Observational vs Experimental Studies Examples in Statistics
Start with the recap, study the fully worked examples, then use the practice problems to check your understanding of Observational vs Experimental Studies.
This page combines explanation, solved examples, and follow-up practice so you can move from recognition to confident problem-solving in Statistics.
Concept Recap
Observational studies gather data by watching subjects in their natural setting without any intervention, while experimental studies deliberately assign treatments to subjects and measure the outcomes. Only experiments, through random assignment, can establish cause-and-effect relationships.
Observational: Compare smokers to non-smokers (you didn't assign smoking). Experimental: Randomly assign people to take a drug or placebo (you controlled the treatment). Only experiments prove causation.
Read the full concept explanation →How to Use These Examples
- Read the first worked example with the solution open so the structure is clear.
- Try the practice problems before revealing each solution.
- Use the related concepts and background knowledge badges if you feel stuck.
What to Focus On
Core idea: Observational vs Experimental Studies checks whether the study design supports a fair comparison before interpreting the outcome.
Common stuck point: Students often know a procedure related to observational vs experimental studies but skip the recognition step: Did the study use a design feature that makes the groups comparable before the outcome is measured? That leads to a calculation or graph that looks reasonable but answers a different question.
Sense of Study hint: Ask: Did the study use a design feature that makes the groups comparable before the outcome is measured?
Worked Examples
Example 1
mediumA cohort study follows smokers and non-smokers for years and compares lung cancer rates. Why is this NOT an experiment?
Answer
Smoking was not randomly assigned
First step
1
The researcher did not assign smoking; participants chose.
See the full worked solution + why-it-works coaching
SetupKey insightWhy it worksCommon pitfallConnection
Unlock answer keys One Family plan — every worked solution, all subjects
Example 2
mediumA teacher randomly assigns half her class to use a new flashcard app and half to study without it, then compares quiz scores. Identify the explanatory and response variables.
Example 3
mediumAn observational study finds chocolate eaters have lower BMI. State one plausible confounder.
Example 4
hardA natural experiment uses a policy change as a treatment. Is it observational or experimental in the strict sense?
Example 5
hardA class of uses two textbooks. The teacher gives book A to the back row and book B to the front. Why isn't this a good experiment?
Example 6
challengeResearchers want to test whether a new website layout boosts purchases. They randomly serve layout A or B to each visitor. Identify the design type and one threat to validity.
Example 7
mediumA medical team wants to know if a new sleep mask improves rest. Design (A) recruits 100 people who already wear masks and 100 who don't. Design (B) randomly gives 100 people masks and withholds them from 100 others. Identify each design and which supports a causal conclusion.
Example 8
mediumResearchers find smokers have higher lung cancer rates than nonsmokers in a 20-year prospective study with no assignment. A critic says 'maybe genetics drive both.' Identify the study type and explain how a randomized trial would address the critic.
Example 9
mediumA meteorologist compares crop yields in two counties that received different rainfall amounts (no intervention). Classify the study and explain why it cannot establish that rain alone caused the yield difference.
Example 10
hardA study finds that towns adopting a new traffic-camera program saw accidents drop 20% the next year. No towns were randomly assigned. Identify the study type and at least two threats to a causal interpretation.
Example 11
hardA government compares income growth in two regions: one received a new tax credit, the other did not, with no random assignment. Researchers call this a 'natural experiment.' Is it strictly observational or experimental? Explain.
Example 12
hardA randomized trial gives one group of patients a new drug and finds 15% recovery vs 5% in the placebo group. A skeptic asks 'How do we know the drug, not chance, caused the gap?' Answer with the design feature and the statistical tool.
Example 13
challengeA school board wants causal evidence that smaller class sizes improve scores. Sketch a randomized design and explain why a comparison of currently small and large classes would not give the same evidence.
Example 14
easyClassify each study as observational or experimental: (a) Researchers track the diets of 1000 people and record their cholesterol levels. (b) Researchers randomly assign 100 people to either a low-fat or high-fat diet and measure cholesterol after 6 months.
Example 15
mediumA study finds that people who drink green tea have lower rates of heart disease. The study surveyed 5000 tea drinkers and 5000 non-tea drinkers. Can we conclude that green tea prevents heart disease? Explain, identifying the type of study.
Practice Problems
Try these problems on your own first, then open the solution to compare your method.
Example 1
easyResearchers record the diets people already follow and track their health, without assigning diets. Observational or experimental?
Example 2
easyResearchers randomly assign people to a low-sugar or normal diet and compare outcomes. Observational or experimental?
Example 3
easyWhich type of study can establish cause-and-effect: observational or experimental?
Example 4
easyA study compares lung health of smokers and nonsmokers, with no one assigned to smoke. What kind of study is this?
Example 5
easyTrue or false: an observational study can still show a correlation between two variables.
Example 6
easyIn which study type does the researcher actively manipulate the explanatory variable?
Example 7
easyA nutritionist surveys what 500 people ate yesterday and their mood. No assignment occurs. Observational or experimental?
Example 8
easyThe key feature that turns a comparison study into an experiment is what?
Example 9
mediumA headline says 'People who drink coffee live longer,' based on a survey of coffee habits and lifespan. Identify the study type and why the headline overreaches.
Example 10
mediumTo study whether exercise lowers stress, design (A) randomly assigns people to exercise or not; design (B) compares people who already exercise to those who don't. Which is experimental and which can claim causation?
Example 11
mediumA study finds neighborhoods with more parks have lower obesity. A reporter concludes parks reduce obesity. Identify the study type and one confounder undermining the causal claim.
Example 12
mediumA pharmaceutical trial randomly gives a drug or placebo and measures recovery. Classify the study and name the feature that allows a causal conclusion.
Example 13
mediumWhy can an experiment claim causation while an observational study with the same correlation cannot?
Example 14
mediumResearchers want to know if a vaccine prevents flu but cannot ethically withhold it once effective. They instead track vaccinated vs unvaccinated volunteers. What study type results, and what limitation must they note?
Example 15
mediumA study compares test scores of students who chose to attend a summer camp versus those who didn't, finding camp-goers score higher. Classify the study and explain why it can't prove the camp helped.
Example 16
mediumClassify each: (1) randomly assigning fertilizer to plots; (2) recording which wild plants grew tallest near a river. Which is experimental?
Example 17
challengeA study reports that hospital patients given a certain therapy die more often than those not given it, concluding the therapy is harmful. Explain how this observational design could mislead even if the therapy is actually helpful.
Example 18
challengeTwo studies on the same question: an observational study (n=100,000) and a randomized experiment (n=500), reach opposite causal conclusions. Which should be trusted for causation and why, despite the size gap?
Example 19
challengeDesign a study to test whether standing desks improve productivity that is genuinely experimental, and explain the single change that would turn it back into an observational study.
Example 20
mediumA school compares grades of students who joined a study app to those who didn't, with no assignment. Classify the study and explain why a cause-effect headline would be unjustified.
Example 21
easyA study tracks whether students who already play music have higher GPAs. Observational or experimental?
Example 22
easyResearchers randomly assign patients to drug A or B and compare cure rates. Observational or experimental?
Example 23
easyA census records ages, incomes, and addresses of every household. Observational or experimental?
Example 24
mediumA confounding variable is one that ____.
Example 25
mediumIn an experiment, blocking is used to ____.
Example 26
mediumA double-blind experiment is one in which ____.
Example 27
mediumA study randomizes which restaurant a customer is sent to. Is this an experiment?
Example 28
hardWhy does random assignment justify a causal claim?
Example 29
hardA study assigns volunteers (not a random sample) to two treatments at random. The sample is volunteer-based. Random assignment yes/no? Random sample yes/no?
Example 30
hardMatched-pair design is best classified as a type of ____ study.
Example 31
easyA pollster phones randomly selected adults to ask whom they will vote for. Observational or experimental?
Example 32
mediumA pharmaceutical trial divides patients into treatment and placebo groups by lottery. Best label: observational, experimental, or sample survey?
Example 33
mediumA study reports that people who own dogs have lower blood pressure. Without random assignment, what cannot be concluded?
Example 34
challengeA researcher claims that a study showing teens who sleep less have lower grades 'proves' sleep deprivation causes poor grades. State the strongest objection.
Example 35
easyA biologist records the songs of wild birds at sunrise without disturbing them. Observational or experimental?
Example 36
easyA teacher randomly assigns half her class to a new reading app and the other half to the old worksheet, then compares quiz scores. Observational or experimental?
Example 37
easyA dentist surveys 200 patients about their flossing habits and records cavity counts from past charts. Observational or experimental?
Example 38
easyResearchers split 60 mice into two groups by coin flip; one group gets a new feed, the other the standard feed. Observational or experimental?
Example 39
easyA government tracks unemployment by phone survey. Observational or experimental?
Example 40
mediumA study compares cancer rates between people living near power lines and those who don't, with no assignment of residence. Identify the study type and one confounder threatening a causal claim.
Example 41
mediumA clinical trial randomly assigns volunteers to drug or placebo, double-blind. Why does this design allow a causal interpretation that an observational study would not?
Example 42
mediumA nutrition lab feeds rats one of three diets chosen by random number table and weighs them after eight weeks. Observational or experimental, and why?
Example 43
mediumA school principal compares standardized scores of students who joined an after-school program with those who didn't. The program was voluntary. Identify the study type and one alternative explanation for any score difference.
Example 44
mediumIn an experiment, researchers give half the subjects a placebo. What is the placebo's role with respect to the treatment group?
Example 45
hardAn ed-tech firm A/B-tests a new homepage by randomly serving version A to half the visitors and version B to the other half, then comparing signup rates. Is this experimental or observational, and what is the practical name for this design?
Example 46
hardA psychologist randomly assigns 40 students to either a meditation class or a control activity, then measures stress. After the study, several students drop out. What design feature is preserved and which is threatened?
Example 47
hardA study tracks 5,000 commuters for ten years to see whether walking to work predicts lower BMI. No assignment occurs. Classify the study and one major confounder.
Example 48
mediumA school wants to know if a new teaching method improves maths scores. Method 1: compare scores of students who chose the new method with those who chose the old method. Method 2: randomly assign students to each method. Which approach is better and why?
Example 49
hardExplain why some research questions can only be studied observationally, not experimentally. Give two examples and discuss the ethical or practical reasons.
Related Concepts
Background Knowledge
These ideas may be useful before you work through the harder examples.
experimental designcorrelation vs causation