Counts show all available guides. Filters combine to narrow your results.
Simplify a complex taskGuide
Tell an association from a causal explanation
Find what a made-up observational dataset supports and propose what would still be needed for a causal claim.
Illustrative resourceNot tool-testedAdults who want a practical, low-stakes way to tell an association from a causal explanation
What you’re working towards
Find what a made-up observational dataset supports and propose what would still be needed for a causal claim. The worked result is a written example for practice; the AI steps have not been tested.
Your choice also saves this guide’s current worked-example checkmarks. Only this browser; no account or sync. View saved guides
Saving needs JavaScript. Reading and downloading this guide does not save it.
Make a start
One step at a time.
3 steps
Start with the synthetic worked example or other non-sensitive inputs. Run one prompt at a time in your chosen AI tool. Replace any bracketed placeholders with the requested material, check each reply, and supply the source or earlier outputs that step needs. Paste them again if the tool lacks that context.
Start with the synthetic example, then adapt each prompt to your task.
01
Open a chat and paste the example
Open https://chatgpt.com and choose New chat. Copy every line under Input in this guide into the message box. Then add the prompt below (the instructions telling the assistant what to do) and send both together. Expect a written response shaped like the example under Output. The wording may differ. If the assistant asks for information already in the input, point to that line; if a necessary fact is missing, keep it unknown rather than guessing.
Your starting prompt
Assess the causal claim using only the four session records and definitions above. Calculate the two averages, separate what is observed from what is unproved, and rewrite the claim carefully. Give one question about missing context. Label possible alternative explanations as possibilities, not facts.
02
Check the specific result
Verify averages of 13 and 6. Check that the rewrite is limited to these four sessions and that the assistant does not turn a possible explanation into a known cause. Compare it with the Input yourself. If the response fails a check, paste the correction prompt below into the same chat. Recheck the corrected version before saving it.
Your starting prompt
Recheck the averages from the four records and distinguish association from cause using the supplied definitions. Remove any inferred motive or claimed causal proof. Keep alternative explanations explicitly hypothetical.
03
Try it with your own example
Open a new chat. Copy the prompt below, then replace [PASTE YOUR INPUT HERE] with your own non-sensitive facts and rules. Include the kinds of details listed under Requirements. Send the whole message. Expect the same kind of result as this guide, using your values. Compare every name, number and rule with your input. If something differs, point to the exact mismatch and ask for a corrected complete answer. Save only a version you have checked.
Your starting prompt
Use only my records, group labels, definitions and claim below. Calculate the group averages from the supplied numbers. Separate the observed association from an unproved cause, rewrite the claim carefully, and ask one useful question about missing context. Mark alternative explanations as possibilities, not facts.
Use only the new input below. If a fact needed to finish is missing, ask me for it before choosing an answer.
MY INPUT
[PASTE YOUR INPUT HERE]
Optional: improve the result
Ask what new information would change confidence
Think about better evidence without claiming the cause can be settled by one extra column. This extension uses the worked example. Return to that example chat, or paste its input and your checked example answer into a new chat first. Do not apply its sample values unchanged to your own input.
Advanced prompt
List three additional non-sensitive facts about each session that could help interpret this pattern. For each, explain what uncertainty it addresses and why collecting it still would not automatically prove that sunshine caused attendance.
Show, then try
See the shape of the result.
This input is synthetic. The output is a written illustration of the intended result, not a verified AI-tool run.
01 / The starting point
Synthetic input
Made-up example for practice. A neighborhood sketch club recorded four sessions: session A, sunny, 12 people attended; B, rainy, 5 attended; C, sunny, 14 attended; D, rainy, 7 attended. No one assigned the weather, and only attendance and weather were recorded. Claim to assess: “Sunshine caused more people to attend.” Correlation means two things vary together; causation means one produces a change in the other. Use the four observations only. Do not infer participants’ motives or claim a real research result.
Copies text only. For image or file tasks, use the source files named in the steps too.
Written example for practice; not a recorded AI response.
What the observations show
Sunny sessions: (12 + 14) ÷ 2 = 13 attendees on average.
Rainy sessions: (5 + 7) ÷ 2 = 6 attendees on average.
In these four sessions, sunny weather occurred alongside higher attendance. That is an observed association, also called correlation.
What they do not establish
The records do not show that sunshine caused the difference. Other differences between sessions, such as scheduling or promotion, were not recorded and could matter; these are possible explanations, not established facts.
Careful rewrite
“In this four-session record, average attendance was 13 on sunny days and 6 on rainy days. The record alone does not establish why.”
A useful next question: were the session days, times and promotion comparable? More observations can help describe a pattern but do not automatically prove a cause.
These checks use the worked example and its particular facts. With your own material, check your actual names, numbers, sources and rules; do not reuse the sample’s values as the answer.
Unsaved checks stay in this tab and reset when you reload. If you choose Save guide, its current example checkmarks are kept in this browser when storage is available.
!
A few things to watch for.
A larger difference is not by itself stronger proof of a cause. How the information was collected matters.
Adding a plausible story about people’s behavior does not add evidence.