The question here is what space vs tab use is actually an proxy, which is a question you generally need to ask when you get weird but valid statistical results.

The question here is what space vs tab use is actually an proxy, which is a question you generally need to ask when you get weird but valid statistical results. If there for instance was an correlation between having lived though the bad old days with primitive build/formatting tool chains and a preference for spaces that would easily explain a difference in salary that have nothing to do with the argument for/against tabs/spaces as it relates to modern tooling.