Unit 2: Research Methods and Causal Inference
How psychologists turn a hunch about the mind into a study that can actually rule out the other explanations.
5 topics · 15 questions
Each topic is a path, not a menu: work down it in order and every question builds on the one above. Click any of them to watch the answer draw itself.
Correlation Is Not Causation
- How it worksHow does a hidden third variable like hot weather create a correlation between two unrelated things?
- Worked exampleFind the missing explanation for a study showing kids who watch more TV have lower reading scores, given that TV time and parental supervision both vary.
- Common mistakeWhy doesn't a strong correlation of 0.8 mean one variable directly causes the other?
Your turn1 of 2
A study reports a correlation of 0.653 between two variables. Find r squared, the percent of variance in one variable statistically associated with the other.Designing a Controlled Experiment
- How it worksHow does random assignment turn two groups of different people into groups that are, on average, the same?
- How it worksHow does an experiment isolate one cause by holding every other variable constant except the one being tested?
- Common mistakeWhy isn't a study with a huge sample automatically a well-controlled one?
Your turn1 of 2
A researcher randomly assigns participants to a treatment group and a control group rather than letting people choose which group to join. A student assumes random assignment mainly just makes the study look more official, without doing anything to the actual composition of the groups. What does random assignment actually accomplish?Confounding Variables
- How it worksHow does a confound sneak into a study by being tied to both the cause and the effect at once?
- Worked exampleFind the confound in a study of 500 people showing that those who exercise regularly report less depression.
- Common mistakeWhy doesn't controlling for one confound guarantee a study has controlled for all of them?
Your turn1 of 2
A study finds that people who exercise regularly report depression scores 9 points lower on average than non-exercisers. Researchers determine that 0.524 of this gap is explained by a confound, higher income among exercisers. Find how many points of the gap remain after accounting for the confound.Reliability and Validity of a Measure
- How it worksHow does reliability, giving the same answer twice, differ from validity, measuring the right thing?
- Worked exampleFind which is the problem, reliability or validity, for a personality test that gives wildly different scores to the same person a week apart.
- Common mistakeWhy isn't a highly reliable test automatically a valid one?
Your turn1 of 2
A test is split into two halves, and scores on the two halves correlate at 0.683. Using the Spearman-Brown formula, find the estimated reliability of the full-length test.The Placebo Effect and Double-Blind Design
- How it worksHow does a double-blind design stop both the patient's and the researcher's expectations from contaminating the result?
- Going furtherWhat happens to a drug's measured effect once you subtract out what the placebo group improved on its own?
- Common mistakeWhy isn't the placebo effect just patients lying about feeling better?
Your turn1 of 2
In a drug trial, the treatment group improves by 22 points on average, while the placebo group improves by 8 points on its own. Find the drug's true effect after subtracting out what the placebo group improved on its own.