Why Experiments Need Controls

A scientific experiment is designed to determine whether changing one factor causes a measurable change in another factor.

However, many things can affect the outcome of an experiment. Scientists therefore use controls to make comparisons more reliable.

A control provides a reference or comparison that helps scientists determine whether the independent variable actually caused the observed result.

A useful idea is:

experimental result + appropriate control → stronger evidence about cause and effect

---

What Is a Control?

A control is a standard or comparison used in an experiment.

Depending on the investigation, the control may be:

• A control group.
• A control condition.
• An untreated sample.
• A sample receiving a known treatment.
• A normal or baseline condition.

The control should be as similar as possible to the experimental treatment except for the factor being investigated.

---

Experimental Group and Control Group

Many experiments contain at least two groups.

Experimental Group

The group that receives the treatment or condition being investigated.

Control Group

The comparison group that does not receive the experimental treatment, or receives a standard treatment.

The results of the two groups can then be compared.

---

Simple Example: Fertilizer

Suppose a student asks:

Does fertilizer increase the growth of bean plants?

The student grows two groups of identical bean plants.

Experimental Group

Plants receive fertilizer.

Control Group

Plants do not receive fertilizer.

Both groups receive:

• The same amount of water.
• The same amount of light.
• The same type of soil.
• The same temperature.
• The same growing time.

If the fertilized plants grow more, the control group provides a baseline for comparison.

Without the control group, it would be much harder to determine whether the fertilizer caused the difference.

---

Control Groups Are Not the Same as Controlled Variables

This distinction is extremely important.

Control Group

A group or condition used for comparison.

Controlled Variable

A factor kept the same throughout the experiment.

These terms sound similar, but they describe different things.

---

Example: Fertilizer Investigation

Question:

How does fertilizer concentration affect plant growth?

Independent Variable

Fertilizer concentration.

Dependent Variable

Plant growth.

Controlled Variables

• Plant species.
• Starting plant size.
• Amount of water.
• Soil type.
• Light exposure.
• Pot size.
• Temperature.
• Length of experiment.

Control Group

Plants receiving no fertilizer.

The controlled variables help make the experiment fair.

The control group provides a baseline for comparison.

---

Why Controlled Variables Matter

Suppose one group receives fertilizer and another does not.

However, the fertilized plants also receive twice as much water.

If the fertilized plants grow faster, what caused the difference?

Was it:

fertilizer?

or:

extra water?

We cannot tell.

Water has become a confounding variable.

Keeping water constant helps isolate the effect of fertilizer.

---

Control Conditions

Not every investigation has separate groups of organisms.

Sometimes scientists use a control condition.

Suppose students investigate:

How does temperature affect the rate at which a tablet dissolves?

They might test:

• 10°C.
• 20°C.
• 30°C.
• 40°C.
• 50°C.

A standard temperature, such as 20°C, might serve as a reference condition.

The other results can be compared with this baseline.

---

Baseline Conditions

A baseline represents a normal or starting condition.

For example, scientists investigating the effect of exercise on heart rate might measure a person's:

resting heart rate

before exercise.

They then compare the heart rate:

• Before exercise.
• Immediately after exercise.
• After several minutes of recovery.

The resting measurement acts as a baseline.

---

Negative Controls

A negative control is designed to produce no experimental effect if everything is working correctly.

Suppose scientists test whether a disinfectant kills bacteria.

Experimental Treatment

Bacteria + disinfectant.

Negative Control

Bacteria + no disinfectant.

The negative control shows how the bacteria behave without the treatment.

If bacteria decrease even in the negative control, something else may be affecting the experiment.

---

Positive Controls

A positive control is a condition expected to produce a known result.

Suppose a scientist is testing a new antibacterial substance.

The experiment might contain:

Experimental Group

Bacteria + new substance.

Negative Control

Bacteria + no antibacterial treatment.

Positive Control

Bacteria + an established antibacterial treatment known to work under the test conditions.

The positive control can help show whether the experimental method is capable of detecting an antibacterial effect.

---

Why Positive Controls Are Useful

Imagine the new substance produces no effect.

There are at least two possibilities:

Possibility 1: The new substance does not work.

Possibility 2: Something was wrong with the experimental method.

If the positive control also fails to produce its expected effect, this suggests there may be a problem with the procedure.

Controls therefore help scientists interpret results.

---

Placebos as Controls

In some clinical research, a placebo may be used as part of a control condition.

A placebo resembles the treatment but does not contain the active treatment being tested.

Researchers can compare:

treatment group

with:

placebo control group

This helps distinguish effects associated with the active treatment from effects related to expectations, study participation, or other factors.

Human research requires appropriate ethical safeguards, so not every study can or should use a placebo.

---

Worked Example: Light and Plant Growth

Research question:

How does light intensity affect plant growth?

A student places plants at different distances from a lamp.

Independent Variable

Light intensity.

Dependent Variable

Plant growth.

Controlled Variables

• Plant species.
• Starting size.
• Soil.
• Water.
• Pot size.
• Temperature.
• Length of experiment.

A suitable baseline could be plants grown under the normal lighting condition used for comparison.

The student should change the light condition while keeping other important factors as consistent as possible.

---

Worked Example: Exercise and Heart Rate

Question:

How does exercise affect heart rate?

Independent Variable

Exercise condition or intensity.

Dependent Variable

Heart rate.

Baseline Control Condition

Resting heart rate before exercise.

Controlled Variables Could Include

• Exercise duration.
• Measurement method.
• Recovery time before each trial.
• Environmental conditions where practical.

Comparing exercise measurements with the resting baseline helps show the change associated with exercise.

---

Worked Example: Enzyme Activity

Question:

How does pH affect enzyme activity?

A student tests enzyme activity at:

pH 3, 5, 7, 9 and 11

The student must also control other factors such as:

• Temperature.
• Enzyme concentration.
• Substrate concentration.
• Reaction time.
• Total reaction volume.

If pH 7 represents the normal reference condition for the investigation, it can serve as a useful comparison.

The controlled variables are not control groups.

---

Worked Example: Cleaning Products

Suppose students investigate:

Which cleaning product removes an oil stain most effectively?

They test:

• Product A.
• Product B.
• Product C.
• Water only.

The water-only treatment provides a useful control.

Why?

It shows how much stain removal occurs from the washing procedure without the cleaning products being tested.

---

Worked Example: Seed Germination

Question:

Does a particular treatment affect seed germination?

Students use 40 similar seeds.

Group A

20 seeds receive the treatment.

Group B

20 seeds do not receive the treatment.

Group B is the:

control group

Other factors should remain similar:

• Seed type.
• Water.
• Temperature.
• Light.
• Growing medium.
• Observation time.

If 18 treated seeds germinate but only 10 control seeds germinate, the difference provides evidence that the treatment may have influenced germination.

Replication and further testing would strengthen the conclusion.

---

Identifying a Missing Control

Consider this experiment:

A student gives fertilizer to five plants.

After three weeks, all five plants have grown.

The student concludes:

"The fertilizer caused the plants to grow."

There is a major problem.

Plants normally grow even without fertilizer.

The experiment needs:

similar plants grown without the fertilizer

Without this control group, the student cannot determine how much growth would have occurred anyway.

---

Identifying a Poor Control

Suppose students investigate whether fertilizer increases tomato plant growth.

Experimental Group

Tomato plants + fertilizer, grown outdoors.

Control Group

Tomato plants without fertilizer, grown indoors.

This is a poor design.

Why?

Two factors have changed:

• Fertilizer.
• Growing environment.

If the groups grow differently, the students cannot determine which factor caused the difference.

A better control group would experience the same conditions except for the fertilizer treatment.

---

Effective Controls

A good control should usually:

• Provide a meaningful comparison.
• Be treated similarly to the experimental group.
• Differ mainly in the independent variable or treatment being tested.
• Be measured using the same methods.
• Be observed over the same time period.

This allows the effect of the independent variable to be isolated more effectively.

---

Designing an Experiment with a Control

Suppose you want to investigate:

Does salt concentration affect seed germination?

Step 1: Identify the Independent Variable

Salt concentration.

Possible concentrations:

0%, 0.5%, 1.0%, 1.5%, 2.0%

Step 2: Identify the Dependent Variable

Percentage of seeds that germinate.

Step 3: Establish a Control

Seeds receiving:

0% salt solution

This provides the normal condition without added salt.

Step 4: Identify Controlled Variables

Keep constant:

• Seed species.
• Number of seeds.
• Volume of solution.
• Temperature.
• Light conditions.
• Container.
• Growing medium.
• Time allowed for germination.

Step 5: Compare Results

Compare each salt treatment with the 0% salt control.

---

Controls and Fair Testing

Controls are part of a broader idea called a fair test.

In a fair test:

one independent variable is deliberately changed

while:

other relevant variables are kept as constant as reasonably possible

and:

the dependent variable is measured consistently

An appropriate control provides a comparison that helps interpret the results.

---

Controls Do Not Eliminate All Uncertainty

Even a well-controlled experiment can contain variation.

For example, individual plants may naturally grow at different rates.

Scientists therefore often use:

• Multiple samples.
• Repeated trials.
• Larger sample sizes.
• Random assignment where appropriate.
• Consistent measurement methods.

Controls improve experimental design, but they work together with these other techniques.

---

Control vs Repetition

A control and a repeated trial serve different purposes.

Control

Provides a comparison.

Repetition

Helps determine whether results are consistent and reduces the influence of random variation.

A strong investigation often uses both.

---

Control vs Constant

Consider:

Does water temperature affect how quickly sugar dissolves?

Independent Variable

Water temperature.

Dependent Variable

Time required for sugar to dissolve.

Controlled Variables

• Amount of sugar.
• Volume of water.
• Type of sugar.
• Stirring method.
• Container.

Reference Condition

A standard temperature may be selected for comparison.

Notice again:

controlled variables ≠ control condition

---

Does Every Experiment Need a Control Group?

Not every scientific investigation requires a separate control group.

For example, an investigation measuring how pendulum period changes with length may simply test many different lengths while keeping other variables constant.

A separate "untreated pendulum" would not necessarily be meaningful.

Instead, scientists compare the measurements across the different values of the independent variable.

Therefore, an important question is not simply:

"Where is the control group?"

but:

"What comparison is needed to answer the research question?"

---

Observational Studies

Some scientific investigations are observational rather than experimental.

Scientists might study:

• Animal populations.
• Climate patterns.
• Ecosystems.
• Geological formations.
• Astronomical objects.

Researchers may not directly manipulate an independent variable.

Instead, they compare naturally occurring conditions or groups.

The idea of comparison remains important even when a traditional experimental control is not possible.

---

Evaluating an Experimental Control

When examining an experiment, ask:

What is the independent variable?

What is the dependent variable?

What is being used as the comparison?

Does the control differ from the experimental group mainly in the treatment being investigated?

Are other important variables controlled?

Is the control measured in the same way?

Does the control actually help answer the research question?

These questions help determine whether the control is appropriate.

---

Example: Evaluate the Experiment

A student wants to determine whether caffeine affects the growth of bean plants.

Group A

10 bean plants receive water containing caffeine.

Group B

10 bean plants receive plain water.

Both groups have:

• Identical soil.
• Similar starting sizes.
• Equal water volumes.
• Equal light.
• Equal temperatures.
• Identical pots.

Plant height is measured every two days.

Evaluation

Group B is an appropriate control group because it receives the same basic treatment except for caffeine.

The design therefore allows the effect of caffeine to be investigated more effectively.

---

Example: Find the Problem

A student investigates whether detergent affects seed germination.

Group A

Seeds receive detergent solution and are kept at 25°C.

Group B

Seeds receive water and are kept at 10°C.

The control is poorly designed.

Two variables differ:

detergent exposure

and:

temperature

Therefore, differences in germination could be caused by either variable.

A better design would keep both groups at the same temperature.

---

Designing Stronger Controls

Suppose you want to test a new plant-growth treatment.

A stronger design might include:

Negative Control

Plants receive no active treatment.

Experimental Group

Plants receive the new treatment.

Positive Control

Plants receive an established treatment known to influence plant growth, if such a comparison is relevant.

This design can provide more information than simply observing the experimental group alone.

---

Common Mistakes

Confusing a Control Group with a Controlled Variable

A control group is a comparison group.

A controlled variable is a factor kept constant.

Saying "The Control Is Everything Kept the Same"

Those are controlled variables, not the control group.

Changing Several Variables at Once

This makes it difficult to determine which variable caused the result.

Using an Unfair Control Group

The control should experience comparable conditions except for the treatment being investigated.

Assuming the Control Must Receive Nothing

A control may receive water, a placebo, a standard treatment, or another appropriate baseline condition.

Assuming Every Experiment Needs a Separate Control Group

Some experiments compare different values of an independent variable without a separate untreated group.

Forgetting Repeated Trials

A control provides comparison, but repeated trials help assess consistency.

Assuming a Control Proves Cause and Effect by Itself

Controls strengthen evidence, but experimental design, sample size, measurement quality, replication, and other factors also matter.

---

Check Your Understanding

1. What is the purpose of a control in an experiment?

2. What is a control group?

3. What is a controlled variable?

4. Explain the difference between a control group and a controlled variable.

5. Why does a fertilizer experiment need plants that receive no fertilizer?

6. Why should control and experimental groups be treated similarly?

7. What is a baseline?

8. What is a negative control?

9. What is a positive control?

10. Why might a positive control be useful?

11. What is the purpose of a placebo in some experiments?

12. A student tests whether fertilizer affects plant growth. Identify an appropriate control group.

13. Give four variables that should be controlled in the fertilizer experiment.

14. A student tests detergent on seeds but changes both detergent concentration and temperature. What is wrong with the design?

15. Why is plain water a useful control when testing cleaning products?

16. How does a control help scientists determine cause and effect?

17. What is the difference between a control and a repeated trial?

18. Why does natural variation make repeated trials useful?

19. Does every experiment require a separate control group? Explain.

20. Design an appropriate control for an experiment investigating how salt affects seed germination.

---

Key Terms

• Control – reference or comparison used to help interpret experimental results.
• Control group – group used as a comparison with the experimental group.
• Control condition – reference condition used for comparison.
• Controlled variable – factor kept constant during an experiment.
• Independent variable – factor deliberately changed by the investigator.
• Dependent variable – factor measured in response to the independent variable.
• Experimental group – group receiving the treatment or condition being investigated.
• Baseline – normal or starting condition used for comparison.
• Negative control – control expected to show no experimental effect.
• Positive control – control expected to produce a known response.
• Placebo – inactive treatment used as a comparison in some studies.
• Confounding variable – uncontrolled factor that could influence the dependent variable.
• Fair test – investigation in which the independent variable is deliberately changed while other relevant factors are controlled.
• Replication – repeating an investigation or using multiple experimental units to test the reliability of results.

---

Key Takeaways

• Controls provide a reference against which experimental results can be compared.
• A control group is not the same thing as a controlled variable.
• Controlled variables are factors kept constant.
• Control groups provide comparison conditions.
• Experimental and control groups should usually differ mainly in the treatment being investigated.
• Negative controls show what happens without the experimental treatment.
• Positive controls provide a known expected response.
• Baseline measurements can act as useful comparison conditions.
• Appropriate controls make it easier to determine whether the independent variable caused an observed change.
• Poor controls can introduce confounding variables.
• Controls should be measured using the same methods as experimental treatments.
• A control does not replace repeated trials or adequate sample sizes.
• Not every investigation requires a separate control group.
• The most important question is:

What comparison is needed to answer the research question fairly?