Variables, Controls & Fair Tests
Learn to plan a fair test in Grade 7 science: name the independent, dependent and controlled variables, set up a control group, change one thing at a time, and repeat trials.
What you'll do in this lesson
A voice-first session with the Crimsora tutor on Variables, Controls & Fair Tests, then targeted practice and FRQs — with the tutor adapting to where you get stuck.
What this lesson covers
Imagine two students argue about whether music helps plants grow. One plays music to a plant on a sunny windowsill and puts a silent plant in a dark closet. The music plant grows taller — case closed? Not even close. That experiment changed two things at once, so nobody can tell whether the music or the sunlight did the work.
A fair test fixes this. In a fair test you deliberately change one factor, carefully measure what happens, and hold everything else steady so the result has only one possible cause. This lesson shows you how to name the three kinds of variables, why a control group gives you something to compare against, and why scientists test many individuals instead of one lucky plant. These habits are the backbone of every investigation you will design this year.
A fair test fixes this. In a fair test you deliberately change one factor, carefully measure what happens, and hold everything else steady so the result has only one possible cause. This lesson shows you how to name the three kinds of variables, why a control group gives you something to compare against, and why scientists test many individuals instead of one lucky plant. These habits are the backbone of every investigation you will design this year.
The Three Kinds of Variables
A variable is anything in an investigation that could change or be changed. Every experiment sorts its variables into three jobs.
The independent variable is the one thing you choose to change on purpose. It is the "I change it" variable. In a study of fertilizer and plant height, the independent variable is the amount of fertilizer.
The dependent variable is what you measure to see if the change mattered. It "depends" on the independent variable. Here it is the plant height in centimeters after four weeks. A good dependent variable is measurable with numbers and units, not a vague word like "healthier."
The controlled variables (sometimes called constants) are everything you deliberately keep the same for every group: pot size, soil type, amount of water, hours of light, species and starting height of the seedling, room temperature. Controlled variables are not boring background detail — they are the whole reason your conclusion is trustworthy.
A common mix-up is swapping the independent and dependent variables. Use this test: you can decide the independent variable before the experiment starts, but you cannot know the dependent variable until after you collect data. If you had to wait and measure it, it is dependent.
When you write a hypothesis, both appear: "If I increase the fertilizer, then the plants will grow taller," names the independent variable in the "if" and the dependent variable in the "then."
The independent variable is the one thing you choose to change on purpose. It is the "I change it" variable. In a study of fertilizer and plant height, the independent variable is the amount of fertilizer.
The dependent variable is what you measure to see if the change mattered. It "depends" on the independent variable. Here it is the plant height in centimeters after four weeks. A good dependent variable is measurable with numbers and units, not a vague word like "healthier."
The controlled variables (sometimes called constants) are everything you deliberately keep the same for every group: pot size, soil type, amount of water, hours of light, species and starting height of the seedling, room temperature. Controlled variables are not boring background detail — they are the whole reason your conclusion is trustworthy.
| Variable | Question it answers | Example |
|---|---|---|
| Independent | What did I change? | Grams of fertilizer per pot |
| Dependent | What did I measure? | Plant height in cm |
| Controlled | What did I keep the same? | Water, light, soil, pot, species |
When you write a hypothesis, both appear: "If I increase the fertilizer, then the plants will grow taller," names the independent variable in the "if" and the dependent variable in the "then."
Why a Control Group Gives You Something to Compare
Suppose you give every plant fertilizer and they all grow 12 centimeters. Did the fertilizer help? You cannot say — plants grow on their own. You need a control group: an otherwise identical group that does not receive the treatment. The groups that do receive it are the experimental groups.
The control group answers the question "what would have happened anyway?" Without it, you have a result but no baseline, so there is nothing to compare against.
Be careful with two words that look alike. Controlled variables are the conditions you hold constant. The control group is a set of test subjects that gets the normal or zero-treatment condition. An experiment can have many controlled variables but usually just one control group.
Notice that the control group is not "the group we ignore." You water it, light it and measure it exactly like the others. The only difference is the level of the independent variable.
Sometimes the control is a normal condition rather than nothing at all. Testing a new sports drink? The control group drinks plain water, not nothing, because being thirsty would introduce a second difference. Testing whether a coating keeps apple slices from browning? The control slices get no coating but sit in the same room, on the same plate, cut with the same knife, for the same number of minutes.
The control group answers the question "what would have happened anyway?" Without it, you have a result but no baseline, so there is nothing to compare against.
Be careful with two words that look alike. Controlled variables are the conditions you hold constant. The control group is a set of test subjects that gets the normal or zero-treatment condition. An experiment can have many controlled variables but usually just one control group.
| Setup | Fertilizer | Role |
|---|---|---|
| Group A | 0 grams | Control group |
| Group B | 5 grams | Experimental |
| Group C | 10 grams | Experimental |
Sometimes the control is a normal condition rather than nothing at all. Testing a new sports drink? The control group drinks plain water, not nothing, because being thirsty would introduce a second difference. Testing whether a coating keeps apple slices from browning? The control slices get no coating but sit in the same room, on the same plate, cut with the same knife, for the same number of minutes.
Change One Thing at a Time
The rule that makes a test fair is simple: only the independent variable may differ between groups. If two things differ, the experiment has a confounding variable and the results cannot be interpreted.
Go back to the music-and-plants example. The music plant sat on a windowsill; the silent plant sat in a closet. Two differences: sound and light. When the music plant grew taller, the light is at least as likely an explanation as the music. The experiment produced data but no usable conclusion.
Where students actually go wrong is in the small details they think do not matter. Watching for these will improve almost any plan you write:
Writing your procedure with numbers and units forces the fairness in. "Add a little water" is not repeatable; "add 50 mL with a graduated cylinder" is. A good check before you begin: list every difference between your groups. If that list has more than one item, redesign.
This also explains why you cannot fix an unfair test afterward with clever math. Once sunlight and music changed together, no amount of graphing separates them. Fairness is built in during planning, not rescued during analysis.
Go back to the music-and-plants example. The music plant sat on a windowsill; the silent plant sat in a closet. Two differences: sound and light. When the music plant grew taller, the light is at least as likely an explanation as the music. The experiment produced data but no usable conclusion.
Where students actually go wrong is in the small details they think do not matter. Watching for these will improve almost any plan you write:
| Sloppy choice | Hidden second change | Fix |
|---|---|---|
| "Water them when the soil looks dry" | Different amounts of water | Give each pot 50 mL every morning |
| Different sized pots | Different root space and soil volume | Use identical pots and equal soil |
| Measure one group Monday, one Friday | Different growing time | Measure all groups at the same moment |
| One group by the heater | Different temperature | Same shelf, same room |
This also explains why you cannot fix an unfair test afterward with clever math. Once sunlight and music changed together, no amount of graphing separates them. Fairness is built in during planning, not rescued during analysis.
Repeated Trials Instead of One Individual
One plant is not evidence. Living things and measurements both vary: one seed may be a dud, one thermometer reading may be misread, one sprint time may catch a runner on an off day. Repeated trials — testing many individuals, or repeating the same measurement several times — let you see the pattern instead of the accident.
There are two ways to repeat. You can use more subjects per group (10 plants per fertilizer level instead of 1), and you can repeat a measurement on the same subject (timing the same toy car down the ramp five times). Strong investigations do both where possible.
With repeated trials you can calculate an average, which smooths out random variation, and you can look at the spread. If the control plants range from 9 to 11 centimeters and the fertilized plants range from 10 to 12, the groups overlap so much that the fertilizer effect is doubtful. If control plants cluster near 10 and fertilized plants near 18, the difference is convincing.
A useful phrase for your lab write-ups: "We tested 10 seedlings in each group and repeated the height measurement twice for each plant." That single sentence tells a reader your result is not a fluke.
Repeating also helps you catch mistakes. If four trials give 2.1, 2.0, 2.2 and 7.9 seconds, the last value is an outlier worth investigating — maybe the timer started late. With a single trial you would never have known anything went wrong. In the next lesson, when you graph data and draw conclusions, these averages and spreads are exactly what you will be plotting.
There are two ways to repeat. You can use more subjects per group (10 plants per fertilizer level instead of 1), and you can repeat a measurement on the same subject (timing the same toy car down the ramp five times). Strong investigations do both where possible.
With repeated trials you can calculate an average, which smooths out random variation, and you can look at the spread. If the control plants range from 9 to 11 centimeters and the fertilized plants range from 10 to 12, the groups overlap so much that the fertilizer effect is doubtful. If control plants cluster near 10 and fertilized plants near 18, the difference is convincing.
A useful phrase for your lab write-ups: "We tested 10 seedlings in each group and repeated the height measurement twice for each plant." That single sentence tells a reader your result is not a fluke.
Repeating also helps you catch mistakes. If four trials give 2.1, 2.0, 2.2 and 7.9 seconds, the last value is an outlier worth investigating — maybe the timer started late. With a single trial you would never have known anything went wrong. In the next lesson, when you graph data and draw conclusions, these averages and spreads are exactly what you will be plotting.
Key terms
- Variable.
- Any factor in an investigation that can change or be changed, such as temperature, amount of light, or time.
- Independent variable.
- The single factor the investigator deliberately changes between groups; it is chosen before data collection begins.
- Dependent variable.
- The factor that is measured or observed to see whether it responds to the independent variable; known only after the test runs.
- Controlled variables.
- All the conditions deliberately kept the same for every group so they cannot affect the results.
- Control group.
- A group treated exactly like the others but given no treatment or the normal condition, used as a baseline for comparison.
- Fair test.
- An investigation in which only the independent variable differs between groups, so any difference in results has one likely cause.
- Confounding variable.
- An unintended second difference between groups that makes it impossible to tell what caused the result.
- Repeated trials.
- Testing multiple individuals or repeating a measurement several times so random variation and mistakes show up instead of hiding.
Worked example
Maya wonders whether adding salt to water makes ice melt faster. She has 12 identical ice cubes, four plastic cups, a balance, a timer and table salt. Plan a fair test: name the independent, dependent and controlled variables, describe the control group, and explain how repeated trials fit in.
Step 1 — Name what you will change. Maya wants to know about salt, so the independent variable is the amount of salt sprinkled on the ice: 0 g, 2 g, 4 g and 6 g. That gives four levels, one per cup.
Step 2 — Name what you will measure. The dependent variable is the time in seconds for the ice cube to melt completely. It has a number and a unit, and Maya cannot know it until the test runs. A vague dependent variable like "how melty it looks" would not work.
Step 3 — List what stays the same. Controlled variables: identical cube mass (check each on the balance, about 20 g), same freezer so all cubes start at the same temperature, same cup type, same room temperature and same table, all cubes started at the same moment, salt sprinkled evenly on top, no stirring.
Step 4 — Identify the control group. The 0 g cup is the control group. It shows how long a cube takes to melt with no salt at all, which is the baseline Maya compares the salted cups against. Without it she would only know that salted ice melts, not that salt made a difference.
Step 5 — Build in repetition. One cube per condition is not enough, since cubes vary slightly. Maya has 12 cubes, so she runs three cubes at each of the four salt levels and averages the three melting times. If one time is wildly different from the other two, she investigates before trusting it.
Step 6 — State the comparison. Maya compares the average melting time at each salt level to the 0 g average. If the salted averages are clearly shorter and the pattern grows with more salt, the evidence supports the idea that salt speeds melting.
Step 2 — Name what you will measure. The dependent variable is the time in seconds for the ice cube to melt completely. It has a number and a unit, and Maya cannot know it until the test runs. A vague dependent variable like "how melty it looks" would not work.
Step 3 — List what stays the same. Controlled variables: identical cube mass (check each on the balance, about 20 g), same freezer so all cubes start at the same temperature, same cup type, same room temperature and same table, all cubes started at the same moment, salt sprinkled evenly on top, no stirring.
Step 4 — Identify the control group. The 0 g cup is the control group. It shows how long a cube takes to melt with no salt at all, which is the baseline Maya compares the salted cups against. Without it she would only know that salted ice melts, not that salt made a difference.
Step 5 — Build in repetition. One cube per condition is not enough, since cubes vary slightly. Maya has 12 cubes, so she runs three cubes at each of the four salt levels and averages the three melting times. If one time is wildly different from the other two, she investigates before trusting it.
Step 6 — State the comparison. Maya compares the average melting time at each salt level to the 0 g average. If the salted averages are clearly shorter and the pattern grows with more salt, the evidence supports the idea that salt speeds melting.
Practice questions
A student tests whether a paper airplane flies farther with a paper clip on its nose. She throws the clipped plane outdoors on a breezy day and the unclipped plane indoors in the gym. Which statement best describes the problem with her plan?
- She has no dependent variable, because distance cannot be measured.
- Two things differ between the trials, so she cannot tell whether the clip or the wind caused the difference.
- She should have used the paper clip on both planes to keep the test fair.
- The independent variable and dependent variable have been reversed.
Answer: Two things differ between the trials, so she cannot tell whether the clip or the wind caused the difference.
A fair test allows only the independent variable — the paper clip — to differ. By throwing one plane outdoors in wind and the other indoors, she added a second difference, a confounding variable. Distance is a perfectly good dependent variable, so the first choice is wrong. Putting a clip on both planes would remove the independent variable entirely and leave nothing to compare, so the third choice is wrong too. The fix is to throw both planes in the same gym, from the same spot, with the same throwing motion, several times each.
A gardener claims a new plant food works because the one seedling he fed grew 22 cm in a month. Explain two specific improvements that would make this a fair test, and say what each improvement lets him conclude that he cannot conclude now.
Answer: Add a control group of seedlings that receive no plant food, and test many seedlings in each group instead of one, keeping soil, water, light, pot and species identical.
Right now the gardener has a measurement but no comparison and no evidence that his result is typical. A control group of unfed seedlings grown under identical conditions shows what growth happens anyway; if unfed seedlings also grow 22 cm, the plant food did nothing. Using ten seedlings per group instead of one lets him average results, so a single unusually vigorous seedling cannot fool him, and it reveals whether the fed and unfed ranges actually separate. He also needs the controlled variables — same soil, same 50 mL of water daily, same light, same pot size, same species and starting height — so that food is the only difference between the groups.
In an investigation on how the height of a ramp affects how far a toy car rolls, identify the independent variable, the dependent variable, and two controlled variables.
Answer: Independent variable: the height of the ramp. Dependent variable: the distance the car rolls after leaving the ramp. Controlled variables (any two): the same car, the same ramp surface, the same floor surface, releasing the car from rest at the same starting line, no pushing.
The ramp height is set by the investigator before each trial, which marks it as independent. The rolling distance is measured afterward and responds to that change, which makes it dependent. Everything else must be locked down: swapping cars mid-experiment or pushing the car on some trials would add a second difference and ruin the comparison. Each ramp height should also be tested several times, and the distances averaged, because a single roll can be thrown off by a bumpy release.
FAQ
- What is the difference between a controlled variable and a control group?
- A controlled variable is a condition you keep the same for everyone, like water amount or room temperature. A control group is a set of test subjects that receives no treatment or the normal condition, so you have a baseline to compare against. An experiment usually has many controlled variables and one control group.
- How do I tell the independent variable from the dependent variable?
- Ask when you know its value. You choose the independent variable before the experiment starts — it is written into your procedure. You only learn the dependent variable after you measure. It also fits the hypothesis pattern: if I change the independent variable, then the dependent variable will change.
- How many trials are enough?
- There is no single magic number, but one is never enough. In a classroom investigation, aim for at least three trials at each condition, or three or more individuals in each group, and more when the results are variable. If your repeats give very different values, run more trials and check whether something in your procedure is inconsistent.
- Can an experiment have more than one independent variable?
- Advanced studies do handle several factors at once using special designs, but in a middle school fair test you change only one thing at a time. If you change two, and the result shifts, you cannot say which change caused it. Test one factor, finish it, then design a new investigation for the next factor.
Learn this with a teacher, not a page
The Crimsora tutor teaches Variables, Controls & Fair Tests live — explaining on a whiteboard, asking you questions, and adapting to where you get stuck.