🙃 When your hypothesis is wrong

Why wrong is good

A hypothesis is a guess with a reason. You test it because you don't know the answer. If you already knew, there'd be no point testing.

When the numbers say no, you've found out something you didn't know before. That's what science is for. Scientists are wrong all the time, and they write it down honestly, because the next question starts right there.

Look at what the judge's scoring card asks for in a top conclusion: "Uses the numbers; honest when the hypothesis was wrong." Nothing there says you have to be right.

A surprise, with made-up numbers

Say you guessed that more water would sprout more beans, because seeds need water. You gave the beans none, a little (1 teaspoon), medium (1 tablespoon) and lots (3 tablespoons), with three bags of five beans at each amount. These are made-up means of the three bags:

• None: 0 of 5 sprouted.
• A little (1 teaspoon): 3 of 5.
• Medium (1 tablespoon): 5 of 5.
• Lots (3 tablespoons): 2 of 5.

Your hypothesis said more water, more sprouts. The numbers went up, then came back down at Lots. So your hypothesis was partly right. Partly right is one of the three boxes on the conclusion sheet for a reason: it's one of the most common answers in real science.

The conclusion frame

The conclusion sheet asks three things, in this order:

• My hypothesis was right, partly right or not right. (Tick one.)
• I know because… (point to the means that show it).
• What might have gone wrong, and what I'd test next.

Filled in with the made-up numbers, it might read: "My hypothesis was partly right. More water helped up to Medium, where all 5 beans sprouted, but with Lots only 2 sprouted. Maybe too much water kept the air away from the seeds. Next I'd test 2 tablespoons, to find where it starts to go wrong."

Print the conclusion sheet to fill in yours.

When the experiment itself went wrong

Sometimes it isn't the guess that's wrong, it's the test: nothing happened, the numbers jumped all over the place, or something broke. That still isn't a failed project. Here's what to do for each, and what to write.

• Nothing happened at any level (no seed sprouted, the egg never floated). Check one thing you kept the same: were the seeds too old, the water too cold? If there's time, fix that one thing and test again. If there isn't, write it down: "Nothing changed at any level, so my test couldn't answer the question. I think the reason was ___." That is an honest result, and the judge will ask about it.
• The numbers jumped about, so the three tries at one level don't agree. Do more tries at that level if you can, and look for something that wasn't kept the same (a push instead of a let-go, a different spot on the floor). Show all the tries on the board, not only the good ones, and say the spread made you less sure.
• Something broke or one try was ruined (the ball hit your foot, the cup tipped). Cross that try out in your logbook, say why, and do it again. A crossed-out try with a reason is what a real lab notebook looks like.
• You ran out of time, with fewer tries than you planned. Say how many you did. Two tries at each level is weaker than three, and saying so is better than hiding it. In Next time, write that you'd start earlier and do all three.

In every case, the conclusion has the same three parts as above: what you found, how you know (or why you can't be sure), and what you'd change next time. Put the problem in the conclusion's last part, not in a corner of the board.

What not to do

• Don't change your hypothesis after the test so it looks right. That isn't honest, and judges can usually tell.
• Don't hide the odd numbers. Say what you think happened.
• Don't write "my experiment failed". It didn't. It answered your question, just not the way you guessed.
• Don't say you were right when the numbers only partly agree. Partly right is a fine answer.

Say it out loud

A judge will ask: "Was your hypothesis right? How do you know?" A great answer points at the board: "Partly. I thought more water would always help, but my means went down at Lots, so there's such a thing as too much." Practise it with the judge's questions.

Here's what fairs usually expect from the conclusion at each age:

• 1st–3rd: What happened, in their words.
• 4th–6th: Right, partly or not, and the numbers that show it.
• 7th and up: The numbers, what might have gone wrong, and a next question.