Science fair conclusion examples: right, partly right and wrong

A science fair conclusion answers the question in one sentence, says whether the hypothesis was right, partly right or not right, and points to the numbers that show it. From about 4th grade it adds why that might have happened and one thing to test next. A hypothesis that turned out wrong makes just as good a conclusion as one that came true, as long as it says so plainly and uses the numbers.

I built the conclusion sheet and the report builder on Peabody around those three answers: right, partly right and not right. Here are four examples from real Peabody projects. The numbers are made up for the examples.

The five sentences

1. The answer. Say the question's answer in plain words.
2. Right, partly or not. "My hypothesis was right / partly right / not right."
3. The numbers. The mean at the first level and the last, or the biggest and the smallest.
4. Why (4th grade and up). What you think made it happen, in your own words.
5. Next (4th grade and up). What might have gone wrong, and the next thing you'd test.

Peabody's grade guide, on the grown-ups' page, says the same thing by band: at 1st to 3rd grade, what happened in the child's words; at 4th to 6th, right, partly or not and the numbers that show it; from 7th, the numbers, what might have gone wrong and a next question.

2nd grade: what happened

Project: Which floor makes a ball bounce highest? (1st–5th grade). Floors: carpet, a hard floor, a folded towel. Measured in cm.

The ball bounced highest on the hard floor. I thought the carpet would be best, but I was not right. On the hard floor it bounced 52 cm in the middle try. On the towel it only bounced 9 cm.

Four short sentences, and every one is the child's. That's plenty at 2nd grade. The "middle try" stands in for a mean, which Peabody's tables add from 4th grade.

4th grade: right

Project: What melts ice fastest: salt, sugar or sand? (2nd–6th grade). Levels: nothing, salt, sugar, sand. Measured in minutes.

Salt melted the ice fastest. My hypothesis was right. The ice cube with salt melted in a mean of 14 minutes, and the cube with nothing on it took 41 minutes. Sugar took 29 minutes and sand took 39, almost the same as nothing. I think salt works best because it dissolves into the water on the ice. Next time I would try 1, 2 and 3 teaspoons of salt.

The "nothing" level is what makes this conclusion strong: it shows what the ice does on its own, so the salt's effect is a number, 27 minutes faster, not a feeling.

6th grade: partly right

Project: How much water do bean seeds need to sprout? (2nd–6th grade). Levels: none, a little (1 teaspoon), medium (1 tablespoon), lots (3 tablespoons) a day. Measured in beans sprouted, of 5, in ten days.

Beans need some water to sprout, but more is not always better. My hypothesis was partly right. I thought more water would always mean more sprouts. No beans sprouted with no water, and the mean went up to 4.3 of 5 with a tablespoon a day. But with 3 tablespoons the mean went down to 3.7, and two of those bags smelled bad by day eight. I think the soggy towel kept air away from the beans. Next time I would add a level between 1 and 3 tablespoons to find where it stops helping.

"Partly right" is the most common honest answer at a fair, and the most interesting. The child found a pattern and the place it broke.

7th grade: not right

Project: Does a heavier toy car roll farther? (3rd–7th grade). Levels: 0, 5, 10 and 20 pennies on the car. Measured in cm.

Adding weight did not make the toy car roll farther. My hypothesis was not right. I thought the 20-penny car would go farthest, but the means were 152 cm with 0 pennies, 149 cm with 5, 155 cm with 10 and 150 cm with 20. The differences were smaller than the spread of my three tries at each weight. I think gravity pulls a heavier car harder, but a heavier car is also harder to get moving, so the two cancel out. My tape measure only read to the nearest centimetre, and the car sometimes curved. Next I would run the plain car and the 20-penny car down 1-, 2- and 3-book ramps and draw a line for each.

This is a strong conclusion. It says no, gives the numbers, notices the differences are smaller than the tries' spread, and has a next question. The guide When your hypothesis is wrong has more on saying it well.

Three mistakes judges see

• Restating the hypothesis. "I thought salt would melt it fastest" is the hypothesis again, not a conclusion. The conclusion starts with what happened.
• No numbers. "Salt worked best" needs "14 minutes against 41".
• Hiding a wrong guess. Changing the hypothesis afterwards to match the results is the one thing that costs a project. Peabody's results table template shows how to find the means first.

Try tonight

Print the conclusion sheet for your child's project, or open the report builder and tap Right, Partly right or Not right. It fills in the means from the tries in the Graph Maker and leaves a line for the "but" and the because, which your child writes. Then try the What does it mean? quiz for practice reading results.

Questions grown-ups ask

How long should a conclusion be?

Three or four sentences at 1st to 3rd grade; a short paragraph of five or six from 4th grade up.

Should the conclusion say "I proved"?

No. Say "my numbers show" or "my test found". One test with three tries at each level supports an answer; it doesn't settle it for good.

What if the tries were all over the place?

Say so. "My tries were far apart, so I can't be sure" is an honest conclusion, and doing more tries is the next step.

Peabody (peabody.kids) is free through the end of 2026, with no ads and no sign-up. The library has free stories by reading level.