Experimental design appears on the AP Statistics exam in both MCQ and FRQ format. FRQ questions about designing an experiment are among the most predictable question types on the exam -- they appear almost every year and follow a consistent rubric.
The four elements every experiment design must include: comparison (control group vs treatment group), random assignment (not random sampling -- these are different), replication (enough subjects to detect a real effect), and control of confounding variables (blocking, blinding, or holding constant).
Random assignment vs random sampling: This distinction costs students more points than any other concept in AP Stats. Random sampling is how you select participants from a population (affects generalizability). Random assignment is how you assign participants to treatment groups (affects causation). An experiment requires random assignment. A survey requires random sampling. They are not interchangeable.
Blocking: When you know a variable might affect the response, divide subjects into blocks based on that variable first, then randomly assign within each block. Example: testing a new teaching method on students, blocking by prior GPA ensures that high and low performers are evenly distributed across treatment groups.
Blinding: Single-blind means subjects do not know which treatment they receive. Double-blind means neither subjects nor evaluators know. Blinding prevents placebo effects and evaluator bias. The AP exam expects you to identify when blinding is possible and why it matters for the specific experiment described.
Confounding variables: A confounding variable is associated with both the explanatory variable and the response variable, making it impossible to determine which caused the observed effect. AP exam questions ask you to identify potential confounds in a described experiment and explain how they threaten the causal conclusion.
FRQ scoring tip: When asked to design an experiment, always state the following in order: (1) identify subjects, (2) describe random assignment procedure, (3) describe treatments including control, (4) identify the response variable and how it will be measured, and (5) state that differences in the response will be compared between groups. Missing any of these components loses points. Use Cognify to practice writing complete experimental designs under time pressure.