How Brainy Is Your Major?

First, an important disclaimer: the data I'm about to present to you is based on averages. A group average does not imply that everyone who is a part of that group is at the average. Therefore, this data should not be used to indicate that someone who majored in subject A is necessarily smarter than someone who majored in subject B, because that is a comparison of two individuals. To adequately make that comparison, we would need their individual scores.

Data from a nationally representative stratified random sample

The following graph is based on Project TALENT, a stratified random sample of the nation's high school population (a sample of over 400,000). In some of my research along with my colleagues David Lubinski and Camilla Benbow of Vanderbilt University, we examined the group of individuals from this sample who had earned a higher educational degree (a bachelor's, master's, or Ph.D.). We looked at measures of verbal, spatial, and math ability (labeled as V, S, and M below). The y-axis indicates the specific ability level and the x-axis (along the bottom of the graph) indicates the general ability level of each group. General ability was determined by taking a composite of the verbal, spatial, and math ability scores. The data are represented in z-score or standard deviation units. For example, social science majors were about 0.8 standard deviations above the mean of the entire sample, whereas engineering majors were about 1.3 standard deviations above the mean of the entire sample. This makes sense because people who earn higher educational degrees are usually above average in ability.

Here is the rank order of the groups examined:

1. Engineering
2. Physical Science
3. Math/Computer Science
4. Biological Science
5. Humanities
6. Social Science
7. Arts
8. Business
9. Education

Data from the Graduate Record Examination (a select sample)

So, are these results replicated in other samples? The answer is yes. We also examined data from the Graduate Record Examination (GRE). But first, let's address a key question you may be asking yourself: does the GRE really measure general ability?

It turns out that Meredith Frey of Otterbein University and Douglas Detterman of Case Western Reserve University have demonstrated in a paper published in the prestigious journal Psychological Science that the Scholastic Assessment Test (SAT) to a large extent measures general ability and that SAT scores can even be translated into IQ scores. Along with their colleague Katherine Koenig, they have also demonstrated similar findings for the American College Test (ACT). Because these standardized tests measure general ability to a large degree, it would make sense that the GRE would as well, although that still needs to be empirically demonstrated. However, regardless of whether the GRE measures general ability or not, the fact that we see the same pattern in a select sample (GRE) in addition to a nationally representative stratified random sample (Project TALENT) illustrates that the average score patterns for majors are quite robust.

When taking the GRE, individuals are asked to indicate which major they intend to pursue for graduate study. Keep in mind that the GRE is a select sample of the population because only a subset of individuals who have graduated from college take this test. The graph below takes an average of the math and verbal subscores to indicate general ability. As can be seen below, the rank order of the major groups is essentially preserved, with minor variations (arts are now ahead of social science).

Here is the rank order of the groups examined:

1. Engineering
2. Math/Computer Science
3. Physical Science
4. Biological Science
5. Humanities
6. Arts
7. Social Science
8. Business
9. Education

So what does this all mean?

Let me give you an anecdote. An individual who saw this data for the first time immediately exclaimed: "Oh, this means I'm smarter than my brother!" (In his head he probably was thinking that he finally had data to confirm what he knew all along). He had majored in math and his brother had majored in a social science. I had to point out to him that this data could not be used to necessarily conclude that he was smarter than his brother, just that on average, math majors are smarter than social science majors. His brother could have been a super smart social science major and he could have been a less smart math major. There is wide variability within each group and much overlap across groups. What was needed in that case to determine who was smarter were their individual ability scores.

So remember: group averages don't necessarily give accurate information about individuals. The flip side of this is that if you had no information about a person other than the fact that they were a math or social science major, it is more likely that the math major is smarter than the social science major. But if you then found out the social science major earned their degree from an elite university and the math major earned their degree from a community college, then perhaps you might then adjust your expectation that the social science major might be smarter, because the level of institution attended also gives you a rough indication of the individual's ability level.

Keep in mind that there are extremely smart people in all disciplines and at all colleges and universities. And perhaps the smartest social scientist is as smart as the smartest mathematician. However, that is a hypothesis that is in need of an empirical test.

Finally, let's not forget a critical factor: what someone chooses to major in has a lot to do with their interests.

© 2011 by Jonathan Wai

You can follow me on Twitter, Facebook, or G+. For more of Finding the Next Einstein: Why Smart is Relative go here.

The initial graphic was created by the artist and website designer Robert Dawson. You can find his ipainting here and his oilpainting here.