Scientific Reasoning and Data Analysis
What this domain tests
Scientific Reasoning and Data Analysis tests experimental design, variable identification, statistics, and interpretation of tables and graphs. You must distinguish observation from inference and statistical significance from biological importance.
Core rules and concepts
- Independent variable is manipulated; dependent variable is measured; controlled variables are held constant.
- A control group provides a baseline; randomization reduces bias; replication estimates variability.
- Correlation does not establish causation; confounding variables can create spurious associations.
- Mean, median, and standard deviation summarize data; standard error reflects precision of an estimate.
- A p-value is the probability of data at least as extreme as observed if the null hypothesis is true; it is not the probability the null is true.
- Effect size and confidence intervals matter more than a significant label alone.
Worked example
A fertilizer trial assigns plots to fertilizer or no fertilizer and measures plant height. Independent variable = fertilizer treatment; dependent variable = height. A control = no-fertilizer plots. If fertilized plants are taller on average but plots differ in sunlight, sunlight is a confounder. Reasoning: random assignment of treatments to plots would reduce confounding; replication across plots would improve confidence.
Traps that cost marks
Do not call a p-value the probability that the hypothesis is correct. Do not confuse statistical significance with practical importance. Do not infer causation from an observational correlation. Do not ignore sample size when judging a difference. Do not treat a controlled variable as a dependent variable.