Data Interpretation Guide

The Prompt

Interpret this data and explain what it means:

[PASTE DATA - tables, charts described, or statistics]

Provide:
1. **What the data shows** (factual description, no interpretation)
2. **What it likely means** (reasonable inferences with confidence levels)
3. **What it does NOT prove** (common misinterpretations to avoid)
4. **Anomalies** (unexpected values and possible explanations)
5. **Missing context** (what other data would strengthen conclusions)
6. **So what?** (actionable implications for decision-making)

Rules:
- Correlation ≠ causation — always note this where relevant
- Distinguish between statistically significant and practically significant
- Flag small sample sizes
- Note if survivorship bias could be at play
- Present confidence as a range, not a point estimate

How To Use It

Paste any data — survey results, A/B test outcomes, sales figures, or research findings. The AI will translate numbers into narrative while maintaining intellectual honesty about what the data can and cannot tell you.

Why It Works

Data literacy is rare. Most people either over-interpret data (treating correlations as proof) or under-interpret it (ignoring clear signals). This prompt enforces the discipline of separating observation from inference and always flagging limitations.

Variations

A/B test interpreter:

Interpret this A/B test result. Is it significant? What's the effect size? Should we ship it? What could be wrong with the experiment design?

Dashboard narrator:

Look at these dashboard metrics and write a weekly narrative: what moved, why it likely moved, and what needs attention this week.

Correlation investigator:

These two variables appear correlated. Brainstorm: 5 causal mechanisms, 3 confounding variables, and 2 experiments that could establish causation.