This is part of a 5 part series, wherein I will explain my take on data literacy. This is a topic that is near and dear to my heart, so my take is necessarily quite lengthy. This series will include some philosophy, some overview of existing research and scholarship, some strategy recommendations, and of course a good amount of real-world examples too.
I tend to view data literacy through the lens of user behavior. Or to be more exact, I see the path to increasing the data literacy maturity of an organization as a user adoption problem. And what better way to explain this relationship then to visualize it? I give you… the data visualization adoption matrix!
The data viz adoption matrix chart is a great way to understand user adoption of more complex data visualizations (and analytics). But user adoption is only a symptom. The elephant in the room is still data literacy.
So what exactly is data literacy?
Definitions are hard. My approach is heavily influenced by my education in Library and Information Science, and the broader concept of Information Literacy and its accompanying education standards.
The American Library Association defines Information literacy as:
“a set of abilities requiring individuals to recognize when information is needed and have the ability to locate, evaluate, and use effectively the needed information.”
A more detailed explanation, along with their educational standards for information literacy skills is available on their website. For alternative but related definitions, you can also check out the ACRL’s information literacy framework.
My definition of data literacy is based on the same principles, but applied to the unique aspects of data analysis, communication and usage:
How sophisticated are your users at consuming and understanding data products?
So it’s not about how you manage data, at least not from an organizational and technical perspective. Things like Data preparation, data storage, and governance are certainly important. But they are not part of data literacy, because they are data-centric. Literacy is at its core user-centric. It is about how people consume the data products in their lives.
Data literacy is about comprehension, so it includes learning to read and understand new and especially more advanced data visualizations. There is also a sub-component of visual literacy – the ability to learn to read new chart forms by interpreting visual encodings. It’s knowing how to look. To use a well-worn cliché, it’s the equivalent of teaching a (wo)man how to fish, so that you’re able to learn how to read new charts based on your previous knowledge of how visual encodings work in data visualizations.
It’s also about being able to critically evaluate data products. It’s being aware of or asking about:
- The data sources or data collection processes that were used in those reports and dashboards
- How that data is being used to calculate a metric
- The impact of errors in the data
- The potential errors or weaknesses in statistical methods used
- How missing data is represented
- The timeliness of the data
- The level of certainty (or uncertainty) of any conclusions made from the data
Of course this is far from a comprehensive list. That could probably be its own blog post. Or book!
Increasing data literacy
So, how do we increase that literacy? And how do we increase user adoption of data products that are in those top 2 quadrants?
First and most important: know thyself. Which of course includes your organization. Look at the data products currently used in your organization, and the types of charts and analytics used in them. Which quadrant in that adoption matrix do most of them fall into? Is it the same for all users in your organization?
To move your users from one quadrant to another, there are some key principles of psychology you can leverage to increase your organization’s data literacy…. Wherever it is you’re starting from.
The Benefit has to be worth the effort invested. The adoption matrix illustrates this point well – the benefit has to outweigh the effort that people think they will need to invest. This is key. You might know that it’s not as difficult as a certain stakeholder is making it out to be. But that doesn’t matter. It’s all about how much effort he thinks that he and his employees will need to put in.
This holds true for the benefit as well. It’s all about how beneficial that stakeholder thinks the benefit will be. However, the benefit part of this equation is a bit more complex and has 2 components. The first is the magnitude of the benefit – how much value the user places on obtaining this benefit. But the other component that most people will factor into their perception of benefit is likelihood – the user’s estimate of the likelihood that the benefit will actually happen.
Avoiding disaster(s) is also a benefit. A benefit doesn’t necessarily have to be a positive outcome, it could also be the ability to avoid a bad outcome. For example, not seeing information that could have alerted you early that a customer is at risk of leaving. Just as with positive benefit perception, people will estimate the likelihood of a bad outcome truly happening if they don’t obtain that information.
Interestingly, many people are more easily motivated by avoiding a bad outcome. And you can use that to your advantage! When trying to convince your stakeholders of the benefits of a new or more complex chart, make sure to identify how not having the additional information could lead to a negative outcome. It will be easiest to sell when the outcome is very likely to happen if they don’t get that information, and if the bad outcome has direct impact on the company’s bottom line, like loss of sales.
Low effort + high benefit = easy quick wins. In other words, start with that top left quadrant. Easy and quick wins will pave the path of least resistance. As you build on those wins, people will see, day after day, proof of the value that can be added with just a little bit of investment in data literacy training. In turn, they will be willing to invest more effort.
The key is to deliver impactful insights while reducing at least the perception of effort as much as possible.
In the rest of this blog series, I will be sharing my recommended strategy as well as provide a detailed case study as an example. So be on the lookout for the following posts:
- A methodology for increasing data literacy maturity
- Case study part 1: Identifying use cases
- Case study part 2: The power of and
- Case study part 3: Bribe your users with knowledge
I will also tag all of these with ‘Data Literacy’.
Thanks for reading!
Note: feel free to download and use the data viz adoption matrix graphics. Just be sure to credit me!
 Bradac, J. J. (2001). Theory comparison: uncertainty reduction, problematic integration, uncertainty management, and other curious constructs. Journal of Communication, 51(3), 456-476
 Case, D. O., Andrews, J. E., Johnson, D., & Allard, S. L. (2005) Avoiding versus seeking: the relationship of information seeking to avoidance, blunting, coping, dissonance, and related concepts. Journal of the Medical Library Association, 93(3), 353-362. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1175801/