An experiment is a research method in which the researcher manipulates an independent variable, uses random assignment to control confounding variables, and measures changes in a dependent variable. It is the only research method that can establish cause-and-effect relationships.
An experiment is the one research method where psychologists don't just watch behavior, they change something on purpose and see what happens. The researcher manipulates the independent variable (the thing being changed), measures the dependent variable (the outcome), and keeps everything else as equal as possible between an experimental group (gets the treatment) and a control group (doesn't). Because the researcher controls the manipulation, an experiment can answer the question every other method can only hint at. Did X actually cause Y?
Two features make an experiment legit. First, random assignment gives every participant an equal chance of landing in either group, which spreads out individual differences and shuts down confounding variables. Second, operational definitions spell out exactly how each variable is manipulated and measured (not "stress" but "score on a 10-item stress survey"), so the study can be replicated. Strip away random assignment or the manipulation and you no longer have a true experiment, no matter how scientific the study looks.
The experiment lives in Topics 1.2-1.4 (Research Methods, Defining Psychological Science: The Experimental Method, and Selecting a Research Method), but it's really the engine behind the entire course. Almost every finding you learn, from Pavlov's conditioning (Topic 4.2) to conformity studies (Topic 9.3) to drug-treatment trials (Topic 8.10), exists because someone ran an experiment. The AP Psych science practices ask you to evaluate research design, interpret data, and argue from evidence, and you can't do any of that without knowing what makes an experiment different from a correlational study or a survey. The single most-tested idea attached to this term is simple to say and easy to mess up under pressure. Only experiments can establish causation.
Keep studying AP Psychology Unit 1
Random Assignment (Topic 1.3)
Random assignment is what turns a comparison into a true experiment. Without it, your groups might differ before the manipulation even starts, and any "effect" you find could just be a pre-existing difference.
Selecting a Research Method (Topic 1.4)
Experiments aren't always possible or ethical. You can't randomly assign children to trauma or force people into sleep deprivation for years, which is exactly why correlational studies, case studies, and longitudinal designs exist as alternatives.
Classical and Operant Conditioning (Topics 4.2-4.3)
Behaviorism was built on lab experiments. Pavlov manipulated which stimuli paired with food, and Skinner manipulated reinforcement schedules, so Unit 4 learning content doubles as a gallery of independent and dependent variables in action.
Evaluating Treatments of Disorders (Topic 8.10)
Whether a therapy or medication actually works is an experimental question. Randomized trials with placebo control groups are how psychologists separate real treatment effects from people simply expecting to feel better.
Multiple-choice questions love handing you a study scenario and asking you to identify the independent variable, the dependent variable, the experimental group, or whether the study is even an experiment at all. A classic stem looks like the practice question "Which term refers to a research method in which variables are manipulated to measure their impact on other variables?" Free-response questions go further. The 2021 SAQ described Mr. Gomez splitting his class by who arrived early, and the points hinged on recognizing that early arrivers were NOT randomly assigned, so causal conclusions were off the table. The 2018 SAQ tested whether you could tell a survey apart from an experiment. On the revised exam's research-based free-response questions, expect to design or critique a study, which means naming the IV and DV, explaining random assignment, and writing operational definitions. Vague answers like "the experiment proves music helps memory" don't earn points; precise method language does.
Both can involve two variables and a pile of data, but only an experiment manipulates a variable and randomly assigns participants. A correlational study just measures variables as they naturally occur, so it can show that stress and absences move together (like the 2018 SAQ survey) but it can never tell you which one causes the other, or whether a third variable causes both. If nothing was manipulated, it's not an experiment, full stop.
An experiment manipulates an independent variable and measures a dependent variable, making it the only research method that can establish cause and effect.
Random assignment, not just having two groups, is what makes an experiment a true experiment because it controls for confounding variables.
Operational definitions specify exactly how variables are manipulated and measured, which makes an experiment replicable and earns you points on research-design FRQs.
If participants ended up in groups by choice, timing, or convenience (like students arriving early to class), the study is not a true experiment and cannot support causal claims.
Experiments aren't always ethical or practical, which is why psychologists also use correlational studies, case studies, and longitudinal designs covered in Topic 1.4.
An experiment is a research method where the researcher manipulates an independent variable, randomly assigns participants to experimental and control groups, and measures a dependent variable. It's defined in Topic 1.3 and is the only method that can prove causation.
No. A correlational study only shows that two variables move together, like the stress-and-absences survey in the 2018 SAQ. Only an experiment, with manipulation and random assignment, can establish that one variable causes another.
An experiment changes a variable on purpose and randomly assigns participants to groups; a correlational study just measures variables as they naturally occur. The exam loves scenarios that look experimental but lack manipulation or random assignment.
No. The 2021 SAQ described a teacher who grouped students by who happened to arrive early, which is not random assignment, so it wasn't a true experiment. Two groups plus a comparison is not enough; the researcher must manipulate the variable and randomly assign participants.
The independent variable is what the researcher manipulates (like whether participants hear music while studying), and the dependent variable is the measured outcome (like their test score). A quick check is to ask what the researcher changed versus what the researcher measured.