Experimental Design
Learn
The Scientific Method
The scientific method is a systematic approach to investigating questions about the natural world. While often presented as linear steps, real science is iterative - scientists often cycle back through steps as they gather new information.
Steps of the Scientific Method
- Observation: Notice something interesting or puzzling
- Question: Formulate a specific, testable question
- Hypothesis: Propose a tentative explanation
- Experiment: Design and conduct a controlled test
- Analysis: Examine and interpret the data
- Conclusion: Accept, reject, or modify the hypothesis
- Communication: Share findings with others
Types of Variables
| Variable Type | Definition | Example |
|---|---|---|
| Independent Variable | What the researcher changes/manipulates | Amount of fertilizer applied |
| Dependent Variable | What the researcher measures (the outcome) | Plant height |
| Controlled Variables | Factors kept constant to ensure fair test | Light, water, soil type, temperature |
Control Groups
A control group receives no treatment or a standard treatment, serving as a baseline for comparison. The experimental group receives the treatment being tested.
ACT Science Connection
The ACT Science section frequently tests your ability to identify variables, understand experimental design, and predict how changing variables would affect results. You don't need to memorize science facts - you need to understand how experiments work.
Characteristics of Good Experiments
- Controlled: Only one variable changes at a time
- Repeatable: Others can replicate the procedure
- Measurable: Results can be quantified
- Large sample size: More subjects = more reliable results
- Randomized: Subjects randomly assigned to groups to reduce bias
Hypothesis vs. Theory
- Hypothesis: A testable prediction about a specific phenomenon
- Theory: A well-supported explanation for a broad range of observations, backed by extensive evidence
In science, "theory" does not mean "guess" - it represents our best understanding supported by evidence.
Examples
Example 1: Identifying Variables
Experiment: A scientist tests whether different amounts of sleep affect test scores.
- Independent variable: Hours of sleep (4, 6, 8, or 10 hours)
- Dependent variable: Test scores
- Controlled variables: Same test, same time of day, same conditions, similar participants
Example 2: Control Group Design
Research question: Does a new drug reduce blood pressure?
Experimental group: Patients receive the new drug
Control group: Patients receive a placebo (inactive pill)
Why? The control group shows what happens without treatment, accounting for placebo effects and natural variation.
Example 3: Identifying Design Flaws
Claim: A study found that students who drink coffee score higher on exams.
Design flaw: This is an observational study, not an experiment. Coffee drinking wasn't randomly assigned. Perhaps students who drink coffee also study more, sleep less, or have other differences. Correlation does not prove causation.
Better design: Randomly assign students to drink coffee or not before an exam, controlling for other variables.
Practice
Solve these problems. Answers are provided below for self-checking.
1. A researcher wants to test if exercise improves memory. What should be the independent and dependent variables?
2. Why is it important to have a control group in an experiment?
3. A scientist tests fertilizer effects on plants but uses different soil for each group. What is wrong with this design?
4. What is the difference between a hypothesis and a theory?
5. Design an experiment to test whether music affects plant growth. Include variables and control group.
Click to reveal answers
- Independent variable: Amount of exercise (e.g., 0, 30, or 60 minutes). Dependent variable: Memory performance on a standardized test.
- A control group provides a baseline for comparison, showing what happens without the treatment. Without it, you cannot attribute results to the treatment - changes might be due to other factors or natural variation.
- Using different soil is a confounding variable - the design lacks proper control. Any differences in plant growth could be due to soil differences rather than fertilizer. The soil type should be a controlled variable (kept the same for all groups).
- A hypothesis is a specific, testable prediction about a single phenomenon. A theory is a broad, well-supported explanation for many observations, backed by extensive evidence from multiple studies. Theories are not "promoted hypotheses" - they serve different purposes.
- Sample design: Independent variable: Type of music (classical, rock, silence). Dependent variable: Plant height/growth over time. Control group: Plants with no music. Controlled variables: Same plant species, same pot size, same soil, same light, same water, same temperature. Multiple plants per group for reliability.
Check Your Understanding
1. Why can't observational studies prove causation?
Show answer
Observational studies can show correlation (two things occurring together) but cannot prove causation (one thing causing another). Without random assignment and controlled conditions, there may be confounding variables - other factors that explain the relationship. For example, observing that ice cream sales and drowning rates both increase in summer doesn't mean ice cream causes drowning - both are caused by warm weather.
2. Why is sample size important in experiments?
Show answer
Larger sample sizes provide more reliable results because they reduce the impact of random variation and outliers. With small samples, unusual results might be due to chance rather than the treatment. Larger samples give a better picture of the true effect and allow for statistical significance testing.
Next Steps
- Review any concepts that felt challenging
- Move on to the next lesson when ready
- Return to practice problems periodically for review