PSYB04H3 Lecture Notes - Lecture 8: Design Of Experiments, Confounding, Random Assignment
Document Summary
Temporal order must be correct: cause comes before effect. No other variable is causing the outcome. The only way to establish causality is through an experiment a controlled environment. Research design in which a variable is manipulated and the effect on another variable is measured. Iv is manipulated: dv is measured, all other variables are controlled, control group, randomized effects or matching participants. Temporal order: iv is presented first, then dv. Differing values of the iv are (cid:272)alled (cid:862)le(cid:448)els(cid:863) or (cid:862)(cid:272)onditions(cid:863) Remains the same for all levels of the iv. A well designed experiment controls as many cv as possible. Any other variables that influence dv besides iv: confound could be an alternative explanation. Any variable whose value changes systematically across different levels of the iv. Opposite of systematic: random: affects the dv differently from observations to observation, ex. A random confound could be personality differences: ex. A systematic confound could be an alternative explanation.
