The Death of the Citizen Data Scientist

For years, companies big and small in the analytics, data science, and machine learning space have been advocating for the citizen data scientist. I’m sad to say that the concept is more or less dead.

What is a Citizen Data Scientist?

As far back as 2015, IBM (disclosure: Trust Insights is an IBM Business Partner) advocated for what was called a citizen analyst – a member of the general public, armed with simple, easy to use tools like Watson Analytics, who would use their curiosity to analyze data and create insights for social good. In the years since other companies like Alteryx have promoted similar ideas with the concept of the citizen data scientist.

The hope these companies had was the democratization of analytics and data science – analytics for everyone. Data science for all. People who were curious and motivated to investigate phenomena like elections, public health issues, or even fun things like Pokemon character references would pick up any of the modern tools and dive into the available data, exploring it and revealing its mysteries.

The promise of democratized analytics was compelling – by offering low cost or free tools to the general public, we would jumpstart the talent pool necessary to fulfill the deficit of skilled workers in analytics, data science, and AI. In turn, this would also create rich economic opportunity for those who took up the charge, as data science job openings and salaries continue to skyrocket.

So, what happened?

The Attributes of a Data Scientist

The prepending of “citizen” on the role of data scientist doesn’t change the nature of the role. A data scientist is four things:

• A curious, motivated seeker of knowledge using the scientific method
• A domain expert of some kind
• A technical practitioner comfortable working with code and data storage technologies
• A mathematical practitioner comfortable and familiar with statistics, probability, and advanced mathematics such as linear algebra

It’s no wonder data scientists are so rare and expensive; they are essentially four careers rolled into one.

So why did the citizen data scientist concept fail? In some ways, the general public has opposite attributes of data scientists.

Incuriosity

Citizens today seem less curious than ever. Especially around hot topics such as climate change, public health, wages, firearms safety, fossil fuels, whether the planet is round or not, citizens have instead embraced what can only be described as hyper-partisanship. Beliefs have become so rigid and uncompromising that they are comparable to, or even exceed, religious zeal. Combine this with massive information overload, and few people have the motivation to dig deep into a topic and analyze the data around it. Some people are so entrenched in their beliefs that very idea of exploring that their beliefs could be factually incorrect is anathema to them. They would rather feel right than be right.

Ubiquitous Satisfaction

Data science is work. Make no mistake about it, even casual exploration of a topic can feel like the same work one performs at a 9-to-5 job – extraction, cleansing, analysis, and reporting of data. Citizens today have far easier ways of obtaining momentary satisfaction; on the smartphone in our pockets is a nearly unlimited buffet of world-class entertainment. Those who have some analytical skills often are so tired or run down from their day jobs that the last thing they want to do in their free time is do more work, even if it’s for something they are otherwise passionate about.

Educational Failure

The educational system has failed. In many places – not all, but many – citizens lack the fundamental knowledge needed to perform data science well. Even with the best tools and low-code/no-code environments like IBM Watson Studio, citizens still need to know what they’re after and what the outputs of software mean. Many citizens, especially in the United States of America, have little or no understanding of the mathematics behind data science such as statistics, probability, linear algebra, or calculus – essentials for knowing how the different techniques in data science work. The techniques that would often lend insightful answers are therefore out of reach, even in the best, easiest tools – not because the tool can’t do it, but because the uneducated user doesn’t know to ask for it.

Combine that with an equal lack of skill in the data and technologies needed to make data science work at more advanced levels (which are required for the really big questions) such as SQL databases, JSON data structures, and multidimensional array management, and it’s no wonder citizens don’t attempt data science.

What Comes Next?

So, if the citizen data scientist is but a dream, what next? How do we continue to encourage citizens to be active participants in the use of data? How do we broaden the embrace of data-driven business in a world which values facts less and less? Existing data scientists and analytics professionals instead have to adopt a fifth skill: the skill of data storytelling, of using data to craft compelling narratives that make our data easier to work with, easier to understand, and easier to take action on. That skill, when executed well, motivates people to slowly and cautiously take the first step towards embracing data, facts, and truth.

For those causes and positions that need the help, it’s unlikely we’ll find all the skills of a data scientist in one person, but those skills can exist individually in many people. If we intend to use data for good, we can instead put a group together. Like a band, each person brings their specialty and as long as someone can conduct the individuals, we can achieve similar results.

And finally, for those few people who are data scientists professionally, it’s our obligation to keep educating the world and as time permits, be the citizen data scientist that others dreamed of years ago. The truth is out there, and it’s in our data.

Christopher Penn
Chief Data Scientist