The heroes of the COVID-19 crisis are the doctors and nurses, first responders, grocery store workers, and other essential service providers who have been needed during the pandemic as never before.
Though more behind the scenes, another group also belongs on the list: Data scientists. Tens of thousands of them globally are collecting and combing through data to better understand the virus, track its spread, and develop drugs, as well as analyze the pandemic’s effect on the economy, supply chains, and other aspects of our lives.
Data science, of course, already was having a big moment well before the coronavirus hit. Data has become ubiquitous, and innovations like cloud computing and machine learning are providing new opportunities to mine it for actionable insights and enable predictive and prescriptive analytics. As a result, every industry has been growing its quant workforce.
This has created an intensifying skills shortage. According to a report by the tech careers site Dice, data engineer was the fastest-growing job in technology last year, with 50 percent growth in the number of open positions. An August 2018 study by LinkedIn found a shortage of more than 150,000 people in the United States.
As COVID-19 further heightens the importance of and interest in Data Science, the need to develop new data scientists is becoming all the more urgent. This will require change – not just better education and training programs but a fresh assessment of what makes a top-notch data scientist and where to find him or her.
What should the world be doing differently? I have three ideas.
- Look beyond traditional skill sets.
When hiring data scientists, most organizations seek people with degrees or experience in applied mathematics, statistics, computer science, engineering or anything else that requires quantitative analysis. That makes perfect sense, but we should be as imaginative as possible about where good data scientists can come from. The lack of a fancy degree, for example, shouldn’t be a showstopper.
I have advanced degrees, but I didn’t start out as a data scientist. My bachelor’s and master’s are in aerospace, aeronautical, and astronautical engineering.
But during my master’s studies, I was chosen to participate in a project far afield from rocket science: development of new multi-channel polygraph technology that is still in use today. While in the thick of this data-intensive work, which required the development of algorithms and machine learning models to detect anomalies during lie-detector tests, I realized I had found my passion: data science. I’ve been in the field ever since.
Moral of the story: Data scientists may not always come from the typical talent pools. Anyone with the drive and know-how to connect data to insights can be a great fit.
- Enough already: Bring more women into the field.
While gender balance in the workplace has improved dramatically over the decades, the proportion of women in technology has actually dropped. According to the National Center for Women & Information Technology, the number of women in the industry peaked at 36 percent in 1991 and has fallen since. Fewer than one in three people working in data science and other data-oriented jobs is female, Analytics Insight says.
I’m all too familiar with this gender gap in technical fields. I was one of 10 women among 400 students during my first year in the college engineering program. I remember my childhood teachers encouraging the boys toward math and science and the girls toward arts and social studies. While in primary school, I was the only female participant in the Math Olympics.
It’s time to stop the madness. At a time when the world needs all the data scientists it can get, let’s stop pushing half the population away from the field with overt behavior or covert messages that can make women feel unwelcome.
- Seek out the hungry and curious.
Data science for data science’s sake isn’t very useful. A data scientist can be a technical wunderkind, with formidable knowledge about algorithms, models, and programming, but unless they’re intensely driven to find patterns hidden in data to glean valuable and actionable insights and turn them into prescriptive analytics to support decision-making, they’re just playing with tools.
From serving industries undergoing digital transformation to studying pandemics, the best data science has a real mission and purpose behind it. You need to first focus on the problem at hand, and then like a data science project, decompose it into distinct, solvable elements that you can address.
I was once asked in a job interview what I’m passionate about, and I talked about my lifelong love of history – how thrilling I find it to study the past where the history just like the time series data with multi-dimensions, and, just as we do in data science, detect patterns that can reveal lessons for the future. It helped get me the job.
My point is that it’s important for organizations to be thoughtful and creative about the types of people they recruit as data scientists, and that means thinking of data science as an art too. Curiosity matters. An obsession with problem solving matters. These natural qualities can’t be taught. Data science-specific skills can be.
We live in a challenging time when data science has never been more valuable. Let’s make sure we’re attracting the best and the brightest to the field… in the high numbers we need.
