Ask a testable question

The idea

A claim or prediction is made scientifically testable by asking about measurable evidence and past patterns; questions with simple yes/no answers are usually not so useful.

Dark clouds

Varsha tells her friend Meghna, “It will rain this afternoon because the clouds look dark.” Meghna wants to test this.

Which kind of question will help most?

• Do you really think so?
• What is the humidity? Did it rain at this humidity before?
• Are the clouds dark?

Sort the questions

Sort Meghna’s possible questions: good (measurable evidence, past patterns) or not so useful (yes/no).

Ask for numbers and patterns

A claim or prediction becomes scientifically testable when we ask about measurable evidence and past patterns. Questions with simple yes/no answers are usually not so useful.

Questions like these go beyond “the clouds look dark”.

• Good question | What it uses
• What is the humidity today? | a measurement
• What was the sky like the last time it rained? | a past pattern
• Is the temperature dropping like before? | a measured change and a past pattern

Notes

A prediction is testable when we ask about measurable evidence and past patterns. Simple yes/no questions are usually not so useful.

• Ask this | Not this
• What is the humidity? | Are the clouds dark?
• Did it rain at this humidity before? | Will it rain?

Check yourself

Which question helps test “It will rain this afternoon”?

Why are simple yes/no questions usually not so useful?

Claim: “Food becomes harmful during an eclipse.” Which question helps test it?

Meghna checks humidity on 5 earlier rainy afternoons: it was above 80% on 4 of them. On how many was it not above 80%?

Answer: 1

5 − 4 = 1. Numbers like these are measurable evidence from past patterns.

• Will it rain?. That repeats the claim.
• Was the humidity above 80% the last time it rained? — correct. Yes. It asks for a measurable number and a past pattern.
• Do you feel it will rain?. A feeling is not measurable evidence.
• They do not give measurable evidence to compare with past patterns — correct. Yes. A number or a pattern gives more to test.
• Because yes and no are difficult words. The problem is the lack of evidence.
• Because they are always wrong. They can be right, but they tell us little.
• Do many people believe it?. Belief is not evidence.
• What physical change happens during an eclipse? Does the temperature change a lot? — correct. Yes. These ask about measurable changes.
• Is the eclipse exciting?. That is an opinion.