I see now, thanks. So: For medium/high caffeine consumers: placebo -> decrease (and headache) caffeine -> 'level 1' Non/low caffeine consumers: p

I see now, thanks. So: For medium/high caffeine consumers: placebo -> decrease (and headache) caffeine -> 'level 1' Non/low caffeine consumers: placebo-> 'level 1' caffeine -> 'level 2' (where level 2 > level 1). However, I still think there might be some confusion in using a single instance for testing, esp. testing alertness. There might be a large withdrawal symptom going on in the brain that first time that makes alertness more difficult in general.