Nature of Science
The Nature of Science domain tests how scientific claims are generated, tested, and revised. Expect questions on experimental design, variables, measurement, data interpretation, and the distinctions among hypothesis, theory, and law.
Core rules and concepts
- A hypothesis is a testable, falsifiable prediction. A scientific theory is a broad explanation supported by many lines of evidence; a law describes a consistent relationship (for example, Newton's second law) and does not "graduate" into a theory.
- In a controlled experiment, the independent variable is manipulated, the dependent variable is measured, and controlled variables are held constant. A control group provides a baseline.
- Reliability means repeated measurements give consistent results; validity means the procedure measures what it claims to measure.
- Precision describes closeness of repeated measurements to each other; accuracy describes closeness to the true value. Random error scatters results; systematic error shifts them in one direction.
- Scientific knowledge is peer-reviewed, replicable, and open to revision. Models are useful approximations, not perfect copies of reality.
Worked example
A student predicts that longer light exposure increases duckweed growth. She sets up 10 identical containers with equal water, nutrients, and duckweed. Five receive 6 hours of light daily; five receive 12 hours. Temperature and container size are identical. After 14 days she counts fronds. The independent variable is light duration; the dependent variable is frond count; controlled variables include temperature, nutrients, water volume, and container size. The 6-hour group is a comparison baseline, not a true no-treatment control, because duckweed needs light. She should compare mean frond counts and variability, not just one container.
Traps that cost marks
- Calling a well-supported explanation a "hypothesis" or saying a law is "proven."
- Ignoring a confounding variable, such as temperature differing between light groups.
- Treating correlation as causation.
- Confusing accuracy with precision, or reporting more significant figures than the instrument supports.