relationship independent variable dependent variable problem experiment 's always called potential confounds right might alternative explanation effect different proposed independent variable 's called controlling confounds structure experiment rule uh alternative explanations n't results inherently confounded means 're deeply ambiguous derive clear um uh uh conclusion well problem well 's problem alternative explanations number right like 're trying see uh know like know water dissolve 's water salt dissolve water well course go somebody says n't go pour salt 're pouring spray salt plastic n't dissolve go aha salt dissolve water go mean n't mean salt ha recently covered uh plastic go okay go 're put flash freeze salt stays surface say haha n't dissolve water say n't mean salt frozen water mean room temperature water 're going pour salt pour another chemical prevents salt dissolving go aha oh mean pure water 're pure water see goes right problem brought long time ago hempel japanese published number possible counterfactuals indefinitely large n't rule right ca n't ones control well control ones strike us reasonable make good sense taken seriously pay attent