Fair Tests in Primary Science: What to Change, What to Keep the Same

17 October 2026 · by Larry

The question shows three flower pots with bean seeds. Pot A is on a sunny windowsill and gets 50 mL of water a day. Pot B is in a dark cupboard and gets 50 mL a day. Pot C is in the same cupboard and gets 100 mL a day. Which two pots should Mei Ling compare to find out whether sunlight affects how tall the plants grow?

Your child picks A and C. The reasoning sounds sensible: the pots are different, so surely that is the test. But A and C differ in two ways, light and water, so if A grows taller there is no way to say which one did it. The pots to compare are A and B. Only the light is different.

That is the whole idea of a fair test, and a child can understand all of it and still miss it, because the question often never uses the word “fair”. It just asks which pots to pick, or what to change, or what to keep the same.

Three jobs in every experiment

Many experiment questions in primary Science are asking about one of three jobs. It helps to say them to your child in plain words.

• What do I change? Exactly one thing. In the pots, it is the light.
• What do I measure? The result I will look at, and it should be something with a number or a clear count. In the pots, it is the height of the plant after a set number of days.
• What do I keep the same? Everything else that could change the result. Here: the size of the pot, the type of seed, the amount of soil, the amount of water, the temperature, and the day the height is measured.

A test is fair when the answer to the first job is one thing, and the list for the third job is long.

1. Changing two things at once

This is the A-and-C mistake above. It usually comes from a child looking for the pots that look most different, as if a bigger difference gives a clearer answer. The opposite is true. The clearest test is the one where the pots are as alike as possible, apart from the one thing being tested.

A quick way to check any pair: say out loud what is different between the two, and count. One difference is a fair test. Two or more is not. In the example, A and B differ in light only. B and C differ in water only, so that pair is a fair test of water, not of light. A and C differ in both, and cannot test either.

2. Mixing up “change” and “measure”

Ask what Mei Ling changed and some children say “the height of the plant”. It did change, over two weeks, so it feels like the answer. But the height is what she looked at to find out. What she did differently from one pot to the next was the light.

The sentence that sorts it out: I change one thing on purpose, and I measure what happens because of it. The thing I change is the cause I am testing. The thing I measure is the effect I am watching for. If your child can say the sentence about any experiment in a question, the two usually end up in the right boxes.

3. “Keep everything else the same” names nothing

It is the answer many children give when asked what to keep the same, and it is true. But “everything” points at no particular thing. A question that asks what to keep the same is usually looking for something the child can point to: the amount of water, the size of the pot, the type of seed. The exact detail, such as “50 mL of water each day for every pot”, is stronger than a general statement.

A good way in is to ask, what else could make a plant grow taller or shorter? Each answer to that becomes something to keep the same.

4. A measurement nobody could check

“Which plant looks healthier?” is hard to compare, because two people can disagree. “How tall is each plant after 14 days, in centimetres?” can be checked by anyone with a ruler. When a question asks how to measure the result, an answer that includes a unit, a time or a count is firmer than a word like “better”, “healthier” or “nicer”.

5. A conclusion that goes further than the evidence

Suppose the plants measured 18 cm in pot A, 6 cm in pot B and 9 cm in pot C. These are made-up numbers, only to show the idea. From A and B, a correct conclusion is: the plant in the sunny window grew taller than the plant in the cupboard, so light affects how tall these bean plants grow.

A common slip is a bigger claim: “all plants need sunlight to grow.” The test did not look at all plants, only these seeds in these pots. Another slip is using the wrong pots for the claim: “C grew taller than B, so light makes plants grow”. B and C differ in water, so that pair says something about water.

Reading the result out of a table in the first place is a skill of its own; reading graphs and tables: where children misread them covers it.

6. Not reading the question

Some of the lost marks are not about science at all. A child answers what to change when the question asked what to measure, or lists three things to keep the same when the question asked for one. This is the same pattern that sits behind many “careless” mistakes: the child knows an answer, but to a different question from the one on the page. Underlining the question word (which, what, why, how many) before writing is a small habit with a large effect, and it is the same habit that helps in word problems.

A fair test you can do in the kitchen

Ask your child: does ice melt faster in the sun? Take two identical cups and put the same number of ice cubes in each. Put one on a sunny windowsill and the other in a cupboard. After 30 minutes, tip the water from each cup into a measuring jug and read the amounts.

Then ask the three questions. What did we change? (Where the cup was.) What did we measure? (How much water.) What did we keep the same? (The cup, the ice, the number of cubes, the time.) Now make it unfair on purpose: use a big cup in the sun and a small cup in the cupboard, and ask whether the result still means anything. A child who sees why it does not has understood the idea.

What helps

• Count the differences between any two setups. One is fair. More than one is not.
• Say the sentence: I change one thing on purpose, and I measure what happens because of it.
• Name each thing kept the same, with its exact detail, not “everything”.
• Prefer a result with a number or a unit over a word like “better”.
• Keep the conclusion to the pots that were compared, and to what they showed.

What does not help

• Memorising a model answer for the bean plant. The next question will be about a ball on a ramp or sugar in water, and the words will not match.
• Telling your child to “think harder about the experiment”. The three jobs give a way into the question that thinking harder does not.
• Rushing to the answer. Reading the setup twice, once to see what is different and once to see what is measured, is where most of the time should go.

A ten-minute check you can do at home

Read each setup aloud and ask your child the three questions.

• Two identical toy cars roll down the same ramp. At the bottom, one rolls onto a carpet and the other onto a tile floor. The distance each rolls is measured. What is changed? What is measured? Name two things to keep the same.
• Sugar is stirred into two cups of water. Cup 1 has one spoon of sugar and cup 2 has three spoons, and cup 2 is also warmer. Is this a fair test of how the amount of sugar affects how quickly it dissolves? Why not?
• Which is the better thing to measure: “which paper towel is strongest”, or “how many marbles a towel holds before it tears”?

The answers: the floor is changed, the distance rolled is measured, and the same car, the same ramp and the same release point are kept the same. The sugar test is not fair, because the temperature is different as well as the sugar, so cup 2 has two differences. The marble count is the better measure, because it is a number anyone can check.

Where this fits

StudyLab does not have Science practice yet; the home page lists it as coming later, so this piece is reading for parents and not a game. The skill carries over, though: stating what was changed, measured and kept the same is a way of reading any question carefully. Which Science topics are taught at which level, and how questions are worded and marked, is set by the school and MOE, so use your child’s own worksheets for the real wording. The free maths practice is there for the numbers side.