AI did not end the data literacy crisis. It changed it. As machines move from providing information to interpreting it and recommending what to do, the scarce skill is becoming judgment: knowing what to trust, what to question, and what decisions are still ours to make.
Evidence does not make decisions. People interpret it, weigh it, challenge it, and make judgments. A courtroom reveals what leaders can learn about evidence, expertise, confidence, and data-informed decision-making.
Feeling sure doesn't mean you're right. Learn five simple ways to question what feels true, spot what might be influencing you, and make better decisions about what to believe.
A Five-Step Framework for Data-Informed Thinking
This course taught you how to choose the right chart. But visualization is the final step, not the first. Before you ever open Excel, Tableau, or Python, you need a framework for thinking about what you are actually trying to understand. This is that framework.
The loop has five steps: Think, Analyze, Interpret, Decide, Communicate. Each one is a distinct kind of work. Each one has its own questions to ask. Skipping a step (or collapsing two together) is the most common way good analytical instinct gets short-circuited. The loop is not a checklist; it is a cycle. Real analytical work moves between steps, returns to earlier ones when new evidence surfaces, and ends only when you have something to say that earns the visualization at the end.
Use this as a working tool when you are mid-analysis. The Loop section walks through the five steps in order. The Step Library catalogs each step with the questions to ask, what commonly goes wrong, and a short example. The Worked Examples section shows the loop applied to three different kinds of analytical situations. The Quick Reference is a one-page summary you can print and pin to your wall.
Each of the five steps catalogued with its purpose, the four questions to ask, what commonly goes wrong, and a short example. Click a step header to expand its details.
The loop applied to three different kinds of analytical situations: a decision-support case, a diagnostic case, and an awareness case. The same five steps work in all three. What changes is what the output looks like and what it is for.
Do not jump to "what chart shows turnover?" Instead ask:
Use multiple views, not just one chart:
Slow down. Consider alternative interpretations:
What can we now do with more confidence?
Now decide how to communicate this to the HR director:
Resist the urge to plot conversion over time immediately. Ask:
Examine the situation from multiple perspectives:
Consider alternative explanations before concluding:
What actions does this support?
How should you present this finding?
This is the easiest step to skip when no decision is pending. Ask anyway:
Even without a specific decision, the analytical work is real:
The slow decline at one site is the most interesting signal:
The board is not deciding anything in this meeting. So what does this analysis enable?
How does the form differ from a decision-support artifact?
The contrast between the way most people approach data and the way the loop teaches you to. Same starting point, different first move, dramatically different result.
The five steps with their core questions and what to watch for. Use this when you need a fast reminder of where you are in the loop.
| Step | Core question | Key discipline | What to watch for |
|---|---|---|---|
| 1 · THINK | What is this analysis for, and for whom? | Refine the question before touching data | Skipping straight to the chart; naming the vehicle, not the purpose |
| 2 · ANALYZE | What do multiple angles reveal that one would not? | Use at least 2-3 cuts before trusting any pattern | Stopping at the first cut; mistaking visualization for analysis |
| 3 · INTERPRET | What are the alternative explanations for this pattern? | Name at least two interpretations before concluding | Confirmation bias; treating correlation as causation |
| 4 · DECIDE | What does this analysis enable, and for whom? | Reduced uncertainty, not certainty. Name the form of decision support. | Showing data without naming what it enables; assuming all analysis is immediate-decision support |
| 5 · COMMUNICATE | What is the simplest form that serves this audience? | Form follows purpose; lead with the answer when one is needed | Treating every analysis as a chart problem; mismatching detail to audience |
The loop is a cycle, not a checklist. Real analytical work moves between steps, returns to earlier ones when new evidence surfaces, and ends only when you have something to say that earns the visualization at the end.
