Check whether my data shows causation or just correlation
Separates trend, correlation, and possible causation in your analysis and gives you the strongest claim you can honestly make.
The prompt
Help me work out how strongly my data supports the claim I want to make, separating trend, correlation, and possible causation.
Claim I want to make: [e.g. "The training program increased sales"]
Data and analysis: [PASTE - what was measured, for whom, over what period]
How the data was collected: [e.g. "people opted in", "pilot in two regions", "A/B test", "before-and-after comparison"]
What else changed at the same time: [OPTIONAL - other launches, seasonality, staffing, pricing]
Assess:
1. Trend: is there a real movement over time, or is it within normal variation? How many data points support it?
2. Correlation: do the two things move together, how strongly, and in which segments?
3. Causation: test against the common alternatives: reverse cause, a third factor driving both, selection (the group was different to begin with), regression to the mean, seasonality, and coincidental timing.
4. Evidence needed: what would support a causal claim (a control group, staggered rollout, dose-response, comparison with a prior period).
Return:
- A verdict: Trend only / Correlation / Suggestive of causation / Strong causal evidence, with reasons.
- A table: Alternative explanation | How plausible here | How to rule it out.
- The strongest honest sentence I can put in a report or slide, plus one sentence I should avoid.
- The cheapest next test that would raise confidence.
Do not overstate the evidence. If the data cannot tell the explanations apart, say so.
Highlighted [brackets] are blanks to replace with your own details.
Fill in the blanks and use this promptMore Data & Analysis prompts
What to fill in
- e.g. "The training program increased sales"
- PASTE - what was measured, for whom, over what period
- OPTIONAL - other launches, seasonality, staffing, pricing
How this prompt is built
This prompt follows the Few-Shot structure (Few-shot prompting). A few-shot prompt includes one or more worked examples of the input and the output you want, so the model can copy the pattern instead of guessing it.
Related prompts
- Identify the leading and lagging indicators for my business or teamSorts your metrics into leading and lagging indicators, links each lead to the outcome it predicts, and fills gaps in early-warning measures.
- Group students for intervention from screening and assessment dataSorts students into MTSS or RTI tiers and intervention groups from universal screening and diagnostic results.
- Diagnose where our college admissions funnel is losing studentsAnalyzes inquiry-to-enrollment conversion, yield, and summer melt to find where prospective students drop off.
- Analyze my clinic's claim denials and revenue cycle KPIsTurns denial and A/R data into root causes, benchmark gaps, and a prioritized fix list for the front end, coding, and billing teams.