Research methods run through the whole redesigned AP Psychology course (Science Practice 2, Research Methods and Design, in the CED) for a reason: They anchor both the AAQ (Article Analysis Question) and the EBQ (Evidence-Based Question) in the free-response section, and they thread through MCQs across every other unit. Getting comfortable with research methods early makes the rest of the course easier to connect.
This article covers everything the exam expects you to know about research methods, at the level of depth the AAQ and EBQ test.
Research designs
The CED distinguishes experimental from non-experimental designs (case study, correlation, meta-analysis and naturalistic observation), and surveys are a common way of collecting data. Know what each does, when it's used, and its main limitation.
Experiment. Manipulates an independent variable, measures a dependent variable, and uses random assignment to control for confounds. The only design that can establish causation. Limitation: Artificial settings can limit generalizability (external validity).
Correlational study. Measures the relationship between two variables without manipulating either. Reports a correlation coefficient (r, ranging from -1 to +1). Can identify relationships but not causes. Limitation: Cannot establish causation; confounding variables may explain the relationship.
Case study. In-depth study of a single individual, group or event. Useful for rare phenomena and generating hypotheses. Limitation: Findings often don't generalize; researcher bias can shape interpretation.
Meta-analysis. Statistically combines results of many studies on the same question. Limitation: Only as good as the studies included.
Naturalistic observation. Observing behavior in its natural setting without intervention. High ecological validity. Limitation: No control over variables; observer bias possible.
Survey. Uses questionnaires or interviews to collect self-reported data from a sample. Efficient for large samples. Limitation: Social desirability bias, sampling bias, low response rates.
Variables
Every experimental design revolves around variables. The exam consistently tests four types.
Independent variable (IV): The variable the researcher manipulates. In a sleep-and-memory study, hours of sleep is the IV.
Dependent variable (DV): The variable the researcher measures. Memory recall is the DV in the sleep study.
Confounding variable: An uncontrolled variable that varies with the IV and could explain results. If the well-rested group also happened to eat breakfast and the sleep-deprived group didn't, breakfast is a confound.
Operational definition: The specific, measurable way a variable is defined in the study. Memory operationalized as the number of words correctly recalled from a 30-word list after a 24-hour delay is an operational definition. Just saying memory isn't.
Sampling
Two distinctions matter for the exam.
Random sampling vs random assignment. Random sampling is selecting your sample from the population randomly (giving every population member an equal chance). It improves generalizability. Random assignment is assigning participants to experimental conditions randomly (giving every participant an equal chance of ending up in any group). It controls for confounds within the study.
A study can have one, both, or neither. Manipulating an independent variable and randomly assigning participants to conditions is what makes a study an experiment.
Sample types. Convenience sample (using whoever's available, like college students) is common but limits generalizability. Stratified sample (selecting from defined subgroups in proportion to the population) improves representativeness.
Threats to validity
The exam tests both internal and external validity threats.
Internal validity is the degree to which a study shows that the independent variable, and not something else, caused the change in the dependent variable. Threats: Confounding variables, experimenter bias (unintentionally influencing participants), demand characteristics (participants guessing the hypothesis and adjusting behavior), placebo effect (change from expectation, not treatment).
External validity is the degree to which findings generalize beyond the study. Threats: Small or unrepresentative sample (WEIRD samples, Western Educated Industrialized Rich Democratic, are a common example), artificial lab settings that don't match real-world conditions.
Controls that address these threats include random assignment (internal), single-blind and double-blind procedures (internal), control groups (internal), and larger, more diverse samples (external).
Ready to boost your grades?
Join 1M+ students who have used Cognito to ace their exams.
Get started for free!
Statistics
The exam covers descriptive and inferential statistics mostly at a conceptual level. You may need to calculate simple measures (mean, median, mode, range) without a calculator, but most questions ask you to interpret.
Descriptive statistics: Measures of central tendency (mean, median, mode) and variability (range, standard deviation). Know when each is most appropriate. Median resists outliers; mean is pulled by them.
Correlation coefficients (r): Range from -1 (perfect negative) to +1 (perfect positive), with 0 being no relationship. Values around 0.1-0.3 are weak, 0.3-0.5 moderate, 0.5+ strong (roughly). Correlation does not imply causation, ever.
Statistical significance: A finding is statistically significant when its p-value is below a threshold (conventionally 0.05), meaning it's unlikely to have occurred by chance. Significant doesn't mean large or important; it means unlikely to be chance.
Effect size: How large the difference or relationship is, independent of significance. A study with a huge sample can produce a statistically significant finding with tiny effect size, which may not be practically meaningful.
Ethics
The APA Ethics Code specifies protections for research participants. The exam tests recognition of these principles.
Informed consent: Participants must be told the general nature of the study and any risks, then agree to participate. Special protections for minors (parental consent).
Confidentiality: Individual data must be kept private and reported in aggregate.
Right to withdraw: Participants can leave at any time without penalty.
Debriefing: After the study, participants must be told the full purpose and any deception used, and any harm addressed.
Protection from harm: Studies must minimize physical or psychological harm. Research with human participants is reviewed by an IRB (Institutional Review Board) before it begins, and deception is allowed only when it's justified and followed by debriefing.
Historical cases the exam may reference: Milgram (obedience, ethical concerns about psychological harm and consent), Zimbardo Stanford Prison Experiment (ethical concerns about researcher intervention and participant harm), Watson's Little Albert (ethical concerns about consent and harm to a child).
For the AAQ, the recurring high-value skill is naming the research design correctly and giving an operational definition. Those two points appear on almost every AAQ. Practice them until they're automatic.
How research methods shows up across the exam
Research methods appears throughout the MCQs. Questions about classical or operant conditioning often reference specific studies (Pavlov, Skinner, Watson) and ask about design elements. Questions about memory reference Loftus (misinformation effect), Sperling (iconic memory), and others. Questions about cognitive biases reference Tversky and Kahneman.
Understanding the research method behind each of these studies is often the difference between a right answer and a wrong one. When an MCQ asks why we know a phenomenon exists, it's asking about the study design, not just the finding.
Study AP Psychology with Cognito
Cognito's AP Psychology course covers research methods in depth with practice AAQs, EBQs and MCQs across every unit.
