Psychology research methods explain how psychologists ask questions, collect evidence, and draw careful conclusions about behavior and mental processes. This cheat sheet helps students compare experiments, surveys, case studies, observations, and correlational research. It is useful for planning studies, reading research summaries, and avoiding weak claims about cause and effect.
The core ideas include forming testable hypotheses, identifying independent and dependent variables, choosing representative samples, and controlling confounding variables. Students also need basic descriptive statistics such as mean, median, range, and standard deviation to summarize data. Strong research depends on validity, reliability, ethical treatment of participants, and careful interpretation of results.
Key Facts
- A hypothesis is a testable prediction written in a clear form, such as if the independent variable changes, then the dependent variable will change.
- In an experiment, the independent variable is manipulated by the researcher, and the dependent variable is the measured outcome.
- An experimental group receives the treatment or condition being tested, while a control group does not receive it or receives a standard comparison condition.
- Random assignment means each participant has an equal chance of being placed in any condition, which helps reduce participant differences between groups.
- Correlation is measured from -1 to +1, where r = +1 is a perfect positive relationship, r = -1 is a perfect negative relationship, and r = 0 is no linear relationship.
- Mean is calculated as mean = sum of all scores divided by number of scores.
- Range is calculated as range = highest score - lowest score, and it shows the spread between the extremes.
- Percent change is calculated as percent change = (new value - old value) / old value x 100 percent.
Vocabulary
- Independent Variable
- The factor that the researcher changes or manipulates to test its effect.
- Dependent Variable
- The outcome that is measured to see whether it changes because of the independent variable.
- Confounding Variable
- An outside factor that may affect the dependent variable and make results harder to interpret.
- Operational Definition
- A precise description of how a variable will be measured or manipulated in a study.
- Validity
- The degree to which a study or measurement actually tests what it claims to test.
- Reliability
- The degree to which a study or measurement gives consistent results over time or across observers.
Common Mistakes to Avoid
- Confusing correlation with causation is wrong because a relationship between two variables does not prove that one variable caused the other.
- Forgetting to define variables operationally is wrong because vague terms like stress or memory must be measured in a specific, repeatable way.
- Using a biased sample is wrong because results from a nonrepresentative group may not generalize to the larger population.
- Changing more than one independent variable at once is wrong because it becomes unclear which change caused the difference in the dependent variable.
- Ignoring ethics is wrong because participants must be protected through informed consent, privacy, the right to withdraw, and debriefing when needed.
Practice Questions
- 1 A study records reaction times of 220 ms, 260 ms, 240 ms, 280 ms, and 250 ms. What is the mean reaction time?
- 2 A researcher finds that students studied an average of 4 hours before an intervention and 5 hours after it. What is the percent change in study time?
- 3 A data set has a highest score of 92 and a lowest score of 57. What is the range?
- 4 A researcher finds a strong positive correlation between sleep hours and test scores. Explain why this result does not prove that more sleep directly caused higher scores.
Understanding Research Methods and Experimental Design
Good studies turn broad ideas into actions that can be observed or scored. This is called operational definition. For example, a researcher might define stress as a score on a questionnaire, the amount of cortisol in saliva, or the number of missed school days.
Each choice captures part of the idea, but none captures everything. A measure has construct validity when it truly represents the concept being studied.
Researchers often run a small pilot study first. This can reveal confusing instructions, measures that are too easy, or questions that make participants guess the purpose of the study.
To support a cause and effect claim, researchers must make the comparison groups as similar as possible except for the treatment. Random assignment helps with this, but it does not remove every problem. Participants may improve because they expect help, which is called a placebo effect.
Researchers can reduce this effect by using a comparison activity that seems equally credible. In a single blind study, participants do not know their condition. In a double blind study, the people collecting results do not know either.
Repeated measures designs need care too. When the same people complete several conditions, practice, fatigue, and order can change results. Counterbalancing varies the order across participants.
The people who join a study shape what its findings can mean outside the study. Random assignment is different from random sampling. Assignment creates fairer groups within a study.
Sampling affects whether the sample represents the larger population. Psychology studies often use volunteers, students, or people recruited online because they are available. These groups may differ from the public in age, education, culture, or motivation.
This can limit generalisation. Researchers should report who took part, how many people left before the end, and where data were collected. A large sample does not fix a biased sample.
Numbers need interpretation, not just calculation. A mean can be pulled strongly by one extreme score, such as a participant who sleeps only one hour per night. The median may better describe a skewed set of data.
Standard deviation shows whether scores cluster closely or vary widely. Statistical significance means a result would be unlikely if there were truly no effect in the population. It does not prove that an effect is large, useful, or certain.
Effect size shows the size of a difference or relationship. Replication matters because one result can occur by chance, measurement error, or unusual participants.
Correlations are useful for finding patterns, especially when experiments would be impractical or unethical. A link between social media use and anxiety does not show which one came first. Anxiety might increase online activity, online activity might increase anxiety, or another factor such as loneliness could affect both.
Longitudinal studies measure people over time and can clarify the order of events, though they still cannot prove causation by themselves. Ethical rules protect participants throughout every method. Informed consent, privacy, the right to withdraw, careful use of deception, and a full debriefing help prevent harm and build trust in the evidence.