PSYCH 100B Lecture Notes - Lecture 5: Statistical Power, Confounding, Caffeine
Document Summary
Confounding variable: extraneous variables that varies systematically with the levels of. All variables are controlled for except for weight, varies systematically with level. Caffeine - higher weight; no caffeine - lower weight. Either caffeine or weight has an effect on sleep. Counterbalancing: counter effect of the variable by balancing it across levels. Only way to be sure that we are distributing evs. E. g. in the case of the rat study - carefully distribute rats into two conditions according to age. If we really care about a particular variable, counterbalancing ensures that we control for it. Within subjects study: each person participates in all of the conditions. Each person serves as their own control. There are individual differences in speed but we see that with the drug, both run the maze faster. If we conducted this as a between subjects study and assigned bob to no drug and bill to drug, it will look as though there is no effect.