Training in statistics, linear algebra and algorithmic thinking is more relevant for today’s educated workforce.
This article is written by Data Incubator founder Michael Li and Columbia University Professor Allison Bishop. It was originally featured on The Wall Street Journal.
Can you remember the last time you did calculus? Unless you are a researcher or engineer, chances are good it was in a high-school or college class you’d rather forget. For most Americans, solving a calculus problem is not a skill they need to perform well at work.
This is not to say that America’s workforce doesn’t need advanced mathematics—quite the opposite. An extensive 2011 McKinsey Global Institute study found that by 2018 the U.S will face a 1.5 million worker shortfall in analysts and managers who have the mathematical training necessary to deal with analysis of “large data sets,” the bread and butter of the big-data revolution.
The question is not whether advanced mathematics is needed but rather what kind of advanced mathematics. Calculus is the handmaiden of physics; it was invented by Newton to explain planetary and projectile motion. While its place at the core of math education may have made sense for Cold War adversaries engaged in a missile and space race, Minute-Man and Apollo no longer occupy the same prominent role in national security and continued prosperity that they once did.
The future of 21st-century America lies in fields like biotechnology and information technology, and these fields require very different math—the kinds designed to handle the vast amounts of data we generate each day. Each individual’s genome contains more than three billion base pairs and a quarter of a million genomes are sequenced every year. In Silicon Valley, computers store over 100 GBs of data—more information than contained in the ancient library at Alexandria—for every man, woman and child on the planet.
Accompanying the proliferation of new data is noise, and a major job for data analysts and scientists is to tease out true signal from coincidence and noise. Knowing when a result is due to chance versus when it is statistically significant requires a firm grasp of probability and statistics and an advanced understanding of mathematics.
We no longer think of outcomes as being triggered by a single factor but multiple ones—possibly thousands. To understand these large and complex data sets, we need an educated workforce that is also equipped with a firm understanding of multivariate mathematics and linear algebra.
Computers and computation are ubiquitous and everyone—not just software engineers—needs to learn how to think algorithmically. Yet the typical calculus curriculum’s emphasis on differentiation and integration rules leaves U.S. students ill-equipped at posing the questions that lead to innovations in computation. Instead, it leaves them well-equipped at performing rote computations that can be easily done by a computer.
We’re not saying calculus shouldn’t be taught. Calculus, like any rigorous technical discipline, is great mental training. We would love for everyone to take it. But the singular drive toward calculus in high school and college displaces other topics more important for today’s economy and society. Statistics, linear algebra and algorithmic thinking are not just useful for data scientists in Silicon Valley or researchers for the Human Genome Project. They are becoming vital to the way we think about manufacturing, finance, public health, politics and even journalism.
It is time to ask ourselves: Is calculus really the best choice to serve as the culminating mathematical experience for a vast majority of students or are they, and society, better served by other mathematical subjects?
Editor’s Note: The Data Incubator is a data science education company. We offer a free eight-week Fellowship helping candidates with PhDs and masters degrees enter data science careers. Companies can hire talented data scientists or enroll employees in our data science corporate training.