Suppose that a large controlled experiment tests whether caffeine improves reaction times. a very large number of randomly selected participants are randomly given identical-seeming pills with varying doses of caffeine (including none) and then given tests of reaction times under the same conditions. the experiment finds a strong negative correlation between caffeine dose and reaction time. (note that lower reaction times are better.)1. there is evidence that caffeine causes lower (better) reaction times.2. people with lower (better) reaction times generally choose to consume more caffeine.3. lower reaction times are associated with higher caffeine doses, but this study provides no evidence that caffeine causes lower reaction times.4. there is no relationship between reaction time and caffeine.