Free for students · Ad-free · WCAG 2.1 AA Compliant · Accessibility
Settings & Accessibility
Experimental Design: Variables and Controls
Scientific Investigation
· Topic 2.1
Introduction
Every ACT Science experiment question is a logic puzzle about cause and effect. Identify what was changed, what was measured, and what was held constant — and you can answer every experimental design question without knowing a single fact about the science.
Experimental design questions appear in every Research Summaries passage — roughly 15–20% of the test. They follow completely predictable patterns once you know the vocabulary.
By the end of this lesson you will be able to:
You will analyze an experiment studying soil type and watering frequency on plant growth — and identify the design flaw that prevents a valid conclusion.
The Concept
The Core Rule
A valid experiment changes exactly one independent variable at a time while holding all others constant. If two things change simultaneously, you cannot know which caused the result.
How the ACT tests this
Identification: 'In Experiment 2, the independent variable was…'
Control purpose: 'Why was Group A included?' — requires explaining the baseline role
Design flaw: 'What change would best allow researchers to determine whether temperature or pH caused the result?'
The Three Variable Types
Independent variable (IV): deliberately changed by the experimenter. Dependent variable (DV): measured as the outcome. Controlled variables (CVs): everything else held constant to prevent confounding.
IV: appears on the x-axis of any graph
DV: appears on the y-axis
CVs: temperature, sample size, time, equipment — all identical across groups
The control group receives no treatment while everything else is identical. It establishes the baseline — what happens when nothing is done.
Control: same conditions as experimental groups except IV = zero or absent
Purpose: provides baseline to measure treatment effect against
ACT question: 'Why was Group X included?' → answer is always 'to serve as a baseline'
Design Flaws
A confounded experiment changes more than one variable at once. You cannot attribute the observed DV change to any single IV.
Confounding: two+ IVs change simultaneously
Missing control: no baseline group
Fix: change only one variable per comparison, add zero-treatment group
Your strategy
1
Step 1 — What did the experimenter deliberately change between groups? That is the IV.
2
Step 2 — What did the experimenter measure at the end? That is the DV.
3
Step 3 — List everything kept the same. Those are CVs. If two things changed at once, flag confounding.
4
Step 4 — Check for a control group: is there a zero-treatment condition?
Worked Examples
Easy
Example 1
Students Confuse Controlled Variables With The IV Because Both Are 'conditions.' The IV Is The One That Was Deliberately Changed, Not Held Constant.
Four groups of 10 bean plants receive 0, 1, 2, or 4 g/L fertilizer in identical pots under identical light. Height measured after 4 weeks.
Table 1: 0 g/L=12 cm, 1 g/L=18 cm, 2 g/L=24 cm, 4 g/L=21 cm.
In this experiment, what is the independent variable?
A.
Mean plant height after 4 weeks
B.
Type of soil used
C.
Fertilizer concentration (Correct answer)
D.
Amount of light provided
Step 1
Step 1 — What changed between groups? Fertilizer concentration: 0, 1, 2, 4 g/L.
Step 2
Step 2 — Plant height was measured (DV). Soil and light were held constant (CVs).
Step 3
Step 3 — The deliberately changed variable is the IV.
Step 4
Step 4 — Answer: Fertilizer concentration. Option C.
Correct answer: C
Why C is correct
Fertilizer concentration was deliberately varied. The IV. Correct.
Why other options are wrong
A: Plant height is what was measured — the DV. Incorrect.
B: Soil was identical across all groups — a CV. Incorrect.
D: Light was held constant — a CV. Incorrect.
⚠ Trap: Students confuse controlled variables with the IV because both are 'conditions.' The IV is the one that was deliberately changed, not held constant.
Medium
Example 2
Students Accept The 40°C Comparison As Valid And Miss That The Overall Design Varies Both Catalyst And Temperature. Thinking 'they Used The Same Temperature Once, So It's Fine.'
Experiment 1: Catalyst A at 20°C and 40°C. Experiment 2: Catalyst B at 40°C and 60°C. Decomposition rate measured.
Catalyst A/20°C=12, A/40°C=28, B/40°C=35, B/60°C=52 mL O₂/min.
A student concludes Catalyst B is more effective based on the 40°C comparison (35 > 28). What is the strongest flaw?
A.
Too small a sample size
B.
Cannot compare catalysts — different temperature ranges tested for each (Correct answer)
C.
Catalyst type is not an IV because it is not numeric
D.
Conclusion is valid — both experiments used hydrogen peroxide
Step 1
Step 1 — What changed between Experiment 1 and 2? Both catalyst type AND temperature range.
Step 2
Step 2 — At 40°C, comparison is possible, but the overall design varies two things simultaneously.
Step 3
Step 3 — Valid comparison requires both catalysts at identical temperature ranges.
Step 4
Step 4 — Answer: B — design is confounded for most comparisons.
Correct answer: B
Why B is correct
Different temperature ranges across experiments confound the catalyst comparison. Correct.
Why other options are wrong
A: Sample size is not the issue here. Incorrect.
C: IVs do not need to be numeric. Incorrect.
D: Same substrate doesn't fix the temperature confound. Incorrect.
⚠ Trap: Students accept the 40°C comparison as valid and miss that the overall design varies both catalyst and temperature. Thinking 'they used the same temperature once, so it's fine.'
Hard
Example 3
Students May Choose A But Misread It As 'test Both Soils At Different Slopes' Rather Than 'at The Same Slope Simultaneously.'
Researcher 1: sandy soil at 10°, 20°, 30°. Researcher 2: clay soil at 20°, 30°, 40°. Rainfall constant at 50 mm/hr. Erosion rate (g/min) measured.
To determine whether soil type alone affects erosion rate, which modification is best?
A.
Test both soil types at 30° only, all other conditions identical (Correct answer)
B.
Test sandy soil at steeper slopes than clay
C.
Increase rainfall to 100 mm/hr for one soil type
D.
Add a third soil type at a new angle
Step 1
Step 1 — To isolate soil type, everything else must be constant — especially slope angle.
Step 2
Step 2 — Current flaw: different slope ranges for each soil. Slope and soil type change simultaneously.
Step 3
Step 3 — Fix: hold slope constant at one value for both soils.
Step 4
Step 4 — Option A: test both at 30° (appears in both current datasets). Answer: A.
Correct answer: A
Why A is correct
Same slope for both soils isolates soil type as the only variable. Correct.
Why other options are wrong
B: Steeper slopes for sandy soil adds more confounding. Incorrect.
C: Changing rainfall introduces a third confounding variable. Incorrect.
D: Adding a third soil type doesn't fix the existing confounding. Incorrect.
⚠ Trap: Students may choose A but misread it as 'test both soils at different slopes' rather than 'at the same slope simultaneously.'
Strategy Tips
Map every experiment: write 'IV=___, DV=___, CV=___' before reading questions
The IV always appears on the x-axis of any graph — confirm your identification
Control group = IV set to zero or absent, not just 'Group A'
To test if a design answers a question: does it change only the relevant variable?
Design-fix answers always: hold the proposed variable constant, vary only the target
Common pitfalls
Confusing IV and DV: IV goes IN (controlled), DV comes OUT (measured)
Accepting a confounded design because 'one comparison seems fine'
Thinking 'control group' means 'comparison group' rather than 'zero-treatment group'
Experimental design questions take 75–90 seconds legitimately. If one takes more than 2 minutes, mark it and return.
Summary
IV = deliberately changed; DV = measured outcome; CVs = everything held constant
Valid experiments change exactly one variable at a time — simultaneous changes create confounding
A control group has no treatment and establishes the baseline
Read any science news article. Identify the IV, DV, and two CVs. Check: did they include a control group? Could two variables have changed simultaneously?