Skip to content
AP Psychology · Development and Learning

Research design

Research design appears in 11 reviewed AP Psychology questions and sits in Development and Learning. Below are worked examples showing exactly how the exam tests it, with the answer marked and every distractor explained.

Research design is the plan for how a study collects data, and it decides what conclusions you can draw. Experiments with random assignment can show cause, while correlational, longitudinal, cross-sectional and quasi-experimental designs show relationships.

Why it matters

The exam wants you to match each claim to the design that can support it. A design that only shows a link cannot prove cause, because a third variable might explain the result.

How the exam tests it

Some items ask which design best tests a causal claim, and the answer is a randomized experiment. Others describe a cross-sectional study where 70-year-olds score lower on intelligence tests than 30-year-olds. The groups grew up with different schooling and health care, so cohort effects may exaggerate the decline, and longitudinal studies that retest the same people usually show a smaller drop.

Don’t confuse it with

Students confuse longitudinal and cross-sectional designs, but a longitudinal study follows the same people over years, while a cross-sectional study compares different age groups at one time and can be skewed by cohort effects.

Research design in real questions

Example 1Medium
Which of the following experimental designs would provide the strongest evidence that violent video games cause increased aggression in adolescents?
  1. AA cross-sectional study comparing aggression levels in heavy vs. light video game users
  2. BA meta-analysis correlating video game violence ratings with national crime statistics across countries
  3. CA randomized experiment where adolescents are assigned to play violent or nonviolent games and aggression is measured immediately afterCorrect
  4. DA longitudinal survey tracking video game use and self-reported aggression over five years

Why C is right

Random assignment to conditions (violent vs. nonviolent game) allows causal inferences because it controls for pre-existing differences between groups. Post-play aggression measurement captures immediate causal effects. Longitudinal surveys, cross-sectional studies, and meta-analyses of correlational data cannot establish causation due to confounding variables.
Why the other 3 choices miss
A
Incorrect because a cross-sectional study is correlational and cannot establish temporal precedence or rule out third-variable explanations for why heavy game users may be more aggressive.
B
Incorrect because correlating national-level statistics is an ecological correlation that cannot establish individual-level causal relationships and is subject to numerous confounds including poverty, inequality, and culture.
D
Incorrect because a longitudinal survey, while stronger than cross-sectional designs, cannot establish causation because participants self-select their media diet; reverse causation (aggressive people seeking violent games) cannot be ruled out.
Example 2Medium
A researcher compares two populations: one with both a family history of bipolar disorder and a recent major stressor (job loss), and one with a family history but no recent stressor. The researcher finds significantly higher rates of first bipolar episodes in the first group. Which theoretical model does this finding BEST support?
  1. AThe diathesis-stress model, because genetic vulnerability interacts with environmental stress to trigger the disorderCorrect
  2. BThe psychoanalytic model, because job loss represents an ego threat
  3. CThe learned helplessness model, because uncontrollable negative outcomes like job loss produce the hopelessness that triggers mood episodes
  4. DThe strictly biological model, because family history predicts the disorder

Why A is right

The finding that individuals with genetic vulnerability (family history) AND environmental stress (job loss) have significantly higher episode rates than those with vulnerability alone directly supports the diathesis-stress model. The diathesis (family history as genetic risk) alone is not sufficient; the stress (job loss) acts as a trigger. Individuals with both factors exceed the threshold for disorder onset. This interactive gene-environment pattern is central to the diathesis-stress framework.
Why the other 3 choices miss
B
Incorrect because the psychoanalytic model focuses on unconscious ego defenses, not on measurable gene-environment interactions; this finding does not specifically test psychoanalytic theory.
C
Incorrect because learned helplessness applies specifically to depression and attributional style about uncontrollability; it does not account for genetic vulnerability or the family history aspect of this finding.
D
Incorrect because if the biological model alone explained the finding, both groups with family history should show similar rates; the differential rates based on stressor presence imply that biology alone is insufficient.

Studied alongside

Other concepts from Development and Learning.