wowprompt

Correlation or causation check for a claimed data finding

Tests a claim like "customers who use feature X spend more, so X causes higher spending" against the evidence, and lists the other explanations to rule out. Use it before a finding from a report goes into a decision, a deck or a press release.

A verdict on whether the evidence supports a causal claim, a table of alternative explanations with how to test each, and a rewritten claim that matches the evidence
The output this prompt produced, unedited.

The prompt

Fill in these first

  • The claim: [e.g. "Users who join our webinar are three times more likely to buy, so webinars drive sales"]
  • The data behind it: [paste the numbers, the comparison groups and the time period]
  • How the groups were formed: [e.g. users chose to attend, randomly assigned, all users before and after a change]
  • Who will use the finding and for what: [e.g. leadership deciding the marketing budget]

Task

Judge how strongly the evidence supports the causal claim, and say what would make it stronger.

Output format

  1. Verdict: supports a causal claim, suggests a link worth testing, or does not support the claim. One paragraph of reasoning.
  2. | Alternative explanation | Why it might explain the result | How to test it |

Cover selection bias, confounding variables, reverse causation, chance, and timing effects such as seasonality.

  1. What is solid in the finding, in two or three bullets.
  2. A rewritten claim that says only what the evidence supports.
  3. The cleanest test to settle it, such as a randomised trial, a holdout group or a before and after comparison with a control.

Rules

  • Do not recalculate numbers that are not given; work with what is pasted.
  • If the groups were randomly assigned, say that this rules out most alternative explanations.
  • Keep the language plain enough for a non-analyst to follow.
  • Never call a result "proven".

Example

Rewritten claim: "Webinar attendees bought at three times the rate of non-attendees. Attendees chose to join, so part of this gap may reflect existing interest."

correlation vs causationconfounding variablesdata claim checkcausal inferenceanalysis reviewmisleading statistics

How to run this

Copy the prompt and paste it into Gemini, then set the aspect ratio to 3:4. No reference photo is needed for this one.

The prompt is written in Markdown with each part under its own heading, which is what the copy action hands over. Change one section at a time, keeping the rest intact, so you can tell which change did the work.

Related prompts

All productivity