20 carefully crafted questions · Multiple choice · Instant feedback with full explanation
Experiment → Treatment applied
Survey → Questions asked
Observational → Just watch
A study has bias if it systematically produces results that don't reflect the true population.
SYSTEMATIC ≠ RANDOM ERROREvery member has an equal chance. Use a random number generator or lottery.
EQUAL CHANCE = SRSSplit into groups by shared traits, then randomly pick from each group.
GROUPS → RANDOM EACHSplit into clusters; randomly pick entire clusters — not individuals within.
WHOLE CLUSTER INBoth have the greatest chance of biased results. Marked with ☆ on your notes.
VOLUNTEER = BIASEDExperimental = receives treatment.
Control = no treatment, used for comparison.
Pick 1 random start, then select every kth member. e.g. every 5th student ID.
START RANDOM → EVERY KthReview the explanations above to strengthen your understanding.