← Back to decks
Sign in to save your progress, vote, and build your own decks.Sign in
Research Methods in Psychology
Psychology research methods explained: hypotheses, independent and dependent variables, random assignment, placebos and double-blind trials, why correlation isn't causation, case studies, surveys, significance and ethics. Learn to spot what a study can and can't show. For AP Psychology students and curious adults.
Hypothesis
A testable prediction about how two or more things are related, stated before the data are collected.
Falsifiable
A claim is falsifiable if some possible result could show it to be wrong. Claims that fit any outcome can't be tested.
Operational definition
Saying exactly how a concept will be measured, e.g. "aggression" as the number of times a child hits a toy in 10 minutes.
Independent variable (IV)
The factor the researcher deliberately changes or assigns, to see what effect it has.
Dependent variable (DV)
The outcome the researcher measures, to see whether the independent variable changed it.
Experimental group vs control group
The experimental group gets the treatment; the control group doesn't (or gets a placebo), giving a baseline to compare against.
Random assignment
Using chance to decide which participants go into which group, so the groups start out similar on average.
Researchers test whether music helps memory. Volunteers who chose to study with music go in one group, the rest in another. What's the main flaw?
- ✓ No random assignment, so the groups may differ in other ways
- No operational definition of music
- The sample is random, so it can't generalize
- The dependent variable is music
Random sampling
Choosing participants so everyone in the population has an equal chance of being picked. It helps results generalize.
Confounding variable
A factor other than the independent variable that differs between groups, so it could explain the result instead.
Placebo
A fake treatment, such as a sugar pill, that lets researchers separate a real effect from the effect of expecting one.
Double-blind procedure
Neither the participants nor the researchers collecting data know who is in which group, which limits both sides' expectations.
Experimenter bias
When a researcher's expectations nudge how they treat participants or record results, often without meaning to.
In a drug trial, the nurses giving pills don't know which are real. The patients don't either. What is this called?
- Single-blind
- Random sampling
- ✓ Double-blind
- Naturalistic observation
Demand characteristics
Cues in a study that hint at what's expected, so participants change their behavior to fit (or to defy) it.
Correlation coefficient (r)
A number from -1 to +1 showing the direction and strength of a linear relationship. 0 means no linear relationship.
Directionality problem
In a correlation, you can't tell which variable affects the other, e.g. does poor sleep cause low mood, or the reverse?
Third-variable problem
A correlation between two things may exist because a third factor drives both, not because one causes the other.
A study finds that teens who use their phones more at night report lower mood. What is the strongest conclusion?
- Night-time phone use causes low mood
- Low mood causes night-time phone use
- ✓ The two are related, but this design can't show which causes which
- There is no real connection, because it's only a correlation
Which correlation coefficient shows the strongest relationship?
- +0.45
- ✓ -0.82
- +0.10
- 0.00
Case study
An in-depth look at one person or a small group. Rich detail, but the findings may not apply to anyone else.
Naturalistic observation
Watching and recording behavior in its usual setting without interfering. It describes, but can't show cause.
Social desirability bias
Survey answers drift toward what makes people look good, such as under-reporting bad habits.
Longitudinal vs cross-sectional study
Longitudinal follows the same people over time; cross-sectional compares different age groups at one moment.
Standard deviation
A measure of how spread out scores are around the mean. A small value means scores cluster tightly.
Statistical significance
A result unlikely to be due to chance alone, by a set threshold (often p < .05). It doesn't tell you how big or important it is.
Effect size
How large a difference or relationship is, separate from whether it's statistically significant.
A study of 20,000 people finds a statistically significant difference in reaction time between two groups. What can you safely say?
- The difference must be large
- ✓ The difference is unlikely to be chance alone, but could still be tiny
- The finding will replicate
- The result proves the hypothesis
Meta-analysis
A method that combines the results of many studies on the same question to estimate an overall effect.
Informed consent and debriefing
Participants agree after being told what a study involves; afterwards they're told its true purpose, including any deception.