The placebo control arm unfortunately doesn't control for population selection bias, scan dyes, or participation stress. It can only control for the placebo effect. The factors you cite do introduce skew in the data, like the placebo effect does. But you control them in other ways. For selection bias, you randomize selection as best as possible. For scan dyes, incidental pill/capsule excipients, etc, you make sure that all patients are subjected to them equally (eg, everyone gets the same number of planned scans, the different dosages get the same amount of non-drug ingredients). Not much you can do to control for participation stress, except to survey it in a questionnaire, and it is so subjective that I doubt it is ever done. So while none of these are controlled with a placebo arm, they are like the placebo effect in that they are sources of skew (the quantified concept of bias).
News 
Market Data 
Discover 
