Key Research Methods & Study Design to Know for AP Psychology

What You Need to Know

AP Psych research questions usually test whether you can: (1) identify a study type, (2) label variables and controls, (3) spot what can and can’t be concluded (especially causation), and (4) evaluate design quality (bias, confounds, ethics, validity/reliability).

Big idea: different methods answer different questions.

• Descriptive methods (case study, naturalistic observation, survey): describe behavior, can’t show cause.
• Correlational studies: measure relationship between variables, can predict, but can’t show cause.
• Experiments: manipulate an independent variable and measure a dependent variable with random assignment to infer cause-and-effect.

Core “language” you must nail:

• Population vs sample
• Random sampling (for representativeness) vs random assignment (for causality)
• Independent variable (IV), dependent variable (DV), confounding variables
• Operational definition (exactly how you measure/manipulate variables)
• Reliability (consistency) vs validity (accuracy)
• Correlation coefficient $r$ (strength/direction)

Critical reminder: Only true experiments with random assignment allow causal conclusions.

Step-by-Step Breakdown

A. How to dissect any research scenario (fast)

1. Name the method
   • Are they just observing/describing? → descriptive
   • Are they measuring two variables to see if they relate? → correlational
   • Are they manipulating something and using random assignment? → experiment
2. Identify the variables (if any)
   • IV: what’s manipulated (or the grouping factor in quasi-experiments)
   • DV: what’s measured as the outcome
3. Check for causation requirements
   • Manipulation of IV + random assignment + control of confounds → causation is possible
4. Find the operational definitions
   • How exactly are “stress,” “aggression,” “memory,” etc. measured?
5. Evaluate design quality
   • Sampling bias? Confounds? Demand characteristics? Experimenter bias? Placebo effects?
6. Evaluate ethics
   • Informed consent? Deception justified + debrief? Protection from harm? Confidentiality?

B. Designing a solid experiment (what AP expects)

1. State a testable hypothesis
   • Example: “If participants sleep $8\,\text{hours}$, then their recall score will be higher than participants who sleep $4\,\text{hours}$.”
2. Operationally define the IV and DV
   • IV levels: $8\,\text{hours}$ vs $4\,\text{hours}$ of sleep
   • DV: “number of words recalled from a $20$-word list”
3. Choose the design
   • Between-subjects: different participants in each condition
   • Within-subjects (repeated measures): same participants experience all conditions
4. Get participants and assign
   • Random sampling helps generalize to population
   • Random assignment helps equalize groups (reduces confounds)
5. Control confounds
   • Hold constant instructions, time of day, environment
   • Use counterbalancing for order effects (within-subjects)
6. Reduce bias
   • Single-blind: participants don’t know condition
   • Double-blind: participants and researchers interacting with them don’t know condition
7. Run procedure + collect DV
8. Analyze + conclude
   • Look for differences between conditions; report whether data support hypothesis
   • If applicable, interpret $p$-value / significance (conceptually)

C. Reading correlational results (quick)

1. Identify the two measured variables.
2. Read $r$:
   • sign (+/−) = direction
   • magnitude = strength
3. State prediction only (never “causes”).
4. Mention the two big correlation problems:
   • directionality problem
   • third-variable problem

Key Formulas, Rules & Facts

Research methods: what they do and don’t prove

Sampling and assignment (high-yield distinctions)

Variables and control

• Independent variable (IV): factor the researcher manipulates.
• Dependent variable (DV): outcome the researcher measures.
• Confounding variable: any factor that varies with the IV and could explain DV changes.
• Control group: baseline comparison group (often placebo).
• Experimental group: receives treatment/IV level.
• Constants: factors kept the same across conditions.

Operational definitions (make vague variables measurable)

• Operational definition: exact procedures used to define/manipulate a variable.
  • Example: “Stress” = score on a perceived-stress questionnaire.
  • Example: “Aggression” = number of times a participant blasts noise at an opponent.

Correlation essentials

• Correlation coefficient $r$ ranges from $-1.00$ to $+1.00$.
  • Sign = direction (positive/negative)
  • Absolute value $|r|$ = strength
• $r = 0$ means no linear relationship (a curved relationship could still exist).

Descriptive statistics (AP-level)

Inferential statistics (what AP expects conceptually)

• Statistical significance: result is unlikely due to chance alone.
  • Often reported as $p < 0.05$ (not a guarantee; just a convention).
• Null hypothesis: assumes “no real effect” or “no difference.”
• Confidence: stronger evidence when effect is large, variability is small, and sample size is larger.

Validity and reliability (often tested)

Key experiment features to reduce bias

• Placebo effect: improvement due to expectations, not treatment.
• Single-blind: participants unaware of condition.
• Double-blind: participants + researchers interacting with them unaware.
• Demand characteristics: cues that tell participants what behavior is expected.
• Experimenter bias: researcher expectations influence results.

Ethics (must-know AP principles)

• Informed consent (unless justified exception)
• Protect from harm / minimal risk
• Confidentiality
• Right to withdraw
• Debriefing (especially if deception used)
• Justified deception only when necessary and followed by debriefing

Examples & Applications

Example 1: Identify method + conclusion

A psychologist watches children on a playground and records how often they share toys.

• Method: naturalistic observation
• What you can conclude: description of sharing behavior in that setting
• What you cannot conclude: whether parenting style causes sharing (no manipulation/control)

Example 2: Correlation interpretation (what to say)

Study finds $r = -0.62$ between hours of sleep and number of days absent from school.

• Direction: negative (more sleep ↔ fewer absences)
• Strength: moderately strong (because $|r|$ is relatively high)
• Correct conclusion: sleep predicts absences (or vice versa)
• Add the caveat: could be third variable (e.g., health) or directionality (chronic illness reduces sleep and increases absences)

Example 3: Turn a question into an experiment

Question: “Does caffeine improve memory?”

• IV: caffeine dose (e.g., $0\,\text{mg}$ vs $200\,\text{mg}$)
• DV: memory score (e.g., number of words recalled)
• Control: same study time, same word list difficulty, same environment
• Random assignment: participants randomly placed into caffeine vs placebo
• Blinding: ideally double-blind to reduce placebo + experimenter bias
• Conclusion allowed: if groups differ reliably, caffeine may cause the memory change

Example 4: Quasi-experiment (common AP trap)

Researcher compares anxiety scores between teenagers who use social media $> 4\,\text{hours/day}$ and those who use it $< 1\,\text{hour/day}$, but teens choose their own usage.

• This is not a true experiment (no random assignment).
• Best label: quasi-experiment or correlational (depending on framing).
• Correct conclusion: association/differences, not causation.

Common Mistakes & Traps

1. Confusing random sampling with random assignment

   • Wrong: “They randomly assigned participants, so it generalizes to everyone.”
   • Why wrong: random assignment helps internal validity, not representativeness.
   • Fix: sampling → generalize; assignment → causation.

2. Claiming correlation proves causation

   • Wrong: “Because $r$ is strong, X causes Y.”
   • Why wrong: directionality + third-variable problems always exist.
   • Fix: say “predicts,” “is associated with,” “relates to.”

3. Mislabeling IV and DV

   • Wrong: calling the outcome the IV.
   • Why wrong: IV is what you manipulate; DV is what you measure.
   • Fix: ask “What did the researcher change?” (IV) and “What did they record?” (DV).

4. Forgetting operational definitions

   • Wrong: “They measured intelligence.”
   • Why wrong: AP expects how (IQ test score? GPA? puzzle performance?).
   • Fix: always restate variables in measurable terms.

5. Ignoring confounds in experiments

   • Wrong: assuming random assignment magically eliminates all confounds.
   • Why wrong: confounds can still occur via procedure differences, attrition, demand characteristics.
   • Fix: look for any variable that differs between groups besides the IV.

6. Thinking $r = 0$ means “no relationship at all”

   • Wrong: “No connection.”
   • Why wrong: it means no linear relationship; could be curvilinear.
   • Fix: phrase it as “no linear association found.”

7. Mixing up reliability and validity

   • Wrong: “It’s reliable, so it’s valid.”
   • Why wrong: a measure can be consistent but consistently wrong.
   • Fix: reliability = consistent; validity = accurate.

8. Overstating what surveys show

   • Wrong: “The survey proves teens are more depressed.”
   • Why wrong: surveys are self-report and often biased by wording and social desirability.
   • Fix: say “reports suggest…” and note possible biases.

Memory Aids & Quick Tricks

Quick Review Checklist

• You can name: case study, naturalistic observation, survey, correlational study, experiment, quasi-experiment.
• You can state what each method can conclude (describe, predict, or cause).
• You can identify population vs sample and explain sampling bias.
• You can distinguish random sampling (generalize) from random assignment (causation).
• You can label IV, DV, and at least one plausible confounding variable.
• You can write an operational definition for vague variables.
• You can interpret $r$ (direction + strength) and say why it’s not causation.
• You can explain placebo effect, demand characteristics, experimenter bias, and why double-blind helps.
• You can define reliability, validity, internal validity, external validity.
• You can list core ethics: informed consent, protect from harm, confidentiality, right to withdraw, debriefing.

You’re closest to full points when you automatically ask: “What method is this, what can it prove, and what’s the biggest flaw?”