By Fatskills Exam Guides Team — the exam nerds behind 28,500+ quizzes and 2.1M practice questions across 500+ global exams.
This topic distinguishes between observational studies (where researchers merely observe and record data without interfering) and experiments (where researchers actively impose treatments to measure effects). Understanding the difference is crucial because only well-designed experiments can establish causation, while observational studies can only suggest associations. This is a frequent FRQ topic on the AP exam, often tied to real-world scenarios like testing whether a new teaching method improves test scores, whether a drug reduces symptoms, or whether a policy change affects employment rates.
Example: Surveying students about sleep habits and GPA to see if more sleep is associated with higher grades.
Experiment: Researchers randomly assign treatments to subjects to measure effects. Can establish causation if well-designed (randomization, control, replication).
Example: Randomly assigning students to either 6 or 8 hours of sleep and measuring their test performance.
Confounding Variable: A variable that influences both the explanatory and response variables, making it impossible to isolate causation.
Example: In a study on coffee and heart disease, age could be a confounder (older people drink more coffee and have higher heart disease risk).
Random Assignment: Subjects are randomly assigned to treatment groups to balance confounding variables.
Key phrase: "Randomly assigned to treatment groups to create roughly equivalent groups."
Random Sampling: Selecting a sample randomly from a population to ensure representativeness.
Key phrase: "Randomly selected from the population to generalize results."
Lurking Variable: A variable not accounted for in the study that may affect the response.
Example: Ice cream sales and drowning deaths are correlated, but the lurking variable is temperature (hotter days lead to both).
Causation vs. Association:
Association: Variables are related, but one does not necessarily cause the other (observational studies).
Placebo Effect: Subjects respond to a treatment even if it’s inert (e.g., a sugar pill).
Solution: Use a placebo group in experiments.
Blinding: Subjects (single-blind) or both subjects and researchers (double-blind) don’t know who receives which treatment to reduce bias.
Blocking: Grouping subjects by a characteristic (e.g., age, gender) before random assignment to reduce variability.
Example: Blocking by grade level before assigning a new teaching method.
Matched Pairs Design: Subjects are paired (or the same subject receives both treatments) to control for variability.
Example: "Students were randomly assigned to either a new study app or traditional notes"-experiment.
Explain Causation Limitations
If experiment: "Can establish causation if well-designed (random assignment, control, replication)."
List Potential Confounding Variables
Example: In a study on exercise and weight loss, diet and metabolism could be confounders.
Describe How to Improve the Study
For experiments: "Use random assignment, blinding, and a control group to isolate the treatment effect."
Interpret Results in Context
Correction: Observational studies can only show association, not causation, due to confounding variables.
Mistake: Forgetting to mention random assignment in experiments.
Correction: Always state that random assignment balances confounding variables to allow causation claims.
Mistake: Confusing random sampling (for generalization) with random assignment (for causation).
Correction:
Mistake: Ignoring the placebo effect in experiments.
Correction: Always mention using a placebo group to control for psychological effects.
Mistake: Not identifying confounding variables in observational studies.
Tricky Distinction: The exam often asks you to explain why an observational study cannot prove causation but an experiment can. Always tie this to confounding variables and random assignment.
Common FRQ Setup:
Questions:
Calculator Pitfall: Not relevant for this topic (no calculations), but be precise with terminology (e.g., "random assignment" vs. "random sampling").
Real-World Trap: The exam may describe a study that seems like an experiment but isn’t (e.g., "Doctors gave a new drug to patients who volunteered"). This is not random assignment-observational study.
A researcher wants to study whether a new fertilizer increases crop yield. She applies the fertilizer to one field and leaves another field untreated, then compares yields. This is an example of: (A) An observational study because the researcher did not randomly assign treatments. (B) An experiment because the researcher imposed a treatment. (C) An observational study because the researcher did not control for weather. (D) An experiment because the researcher measured a response variable.
Answer: (A) This is an observational study because the fields were not randomly assigned (e.g., the treated field might have better soil).
A school district surveys 200 high school students about their homework hours and stress levels. They find that students who report more homework hours also report higher stress levels. (a) Is this an observational study or an experiment? Explain. (b) Can the district conclude that more homework causes higher stress? Why or why not?
Answer: (a) This is an observational study because the researchers did not assign homework hours; they only observed existing data. (b) No, the district cannot conclude causation because there may be confounding variables (e.g., students with more difficult classes may have both more homework and higher stress).
Join 4M+ learners. Unlock unlimited quizzes, wrong-answer tracking, flashcards + reminders, study guides, and 1-on-1 challenges.