Given their experimental setup (looking at the effect of randmly assigned ADAs who more or less likely to prosecute), it seems reasonable to assume that any suc

Given their experimental setup (looking at the effect of randmly assigned ADAs who more or less likely to prosecute), it seems reasonable to assume that any such hidden variables would be evenly distributed among the randomly assigned ADAs.