Bridging the Data Divide
Akin to the spread of reading, writing, and arithmetic at the beginning of the 19th century, data literacy has become a prized attribute as organisations and societies seek to transition toward evidence-based decision making. So, what is data literacy? Data literacy can be thought of as the ability to
“read, understand, create and communicate data as information.”
The transition, though, is proving challenging with a gap in the level of fluency required and the current state of affairs. Closing the data literacy gap need not be mystical. The following is a breakdown of the knowledge, skills, and attitudes I have found helpful when reasoning with data.
Start with knowledge. Knowledge is familiarity with facts and information gained through study or experience. Data literacy begins with a grasp of the basic building blocks of data. Knowing how to distinguish between the fundamental types of data, such as categorical and numerical variables, discrete vs continuous values, and ratio and interval data, is vital. It informs the kinds of groupings and aggregations that can be performed on subsequent analysis.
The application of knowledge requires skill and is the second set of desirable traits. Chances are you will have already come across at least one table, chart, graph, or map today. The data-literate read visual representations of data created by others and interact with it to answer questions and formulate new ones. Relatedly, data-literate individuals can create clear and compelling visualisations that reveal insight to others. This often involves choosing appropriate chart types, making sound design and layout choices with the deliberate use of colour and annotation to aid comprehension.
The third category of traits can be organised around attitudes. Attitude is the approach adopted to a particular problem and are patterns of thinking that influence behaviour. And although attitudes stem from knowledge and skills, they are also shaped by our interactions with the world and others. Chief among the attitudes that build the data literacy muscle is the confidence to dive into a relevant data set and apply different techniques to find answers to questions. With that being said, the data-literate also approach new pieces of information with a dose of scepticism and are alert to the common pitfalls when working with data, from epistemic errors to mathematical miscues. Accepting that the underlying data is imperfect and incomplete to a certain degree, data-literate individuals seek to capture and document known issues and uncertainties, committing to making incremental improvements in the underlying data and the associated analysis.
These are just some of the characteristics required to wrestle insight from data, separated into knowledge, skills, and attitudes. While indexing highly on each vertical is a work-in-progress as the volume and variety of data proliferate, focussing on developing the fundamentals can empower one to participate in data-driven conversations.