Introduction to Psychology · Foundations and Research

Psychological Research Methods

8 min read

On this page 7 sections

In 30 seconds
Why this matters
The college version
Eli explains
Worked example
Key takeaway
Study tools

In 30 seconds

Psychologists answer questions with the : they pose a , derive a testable from a , define their variables operationally, collect data, and draw conclusions that others can replicate. Descriptive methods (, , ) describe behavior; measures how two variables relate, expressed as a . Only experiments — which manipulate an , measure a , and use — can establish cause and effect. A correlation never proves causation because of the directionality and third-variable problems.

Why this matters

In healthcare, randomized experiments (randomized controlled trials) underpin the evidence for whether treatments work, while correlational studies of lifestyle and disease must be read carefully so that association is not mistaken for cause. In education, surveys and naturalistic observation inform teaching practice, but only experimental or strong quasi-experimental designs can support the claim that a method causes learning gains. This material explains research concepts for study purposes and is not clinical or treatment advice.

The college version

1. The Scientific Method

The scientific method is a systematic cycle: (1) ask a research question; (2) develop a hypothesis — a specific, testable prediction, usually derived from a theory (a broader explanation that organizes facts and generates predictions); (3) create an operational definition that turns each abstract concept into a concrete, measurable procedure; (4) collect and analyze data; (5) draw conclusions and publish; and (6) allow replication — repeating a study to see whether its results hold. For example, "aggression" might be operationally defined as the number of times a child hits a doll during a set observation period.

2. Descriptive and Correlational Research

Descriptive research observes and records behavior without manipulating anything:

Case study — an in-depth look at one person or a small group. It yields rich detail but limited generalizability.
Naturalistic observation — watching behavior in its natural setting without interference. It is realistic but offers no control over conditions.
Survey — asking people to self-report. It is efficient but depends on honest, accurate answers. Survey quality hinges on sampling, or who is selected to participate. Random sampling means every member of the population has an equal chance of being chosen, which improves representativeness.

Correlational research measures the relationship between two variables without manipulating them. The result is a correlation coefficient (r), ranging from −1.0 to +1.0, where the sign shows direction and the number shows strength. Correlations reveal associations but cannot establish causation because of the directionality problem (does A cause B, or does B cause A?) and the third-variable problem (a hidden factor C may cause both A and B).

3. Experimental Research

Experimental research is the only design that can support causal conclusions. It manipulates an independent variable (the presumed cause the researcher changes) and measures a dependent variable (the outcome). Participants in the experimental group receive the treatment, while those in the control group do not, providing a baseline for comparison. Random assignment — placing participants into conditions by chance — spreads preexisting differences across groups so that any later difference can be attributed to the manipulation rather than to a confounding variable (an outside factor that varies along with the independent variable). Note that random assignment (who goes into which group) is different from random sampling (who is in the study at all).

Two further concepts guard the quality of measurement: reliability means a measure gives consistent results, while validity means it actually measures what it claims to measure. Statistical significance indicates that a result is unlikely to have occurred by chance alone.

How it works

Ask a precise research question.
Turn a theory into a testable hypothesis.
Operationally define every variable.
Choose a design: descriptive, correlational, or experimental.
For experiments, manipulate the independent variable and use random assignment to experimental and control groups.
Measure the dependent variable and check for statistical significance.
Publish so others can replicate and refine the finding.

Common confusions

Do not confuse	With	Difference
Correlation	Causation	Association is not one variable causing the other
Random assignment	Random sampling	Assignment places people into groups; sampling selects who is studied
Reliability	Validity	Consistency vs accuracy
Hypothesis	Theory	A specific prediction vs a broad explanation
Independent variable	Dependent variable	What you change vs what you measure

Memory aids

For variables, remember "M.I.D.": you Manipulate the Independent variable and Measure the Dependent variable. For correlation versus causation, remember the phrase "correlation is not causation" — the two C's sound alike, but they are not the same thing.

Quick review

Topic Recap

Psychology uses the scientific method to turn questions into testable hypotheses with operationally defined variables. Descriptive methods describe behavior, correlational research measures relationships (expressed as correlation coefficients), and experimental research — with independent and dependent variables, experimental and control groups, and random assignment — tests causation. Correlation never proves causation, good measurement must be both reliable and valid, and replication and statistical significance keep the whole system honest.

Knowledge Check

Which research design can support cause-and-effect conclusions?
What does a correlation coefficient of −0.80 tell you?
What is the difference between random assignment and random sampling?
A study finds that ice cream sales and drowning deaths are correlated. Which problem does a hidden "hot weather" variable illustrate?
A bathroom scale gives the same wrong weight every day. Is it reliable, valid, both, or neither?

Answers and Rationales

Experimental research — only experiments manipulate an independent variable and use random assignment.
A strong negative relationship — as one variable rises, the other tends to fall, and the association is strong.
Random assignment places participants into experimental versus control groups by chance; random sampling selects who is in the study to represent the population.
The third-variable problem — a hidden factor (weather) causes both, so neither causes the other.
Reliable but not valid — the scale is consistent (reliable) but does not measure the true value (not valid).

Eli explains

The same idea, in plain words

Explain it like I’m 10

Research methods are the rules psychologists follow so their findings can be trusted. Instead of "I have a hunch," they ask a specific question, make a prediction they can test, and measure things in a way anyone else could copy.

A useful comparison: think of a detective versus a rumor. A rumor spreads because it "feels right"; a detective gathers evidence, checks it, and lets others review the case — psychology's methods are the detective's toolkit. For correlation versus causation, ice cream sales and drowning deaths both rise in summer: they are correlated, but ice cream does not cause drowning — warm weather drives both. To test a causal claim, you need an experiment, not just the observation that two things rise together.

Where it stops being exact: real-world measurement is never perfect. An operational definition chooses one way to measure an abstract idea (like "stress"), so two studies may measure it differently. Samples are never perfectly representative of everyone, and statistical significance is a statement of probability, not a guarantee of truth. Results can also fail to replicate, which is exactly why science keeps checking itself.

Simple Example

You wonder whether sleep affects memory. You form a hypothesis: "Students who sleep 8 hours will recall more words than students who sleep 4 hours." You operationally define "recall" as the number of words from a 20-item list correctly remembered the next day. Then you design a study to test the prediction.

Worked example

A researcher asks: "Does sleep duration affect memory recall?" and hypothesizes that more sleep leads to better recall.
Operationally, "recall" is defined as the number of words remembered from a 20-item list, and "sleep" as hours recorded by a sleep tracker.
In a correlational study, the researcher measures sleep and recall in many students and finds r = +0.45 (a moderate positive relationship). This shows association only — it cannot tell us whether more sleep causes better recall (directionality problem) or whether a third variable such as overall health causes both (third-variable problem).
To test causation, the researcher runs an experiment: randomly assign participants to an 8-hour group (experimental group) or a 4-hour group (control group), hold other factors constant, and measure recall (the dependent variable).
If the 8-hour group recalls significantly more words, and random assignment ruled out confounding variables, the researcher can more confidently infer that sleep influenced recall — though replication is still needed before the finding is trusted.
Methodological limits: laboratory sleep conditions may not reflect real life, the sample may not generalize to other populations, and statistical significance is a probability statement, not proof.

Key takeaways

High yield: Correlation does not imply causation (directionality + third-variable problems).
High yield: Only experiments — with manipulation and random assignment — support causal conclusions.
High yield: Random assignment (to groups) is not the same as random sampling (into the study).
High yield: Independent variable = manipulated; dependent variable = measured.
High yield: Reliability = consistency; validity = accuracy.
High yield: Operational definitions make variables measurable and studies replicable.
A correlation coefficient's sign shows direction; its number shows strength.
Statistical significance is about probability, not proof.

Keep learning

Ready to build on this? Continue to the next lesson.

Study toolsYou’ll learn to · Key vocabulary

You’ll learn to

List the steps of the scientific method and explain how research questions, hypotheses, theories, and operational definitions work together.
Describe the three descriptive methods — case study, naturalistic observation, and survey — and the role of sampling.
Explain correlational research, the correlation coefficient, and the directionality and third-variable problems.
Describe experimental research — independent and dependent variables, experimental and control groups, random assignment, and confounding variables — and distinguish correlation from causation, reliability from validity, and define replication and statistical significance.

Key vocabulary

Scientific method: A systematic cycle of question, hypothesis, measurement, and testable conclusions
Research question: The specific question a study aims to answer
Hypothesis: A testable prediction, often derived from a theory
Theory: A broad explanation that organizes facts and generates hypotheses
Operational definition: A concrete, measurable statement of a variable
Replication: Repeating a study to check whether results hold
Descriptive research: Observing and recording behavior without manipulation
Case study: In-depth study of one person or small group
Naturalistic observation: Watching behavior in its natural setting
Survey: A self-report questionnaire
Sampling: Selecting who participates in a study
Correlational research: Measuring the relationship between variables
Correlation coefficient: A number (−1.0 to +1.0) showing direction and strength of a relationship
Directionality problem: Not knowing which variable causes which
Third-variable problem: A hidden factor may cause both variables
Experimental research: Manipulating a variable to test cause and effect
Independent variable: The variable the researcher changes
Dependent variable: The outcome the researcher measures
Experimental group: Receives the treatment
Control group: Does not receive the treatment
Random assignment: Placing participants into groups by chance
Random sampling: Choosing participants so everyone has an equal chance
Confounding variable: An outside factor that varies with the independent variable
Reliability: Consistency of a measurement
Validity: Whether a measure captures what it claims
Correlation vs causation: Association between variables is not one causing the other
Statistical significance: A result unlikely to be due to chance

Educational content only. It is not medical, legal or professional advice. Found an error? Tell us.