The Scientific Method: Asking a Question and Testing It Fairly
A 45-minute science lesson that runs one real investigation and names the steps as they happen — built on the question students almost never ask about their own experiments: what else is different?
Science
Subject
Grades 3-5
Grade Level
45 minutes
Duration
Scientific Method
Topic
On this page: Materials · Warm-Up · Direct Instruction · Guided Practice · Independent Practice · Assessment · Closure · Related Resources
Learning Objectives
By the end of this lesson, students will be able to:
- Ask a question that can be answered by testing
- Make a prediction and explain the reason behind it
- Explain what makes a test fair, and identify what is being changed and what is kept the same
- Record observations accurately, including results that were not expected
- Draw a conclusion that the evidence actually supports
Materials
- Scientific Method Worksheet — one copy per student
- Sink or Float Worksheet, for the investigation and for follow-up
- A clear tub or large bowl of water per group
- A collection of objects: a large cork, a small metal screw, an orange, a candle, a crayon, a coin
- Four balls of modeling clay of the same size, one per group, for the second test
- Paper towels, and a tray under each tub
Vocabulary
- Question — something you want to find out that a test could answer
- Prediction — what you think will happen before you test it
- Hypothesis — a prediction with a reason behind it, which a test could show is wrong
- Fair test — a test where only one thing is changed and everything else is kept the same
- Variable — something in a test that could be changed
- Observation — what you actually see happen, whether or not you expected it
- Data — the observations you record
- Conclusion — what the results let you say
Preparation
Choose your objects so that weight and floating come apart. A large cork and a small metal screw are the essential pair — the heavier object floats and the lighter one sinks, which is what breaks the prediction the class is about to make. An orange, a candle and a coin fill it out.
Set out one tub per group with a tray or towels underneath. Water on the floor takes over a lesson faster than anything else in elementary science.
Make the clay balls in advance, all the same size, one per group. They are for the second test and they matter: clay lets you change the SHAPE while keeping the material and the amount identical, which is the cleanest fair test available with classroom materials.
Do not put a scientific-method poster up before the lesson. The steps are going to be named as the class does them, and a poster in view invites students to recite ahead of the thinking. Put it up at the end, next to what they actually did.
Warm-Up
5 minutes. Hold up the large cork and the small metal screw, one in each hand, and pass them round so everyone feels the weight. Ask which will sink.
Nearly every class says the cork — it is much bigger — or says the heavier one sinks, which sounds like a rule. Take a show of hands and write the count on the board. Do not hint.
Drop them both in. The big cork floats and the small screw goes straight down.
Let that sit for a second before saying anything. Then ask the only question that matters: we were confident, and we were wrong — so how could we have found out? Somebody says test it. That is where the lesson starts, and now they want to.
Direct Instruction
12 minutes. Run one investigation with the class from start to finish, naming each step out loud as it happens and writing the name on the board only once the class has done it. The steps are labels for things they are doing, not a sequence to follow.
- Question. Does a heavy object always sink? Narrow it with the class until it is testable — "does this object sink?" is answerable, "why do things float?" is not, not yet.
- Hypothesis. Take a prediction WITH a reason. "It will sink because it is metal" is a hypothesis; "it will sink" is a guess. The reason is what makes it possible to learn something when it turns out wrong.
- Test. Drop objects one at a time and record.
- Observation. Write what happened, including the surprises.
- Conclusion. Say what the results allow. Heavy things do not always sink.
Now teach the part that actually distinguishes an experiment from messing about. Ask the class to design a test for "heavy things sink" and take their first suggestion, which will be a big rock against a paper clip.
Then ask the question this lesson exists for: what else is different? Write the differences up as the class finds them — the weight, the size, the material. Three things changed at once.
Say it plainly: if three things are different and the rock sinks, we cannot tell which of the three did it. A fair test changes one thing and keeps everything else the same.
Then show what a fair test looks like using the clay. Roll a ball, drop it, watch it sink. Flatten the same clay into a boat and float it. Same material, same amount, same water — one difference, so the shape is the only thing left that can explain it.
That comparison is worth going slowly over. It is the first time most students see a result they can actually attribute to a cause, and the certainty is noticeably different from the rock-versus-paper-clip kind.
Guided Practice
13 minutes. Groups predict and then test the full set of objects, recording each prediction BEFORE the test and the result after, in two separate columns.
Writing the prediction first is not a formality. Students who record afterwards reconstruct what they "knew all along," entirely sincerely, and the whole value of the exercise — noticing that you were wrong — disappears.
When the results are in, ask each group to count how many predictions were wrong and say the number out loud. Treat a high count as good news, because it is: it means the test told them something they did not already know.
Then each group designs a fair test of their own for one question, and has to write down what they will change and what they will keep the same:
- Does the shape of the clay change whether it floats?
- Does the temperature of the water change anything?
- Does an orange float differently with the peel on and off?
Circulate with one question and ask it of every group: what else is different? That prompt does more for experimental design than any checklist, and students start asking it of each other within a few minutes.
Independent Practice
10 minutes. Students complete the Scientific Method Worksheet alone, writing up one investigation from the lesson with each step named.
Then the transfer task, which is the one worth marking: give them a flawed experiment on paper — Ella tested whether plants need light by putting a big plant in the window and a small plant in the cupboard, and watering the big one more — and ask them to list everything that is different and say what should have been kept the same.
Spotting an unfair test in somebody else’s design is the skill this lesson is for. Designing your own well comes later and comes from having caught a few bad ones first.
Assessment
3 minutes. Exit ticket, three items. Write a question you could answer by testing. What does it mean for a test to be fair? Then: Jack says his hypothesis was wrong so his experiment failed. Is he right?
The third item is the one that reports whether the culture of the lesson landed as well as the content. A student who writes that the experiment worked fine and he just found out something different from what he expected has understood what testing is for.
Closure
2 minutes. Hold up the cork and the screw one last time and ask what the class would say now to somebody who claims heavy things sink. Close on the sentence the lesson earns: you find out by testing, and a test only tells you something if one thing changed and everything else stayed the same.
Differentiation and Accommodations
- Extra support: reduce it to predict, test, record. Three columns, three objects, and the vocabulary can wait — a student who predicts and then honestly records a result that surprised them is doing science, whether or not they can say "hypothesis." The sink or float worksheet gives the recording structure without the process vocabulary on top.
- Extension: variables in an experiment names what this lesson does by feel — the independent variable, the ones held constant, and what is measured. It is written for older students and suits a group who designed a genuinely fair test without much help.
- Common difficulty: a conclusion the evidence does not support. The group tested four objects and concludes that metal always sinks. Ask what they would need to test to be sure, and let them find that four objects is not many. That is a better lesson than being told the conclusion is too strong.
- Watch for: students quietly changing a recorded result once they see another group’s. It is almost never dishonesty — it is the belief that there is a right answer to be got. Say clearly, more than once, that two groups getting different results is itself a finding, and worth investigating rather than tidying away.
Extension Activities
The natural next step is a proper write-up. Writing a lab report takes the same steps into a written form; it is pitched at older students, so use it as a structure to talk through rather than a sheet to hand out at grade 3.
Every content lesson afterwards is a chance to reuse this. When states of matter asks whether air takes up space, that is a testable question with a fair test attached, and naming it as one costs nothing and reinforces everything.
Keep a class question wall. Any question that starts "I wonder if" goes up, and once a fortnight the class picks one and works out how it could be tested — whether or not there is time to run it. Designing the test is most of the thinking.
This page is print-friendly. Use the Print this lesson plan button above, or your browser's own Print command — either one prints the lesson without the site header, navigation, or resource cards.
Browse all lesson plans or see every Science worksheet on InstructorWeb.