Critical Thinking and Scientific Problem-Solving Skills
Critical thinking and scientific problem-solving
This domain tests reasoning with passages and data rather than memorized facts. CARS asks you to identify arguments, assumptions, tone, and inferences; research-design questions ask you to evaluate variables, controls, validity, and causal claims.
Core rules and concepts
- Arguments: separate conclusion from premises. The conclusion is what the author wants you to accept; premises support it. An assumption is an unstated premise needed for the argument.
- CARS: answer from the passage, not outside knowledge. Prefer the most supported, moderate choice. Distinguish restatement from inference and the author's claim from a critic's view.
- Experiments: the independent variable is manipulated; the dependent variable is measured. Control groups, randomization, and blinding reduce bias. Internal validity concerns causal inference; external validity concerns generalization.
- Data: correlation is not causation. Confounders affect both exposure and outcome. Selection, measurement, and attrition bias distort results. Statistical significance is not clinical importance.
Worked example
A cohort study reports that coffee drinkers have lower rates of heart disease. The authors conclude that coffee protects the heart. The exposure is coffee drinking; the outcome is heart disease. Because participants chose their exposure, confounding is likely: coffee drinkers may differ in smoking, income, or activity. The causal conclusion is unsupported. A randomized trial assigning coffee intake would better control confounding, though adherence and blinding would be difficult.
Traps that cost marks
Do not import outside facts into CARS. Extreme wording ("all," "never") is often wrong. In experiments, a significant p-value does not prove a large or causal effect. Do not confuse reliability (consistency) with validity (accuracy). Check whether a study is observational or experimental before accepting causal language.