Controlled Variable: Meaning, Examples & Uses

Controlled Variable Something that you keep the same in an experiment so it can't change your results. Suppose you want to test how light affects plant growth. You give every plant the same water, soil, and pot size, and those are the things you control.

Keeping them the same makes your test fair. If your plants grow differently, you can be fairly sure light caused it, not something else. For a broader comparison of observational and experimental study designs, see our guide.

What does it mean in simple words?

Think of a race. Every runner gets the same track and distance, so the only thing being compared is speed.

In science, those fixed things are called constants or control variables. They protect what scientists call internal validity, which just means confidence that your change, and nothing else, caused the result.

The three kinds of variables in an experiment

An experiment is one of the main types of research methods, and every experiment has three kinds of variables. You change one, you measure one, and you keep the rest the same.

Type	What it is	Plant growth example
Independent variable	The one thing you change on purpose	Hours of light each day
Dependent variable	What you measure to see the effect	Plant height after 3 weeks
Constants	Everything you keep the same	Water, soil, pot size, temperature, type of plant

Examples from different subjects

Here are five simple experiments and what each one keeps fixed.

Experiment	You change	You measure	You keep the same
Paper towel test	Brand of towel	Water soaked up	Towel size, water temperature, soaking time
Pendulum swing	Length of string	Time for 10 swings	Weight on the string, starting angle, timer
Enzyme test	Temperature	Reaction speed	pH, amounts of each chemical, volume
Sleep and memory	Hours of sleep	Words remembered	Word list, study time, noise level
Boyle's law	Volume of gas	Pressure	Temperature, amount of gas

The last one is a classic physics experiment. Keeping temperature steady lets you see how squeezing a gas raises its pressure.

How to control variables in your experiment

List everything that could affect your result, like time, temperature, tools, and materials.
Pick the ones most likely to matter. You can't control everything.
Decide exactly how you'll keep each one steady. "Same water" isn't enough. "50 mL of tap water every morning" is.
Use the same tools and steps for every trial.
If you can't control something, split people or items into groups at random. This spreads hidden differences evenly.
Write down anything you couldn't control, and mention it as a limitation.

How to write it in a lab report

Teachers usually want three things for each controlled variable: what it is, why it matters, and how you kept it the same. A table makes this easy.

What you kept the same	Why it matters	How you kept it the same
Amount of water	More water could make plants grow taller	50 mL of tap water per pot, every morning at 8 a.m.
Soil type	Different soils have different nutrients	Same bag of potting soil for every pot
Temperature	Warmth speeds up growth	All pots in one room at about 21 °C
Type of plant	Different plants grow at different speeds	All seeds from one packet

Each item should be something you can measure. If you can't say how you kept it steady, your teacher can't tell if you did. Our guide to operational definition examples shows how to turn fuzzy ideas into things you can measure.

Controlled Variable vs Control Group vs Confounding Variable

These three sound alike, but they're different.

Term	What it is	Plant example
Controlled variable	A factor kept the same for every group	Every plant gets 50 mL of water
Control group	A group that doesn't get the treatment, used for comparison	Plants grown with normal daylight only
Confounding variable	Something you didn't control that changes along with what you're testing	The plants under extra light also sat near a heater, so they were warmer too

In short, a controlled variable is a factor. A control group is a group of people, plants, or samples. A confounding variable is the problem that constants help you avoid.

Controlling for a variable in statistics

Sometimes you can't keep things the same. In an observational study, you only watch and measure what's already happening. So researchers "control for" extra factors in their math instead.

For example, say coffee drinkers seem to have more heart disease. But coffee drinkers might also smoke more. Adding smoking to the analysis helps separate the effect of coffee from the effect of smoking. If you're not sure which kind of study you're reading, see observational study vs experimental study.

This kind of control has limits. A 2016 study in PLOS ONE showed that when an extra factor is measured poorly, part of its effect still leaks into the results. The problem got worse with bigger samples.

A 2024 paper by Cinelli, Forney, and Pearl also warned that some factors make results worse when you control for them. One example is a factor that sits between the cause and the effect. So a well run experiment is usually stronger evidence than statistical control alone.

What it means in engineering

In engineering, the controlled variable is the thing a machine tries to hold at a target, called the set point. That's how the US National Institute of Standards and Technology (NIST) defines it.

Think of a home heater. Room temperature is what's being controlled, 21 °C is the set point, and the thermostat turns the heater on and off to stay there. In science class, you keep things steady yourself. In engineering, a machine does it for you.

Common mistakes to avoid

Mistake	Better way
Listing what you kept the same but not how	Add an exact amount, time, or tool for each one
Calling the control group a constant	A constant is a factor. A control group is a set of people or samples
Changing two things at once	Change only your independent variable
Keeping things so tightly fixed that results don't match real life	Mention it as a limitation
Believing "controlled for" in a paper means fully controlled	Check how the factor was measured

Conclusion

Keeping things the same is what makes an experiment fair. Change one thing, measure one thing, and hold everything else steady. Then write down exactly how you did it. Before you plan your test, our guide on what is a research question can help you shape a clear question to test.

FAQs

Can a controlled variable change during an experiment?

It shouldn't. If it changes by accident, like the room warming up, write it down and mention it as a limitation.

How many things should I keep the same?

There's no set number. Control anything that could really affect your result. Most school experiments list three to five.

Does this idea work in psychology too?

Yes. Researchers keep things like instructions, test materials, and room conditions the same for every person, so only the factor being studied is different.

Can Canyam help me see how real studies control variables?

Yes. Canyam lets you search more than 65 million research papers and read short AI summaries of their key findings and methods, so you can quickly see how other researchers set up their experiments.

Does it go on my graph?

Usually not. Your graph shows the independent variable on the bottom and the dependent variable up the side. A controlled variable is normally listed in your method or under the graph instead.