computers

Studies point out that their samples can be biased because of over representation of "computer literate" people. But I haven't seen anything more than just stating it. Why don't we check how the numbers look like here?