30 Jul '14 11:21>
Originally posted by SuzianneYes, actually, they can.
Well, my point is that statistics are one thing, but they cannot be used as a predictor.
You cannot assume (in America anyway, on pain of a lawsuit) that any particular black guy is going to be less smart than any particular white guy.
Correct. But you can assume that he is more likely to have a criminal record. Statistics predicts likelihoods, not actual outcome for specific cases.
Of course I agree with you that you should not discriminate based on those likelihoods/predictions - yet people do so quite a lot.
In some cases it is standard practice. For example people get employed for their looks or their educational attainments even when they are not directly relevant to the job in question. Someone with a university degree may be more likely to be good at a given job than someone without a degree, but that isn't always the case, yet such discrimination is standard practice.