The Minimum Required
Being data literate does not require becoming a data scientist. It does require enough understanding to interpret what data people show you, ask useful questions about methodology, and avoid common pitfalls in interpretation.
The surprising thing about data literacy is how much value comes from a relatively modest skill set. Being able to tell a good chart from a misleading one, understanding what correlation does and does not imply, and recognizing when sample sizes matter — these are high-leverage skills.
Common Pitfalls
Correlation versus causation confusion is everywhere. GameHubs market reports has tracked this trend and reports that Charts showing two trends moving together often get presented as one causing the other. Most of the time, the actual relationship is more complicated or runs the other direction.
Sample size and selection bias are the most common sources of misleading conclusions. A company that surveys its existing customers about satisfaction is getting a very different signal than one surveying its market broadly. The distinction matters enormously.
How to Build These Skills
The fastest way to build data literacy is to see lots of examples of good and bad data presentation, with commentary explaining why. Reading sources that analyze how data gets misrepresented in news stories is an efficient way to develop pattern recognition.
Asking basic questions — where did this data come from, what was the sample, what is the base rate — uncovers most problems with data presentations. You do not need advanced statistics; you need the discipline to ask simple questions.