A data scientist's view on data literacy.

Long time, no see – definitely! I need to admit that I feel a little bit ashamed of my last activity here. Almost half of a year with neither blog post nor Instagram one. I was far more proud of myself in the previous year, but you know what? I believe there is the right time for everything and that’s also something I’ve need to learn. Each of us has equal amount of time a day. It’s us to decide where to allocate this energy. If I haven’t allocated majority of mine into my private life, I wouldn’t be a happy married woman now, enjoying every single memory of a dream wedding. Now, I feel with my whole heart, that it’s the right time for me to be back. What’s more important, I have mental resources for that. 🙂

What’s in the agenda of DataScientistDiary for next weeks?

Of course, though being silent, it doesn’t matter I have been out of the industry! I still did my job, read books, listened to hundreds of podcasts episodes and had many ideas and reflections. One of the cool things I have been working on and it’s still high on my agenda is citizen data science concept. As you might now, I am responsible for data science community leadership in my current workplace. Apart from ML projects development and support, I am priviliged to enable future citizen data scientists to grow under the umbrella of our organization’s Data Academy. Every person, every project is completely different. I love this job because I can scale the amount of projects, topics I am involved in and learn with my trainees! Flight risk, sustainability, workload prediction or price elasticities – each business user brings something new to the table.

But end of private stuff! It took me some time to figure out what kind of blog article should I choose as first after the break. “Why not to start with something you’ve been most involved in past months?” – I’ve asked myself. That’s the point! This way, we’ll start from the fundamentals, defining and exploring many of fancy terms which you probably spotted on LinkedIn, job adverts, podcasts or other articles. How to become data-driven? What does data literacy mean? How citizen data science differs from regular data science? Who is it for? Let me invite you to the beautiful journey of tomorrow and begin with the importance of data literacy.

What’s data literacy?

According to Gartner’s definition:

“Data literacy is the ability to read, work with, analyze, and communicate with data in context, including an understanding of data sources and constructs, analytical methods and techniques applied, and the ability to describe the use case, application and resulting value.”

These days, technology evolves so quickly, people need to be able to effectively consume and use data not only for career but also personal life. Data literacy allows organizations to have strong adoption of data and analytics. Coding and data visualizations are necessary, but even more important are the non-technical skills such as interpreting data and communicating insights with it. Organizations need people who can bridge the technical side to the business side. It’s data literacy what provides people with the skills necessary to have confidence in consuming data effectively.

Why is data literacy important?

I hope you already noticed that we can treat data literacy as some kind of shift in people’s mindset. But why is it so necessary? Companies all over the world are constantly seeking ways to differentiate themselves from their competitors, by investing in data as an asset. When volumes of data increase so rapidly, it’s critical for businesses to identify useful signals to make decisions quickly. Employees need to possess data literacy and skills to discover insights, trends, and patterns relevant to solving real business problems.

After gaining the right skills and improving data literacy, employees can apply techniques to derive meaning from data and communicate their discoveries e.g. to grow sales performance or make their processes more efficient. They can ask the right questions to achieve strategic goals, collect and connect the proper data to those questions, determine whether or not information is reliable, and apply this knowledge with relevant business context.

How to become data literate?

In fact, everybody is data literate to some extent. We read data on a daily basis but we need to build the confidence to be on this journey – be able not only to read data but also communicate it or adapt the tools. Everyday tasks like tracking burnt calories on Apple Watch or determining if we should pack umbrella going outside are both examples of data consumption, even if we’re not fully aware of it.

However, in a business context, people become intimidated. Due to the cultural aspects, many people lack confidence in their data literacy. It’s important to remember that this is a progressive journey to become more data literate. It’s hard to indicate the beginning of this learning path as well as the finish line. It’s an iterative process.

4 levels of analytics we all need

When we think about data trends, we think about the big fancy words like machine learning, big data or AI. But aren’t them all just about helping you make smarter, more well-informed decisions? It’s not just access to data that makes you smarter, it’s the way you analyze it. That’s why it’s important to understand all four levels of analytics: descriptive, diagnostic, predictive and prescriptive.

The right tools make things easier as per Apple Watch example. But does it mean it can make people lazy? Absolutely not, as the visualization is not the end game. It’s just the first level of analytics (descriptive). From that level, there are another steps. That’s where decision making begins. Let me showcase below.

(1) With Apple Watch and connected applications I can track my daily level of physical activity (descriptive). Am I making recommended 10 thousands steps a day? What is my heart rate at night? How many hours of deep sleep did I take last night? How many workouts have I done this month?

(2) Then I can ask myself why is it not on the satisfactory level (diagnostic). Is my steps amount at a low level because I am working remotely? Or maybe, I am going everywhere by car because it’s comfortable? Why have I slept so poorly last night? Is it because I finished my workout very late and couldn’t fall asleep afterwards?

(3) What can I predict based on this (predictive)? Am I likely to gain/loose some weight in some period if I compare the number of burnt calories to the ones I am intaking? Or maybe in 20 years I will have troubles to run after my grandkids because of hurting back and worse blood test results?

(4) Finally, what do I need to do to improve the situation (prescriptive)? Would adopting a dog make me more healthy person? Or maybe decreasing daily amount of calories by 300 would make me fit into my favourite pair of jeans again? Would adding more protein to my meals make me less hungry all the time?

Tools enhance our actions. I can avoid repetetive, tideous task and focus on deeper work. They won’t solve all your problems for you. Taking Tableau as an example – it’s great, but we still need storytelling skills to communicate observed patterns. We need people to make decisions.

Does everyone need to become data literate?

“I am not the manager. Should I even think about this? Wouldn’t it be better for me to focus on my tasks instead of trying to understand organizational strategies like that?” Well, data literacy is definitely not a nice-to-have skill. Everyone need to constantly develop in this area. If you don’t want to stay behind, you need to go with the flow. Executives need to understand how to use the tools and for what purpose. Upskilling is a priority for every company today, and the ability to read, analyze, and communicate with data is vital to success in every industry, every position, today. To succeed in the digital world, one must be comfortable with data. This means that everyone should be data literate.

However, not everyone need to reach the top level. Not every employee is required to learn coding in Python or creating complex Power Bi dashboards. Majority (even over 90%) of the organization will remain on level 1 or 2 of the awareness. Then there are the nerds who will be eager for technical development. Together they form the organization. But when you develop speaking with data, this is a progression, being comfortable with all levels. Holistic way of looking at and using data.

What are the criteria of evaluating data literacy?

To understand the value of data literacy assessments you must first understand the nature of what is being assessed. Unfortunately, there are no tech criteria for data literacy evaluation. Having data strategy is a starting point. Do you know what you want to achieve in the organization? What tools do you currently have? What other tools do you need? Building the roadmap is the answer and it should lay in the hands of data and analytics leaders. They are in scope of creating this narrative and highlighting the business value which could be gained.

It’s a common saying that data science is not just about fancy algoritms, but more about solving problems and ability to ask the right questions. Similarly there, evaluating data literacy level won’t be possible without defining the right questions, criteria. Below you can spot some examples:

How many managers can explain the outcome of their processes?
How many data scientists are able to explain their model performance measures?
How many business users have we upskilled to become citizen developers?
…

All depends from a vision and the impact you’d like to make.

What are the examples of good data literacy strategy?

All things considered, have you fully evaluated your current level? Have you spotted the skill gaps across your people? Then let’s make something to make the situation better and roll out some data literacy training programs! What could be better for those confused in data driven reality than building a safe environment to gain data literacy knowledge and skills? It needs to be a part of a organization culture! It needs to be fun! Everyone need to understand the benefits of becoming data literate. They should have clear answers to “what’s in it for me?” and “how it relates to my current or future role?”.

Think outside the box. Go with fail fast approach. Let people be creative and not afraid of failure. Leverage games, workshops over boring presentation decks. Enable communities. Inspire by examples. Make sure all roles are defined. Identify all the personas (e.g. data owners, data stewards) and their responsibilities, investigate who can help you to overcome this communication barrier. Last but not least – lead by example.

In my opinion, subscriptions to education and training platforms like PluralSight or Coursera can fasten the evolution as it’s easier to fit the needs of our nerdy group with high ambitions to learn. Nevertheless it shouldn’t be the only option. That is because learning is much more effective when it is social (done with others), personalized (done with expert feedback), and contextual (connected directly to the business problems you are solving). From my own experience I can say, that focusing too much on courses than practical application of new skills within projects, can be some sign of procrastination. People need to leave their comfort zones and that’s why fail fast approach is so important. Only all the methods combined (e-learning, communities of practice, hands-on sessions, side-by-side learning and more) enable people to grow with the satisfactory pace.

Summary

If you had only one thing to remember from this article, I want you to bear in mind that data is not a vertical — it is not just one job family, like a data scientist or data engineer. Instead, data is a horizontal which means it is a skillset that cuts across a growing number of jobs in every field. A salesman is a better marketer with data skills. A product manager is a better product manager with data skills. And so on for logistics, engineering, marketing, and even HR. Not everyone needs to know how to code. But soon everyone will need data literacy.

Ultimately, data literacy is about much more than machine learning and data science. And it’s about more than AI. Data literacy is simply about people coping better in a the world overflowed with data — which is why we need it now more than ever.

Summing up, I want you to ask yourself now and answer in the comment if you have the courage and willigness to share: Would you consider yourself data literate? What can you do today to be more data literate tomorrow? 🙂

A data scientist’s view on data literacy – what is it and why it should matter to you?