. However, the SAT data allowed a more detailed look due to more categories available. Interestingly, social sciences appear to be at the top along with the STEM majors. And yet, psychology, a social science, appears near the bottom. My hypothesis is that the higher average for social sciences is due in part to the fact that schools who select students with the highest test scores—such as Harvard University, Columbia University, Stanford University, University of Chicago, and Washington University in St. Louis—have “social sciences” as their most popular major according to data from US News. Finally, business appeared near the bottom from 1946 to 2005, but by 2014 had risen to the middle of the pack. This shows business is attracting more able students in recent years, perhaps due to the value of this major among current employers.
Similar patterns are found in many other sources of data, including within a select sample of students in the top 1% of academic aptitude. Even within participants at the World Economic Forum in Davos and billionaires, a similar pattern is found across the sectors in which they operate or made their money.
Why have STEM majors consistently been at the top?
According to a recent Payscale college salary report, STEM majors tend to be the most highly compensated. That STEM majors have consistently had the highest average academic aptitude may also reflect the fact that STEM disciplines are highly complex and require such aptitude. Even scientists in the “hard” STEM fields (e.g. physics, math) tend to believe that these fields require brilliance or genius according to a recent paper published in Science by Sarah-Jane Leslie and colleagues, perhaps because it is true, at least in part. In some of my research, even within the top 1% on the SAT-Mathematics (SAT-M) for talented test takers at age 12, a higher score was associated with a higher likelihood of these students eventually earning a STEM PhD, publication, patent, and university tenure. Additionally, Stephen Hsu and James Schombert used five years of university academic records to show that the probability of success of being at the top of one’s cohort in a physics or math major (but not other majors such as sociology, history, English, or biology) was highly dependent on an individual’s SAT-M score. For example, earning a score of roughly below 600 on the math portion made the probability of attaining a superior academic record in physics or math very low. Perhaps the STEM disciplines have always selected on academic aptitude and employers have rewarded that aptitude and skillset due to STEM’s usefulness in a variety of fields.
Why have education majors consistently been at the bottom?
These data show that US students who choose to major in education, essentially the bulk of people who become teachers, have for at least the last seven decades been selected from students at the lower end of the academic aptitude pool. A 2010 McKinsey report (pdf) by Byron Auguste, Paul Kihn, and Matt Miller noted that top performing school systems, such as those in Singapore, Finland, and South Korea, “recruit 100% of their teacher corps from the top third of the academic cohort.” The US certainly recruits some of its teachers from the top of the aptitude distribution, including at top education schools such as Harvard University and Vanderbilt University. Additionally, Teach for America often selects students from highly selective institutions, which have already filtered students based on academic aptitude.
Andrew Yang, founder of Venture for America, has argued that what top students choose to study greatly influences a society down the road. The McKinsey team stated that closing the talent gap, or following the lead from some other countries and selecting teachers from the high end of the academic aptitude continuum may help improve education for US students. We really don’t know if this strategy would work, but given that the rank order of academic aptitudes for various majors has remained stubbornly constant for the last seven or more decades, it will be extremely difficult to shift what our culture values from traditional STEM (including medical) disciplines to education and teaching, at least in the short term. A recent article by Dan Goldhaber and Joe Walch in Education Next highlights, however, that the SAT scores of first year teachers has recently been on the rise.
How this reflects what US culture values
Why has the rank order of average academic aptitude across various areas been strikingly the same? That remains unclear. For one thing, however, it reflects upon the majors and resulting occupations that US culture has consistently valued for the last seven or more decades. We will have to wait and see if in the next seven decades, this pattern of academic aptitude across majors will change, and if so, in what ways. What majors and occupations future generations of top students choose to pursue directly impacts a nation’s future economy.