What Some Critiques of Data Science Education Get Wrong

A math equation has a clear, correct answer. How to find that answer, on the other hand, is more flexible. It’s the creative journey in math that draws many to the subject and to STEM fields. One answer. Multiple possible approaches.*

Ironically, the debate around math curriculum is similar. 

Most of us agree that we need to solve for our x: improved K–12 education with more opportunities to build mathematical reasoning and access STEM fields, especially for students of color and students experiencing poverty.

Our strategy to solve for that proverbial x is where approaches diverge. 

The present approach to math education has been working for only a small proportion of students, who happen to be disproportionately white, Asian, and male. Black, Latinx, and Indigenous adults remain vastly underrepresented in STEM fields. And the problem begins in school. By the time they reach high school, many students of color have already been tracked into lower-level or dead-end math pathways that don’t lead to advanced math or prepare them to enter rigorous college programs. 

To change this, improving access to high-quality teachers and instructional materials is a must. Strengthening instruction in math sequences that lead through algebra to calculus can enable more students to succeed. But updating K–12 math curricula to provide students more entry points to advanced math is another key strategy. High school math teachers, in fact, have been clamoring for innovative curricula, particularly for students who are not interested in taking a traditional math course or majoring in STEM. 

Data science is a rapidly growing example. Using real-world data and teaching programming languages such as R, high school introductory data science courses are engaging students in quantitative reasoning in ways that students find relevant and meaningful. 

But some STEM professors argue that only traditional high school math courses can adequately prepare students for their futures. Over the next month, we at Just Equations will be breaking down what some of these arguments against high school data science are missing. Here is the first example of a questionable claim:

CLAIM: Data science can’t be taught in high school. Data scientists need to use calculus, statistics, and computer science, so teaching it to high school students who lack mastery in those topics can mislead them into thinking there is a shortcut into data science. 

FACTS: Data science should be treated no differently from biology, chemistry, physics, or computer science. All are taught in high school, and all require multiple prerequisites before a student can earn a degree and become a professional or continue to graduate school. 

When you took chemistry in high school, did you think you were a chemist? Do you worry every spring about the surge in 18-year-olds proclaiming they can practice medicine after passing their AP Biology exam? Or conduct research for NASA after completing AP Physics? 

Of course not. And that’s because most students and families—and certainly most high school teachers—realize that a high school course is an introduction to a topic. In addition to building students’ basic knowledge, that first exposure can spark their interest or tap their talents, giving meaning to their path through future courses. Alternatively, it can lead some who imagined a career in that area to decide their interests lie elsewhere. 

Data science has the same potential, particularly for students who may have been turned off by traditional math courses. Preliminary evidence suggests that some students who take courses such as data science find it so engaging that they are more likely to continue their journey with math than if they had taken, say, precalculus. More research is needed, of course, but that can happen only if innovation in advanced math courses is allowed, not stifled. 

The state of California has encouraged innovation in math, starting with the 2016 California Math Readiness Challenge Initiative, which supported development of new senior-year courses such as Introduction to Data Science, Discrete Math, and Math Reasoning With Connections. Other states, including Virginia, Oregon, and Ohio, are also developing new high school math courses with a similar goal. These innovations need to be evaluated using longitudinal data to understand their outcomes. 

But university faculty need look no further than their own campuses to see the weakness of this argument against teaching high school data science: UC Berkeley, for example, offers a highly popular and well-regarded class called Data 8: Foundations of Data Science. Students can take it in their freshman year to meet their general education math requirement. Students who wish to major in data science need to complete other prerequisites, including three in math and two in computer science. 

Its prerequisite? “On paper, it’s Algebra 2,” one of the teachers, Ani Adhikari, told me when I interviewed her several years ago. “We want them to know that 30 percent is the same as 0.3, and that one-third is greater than 30 percent. But if they don’t, we can teach them.”

That’s not to say data science classes aren’t rigorous. But that’s a discussion for one of our next blogs in the series.

‍

^{* Don’t ignore PEMDAS, though.}