Descriptive Research Methods

Psychologists choose a research method based on the question they want to answer. AP Psychology separates research into experimental and non-experimental methodologies. Experimental methodology manipulates an independent variable and uses random assignment to groups. The named non-experimental methodologies are case study, correlation, meta-analysis, and naturalistic observation.

Research methods answer different questions. Experiments test causation. Non-experimental methods describe behavior, examine relationships, study unusual cases, or combine previous findings. A survey remains important, but it is a self-report data-collection technique rather than one of the five named methodologies.

Naturalistic observation records behavior in a real-world setting without interfering. Its realism is a strength, but realism does not make the method causal. It can collect quantitative counts, qualitative descriptions, or both.

A case study investigates one person, small group, or unusual situation in depth. Evidence can come from observation, interviews, records, tests, or several sources. It is especially useful when the case is rare or ethically impossible to create in a lab, and it can generate hypotheses for later research.

A meta-analysis statistically combines results from many studies that investigated the same question. It examines the broader pattern across a body of evidence, but the conclusion depends on the quality of the included studies and the rules used to select them.

Correlation measures whether two variables are related. It can show direction and strength and support prediction, but directionality and third-variable problems prevent a causal conclusion. Continue to U0.4 for correlation coefficients, scatterplots, and causal traps.

Experimental methodology manipulates an independent variable and uses random assignment to groups. Researchers then measure the dependent variable while trying to control confounding variables. A well-designed experiment can support a cause-and-effect conclusion. Continue to U0.5 for variables, groups, controls, blinding, and validity.

A survey is a technique for collecting people's reports through questionnaires or interviews. It is tested in AP Psychology, but it is not one of the five named research methodologies. A survey can supply data for a descriptive study or a correlational study, depending on the research question and how the data are analyzed.

A structured interview asks every participant the same planned questions. A Likert scale asks for an ordered response, such as a level of agreement. Self-report can be inaccurate because of memory, misunderstanding, embarrassment, or deliberate distortion, and a large sample is not automatically representative.

Non-experimental methodologies do not combine a manipulated independent variable with random assignment to conditions. That means alternative explanations remain open, even when a study identifies a strong pattern or relationship.

Method first, technique second: a survey describes data collection. Ask what researchers did with the responses before deciding whether the broader study was descriptive, correlational, or part of another non-experimental design.

• Experimental methodology: Research that manipulates an independent variable and uses random assignment to groups.
• Non-experimental methodology: Research without the combination of manipulated independent variables and random assignment required for an experiment.
• Descriptive research: A useful teaching category for research that describes behavior without manipulating variables.
• Naturalistic observation: Observing behavior in a real setting without interference.
• Case study: In-depth study of one person, group, or situation.
• Correlation: A non-experimental methodology that measures the relationship between variables.
• Meta-analysis: Research that statistically combines results from many studies on the same question.
• Survey: A questionnaire or interview technique used to collect self-report data.
• Structured interview: A survey technique in which participants are asked the same planned questions.
• Likert scale: An ordered response scale commonly used to measure attitudes or ratings.
• Self-report bias: Inaccuracy or distortion in information participants report about themselves.
• Social desirability bias: The tendency to answer in a way that appears acceptable or favorable.
• Directionality problem: Uncertainty about which correlated variable may influence the other.
• Third-variable problem: The possibility that another variable explains an observed correlation.
• Population: Full group of interest.
• Sample: Participants actually studied.
• Representative sample: Sample that reflects the population.
• Random sample: Selection method where each population member has equal chance.
• Wording effect: Question phrasing changes responses.
• Demand characteristics: Participants change behavior after guessing study purpose.
• Observer bias: Researcher expectations affect observation.

• Five-method trap: Survey is tested, but it is not one of the five methodologies named in the current CED.
• Survey-label trap: A study can collect survey responses and still be correlational if it measures the relationship between two variables.
• Meta-analysis trap: Meta-analysis does not mean a study that merely uses statistics; it combines results from multiple studies.
• Experiment trap: A comparison between groups is not automatically an experiment. Look for a manipulated independent variable and random assignment.
• Correlation trap: A strong correlation is not evidence of causation; directionality and third variables remain possible.
• Case-study trap: Detailed evidence from one unusual person does not automatically apply to everyone.
• Observation trap: Natural behavior can be realistic while still being affected by observer bias, reactivity, or uncontrolled variables.
• Sampling trap: Random sampling helps represent a population; random assignment creates comparable experimental groups. They solve different problems.

You are ready when you can identify the five methodologies, distinguish survey technique from a broader methodology, and explain why non-experimental evidence cannot establish causation.