EXPERIMENT
Randomly assigns treatment → can show cause & effect
OBSERVATIONAL
Watches only, NO treatment → correlation only, NOT causation
SAMPLE SURVEY
Asks questions → opinions & attitudes
BIAS
Result systematically misrepresents population
SIMPLE RANDOM
Every member equally likely — gold standard
STRATIFIED
Divide into similar groups, random sample each group
CLUSTER
Divide into groups, pick entire random groups
SYSTEMATIC
Random start → every k-th member
CONVENIENCE ☆
Easy to reach → most biased
SELF-SELECTED ☆
Volunteers only → most biased