Y Combinator's 2011 startup school conference. (Robert Scoble)

Many tech companies readily admit that their workforces aren't as diverse as they could be — and that they're trying hard to fix the imbalance. To date, company demographics that skew heavily toward whites and men have received most of the attention. But as I've written before, there also exists a substantial gap in gender pay in Silicon Valley.

The wage gap largely mirrors the national average; according to the American Association of University Women, female workers in the computer science field make 77 percent what their male peers do one year out of college. By comparison, Census data suggest that women make 78 percent of what men do, generally.

But we're starting to get a better idea of specific differences between men and women's salaries at different tech companies. A recent survey of median salaries by Glassdoor suggests that at many Silicon Valley firms, men make thousands of dollars more than women every year. At Google, for instance, a male senior software engineer makes 19 percent more than his female counterpart.

A couple of other things are worth pointing out: The numbers are heavily conditional on experience and the number of reports from either sex. There are big discrepancies between men and women in terms of years worked, even within the same job. The male senior software engineer at Google has, on average, nearly a decade of experience compared to just 7.6 years for women with the same title. Experience is likely one factor that helps drive up male wages relative to women.

At the same time, however, where women demonstrate more experience than men, the wage gap doesn't substantially shift in favor of women like you might expect. For instance, female software engineers at Cisco largely have more work experience than their male counterparts — yet that isn't reflected in their salaries. A mid-level female engineer at the company with nearly a year and a half more experience makes just $162 more per year, according to Glassdoor's data.

The precious few number of reports from women in general also reflect what we already know about the industry: There just aren't that many of them. Glassdoor collected hundreds of reports from men at numerous tech firms — but at those same firms, they barely could get a few dozen from women, if that.

Update: The Wall Street Journal's Miriam Gottfried also points out that Glassdoor's figures don't account for payment in company stock, which is a big deal, too.

On Twitter, some readers are pushing back against the study, saying that the men generally have more experience than the women. Again, while the data does seem to show that, I'd like to reiterate that if experience really explained so much about the wage gap, then we might expect women who have a large experience advantage over men to be making at least as much more than men as the men who have a similar experience advantage over women. (Suppose Joe has 1.5 years more experience than Jane and makes 20 percent more than her; if the roles were reversed, then by the "experience" argument Jane should be making 20 percent more than Joe.) But in fact, what we see is that some Cisco employees with 1.4 years more experience than men make just 0.01 percent more. And it gets worse: women with 1.5 years more experience at Oracle actually make 6 percent less than their greener male peers.

Yes, experience matters. But the uneven benefits conferred by experience imply that either a) something else is at work here, and experience doesn't explain as much as some think it does; or b) compared to men, women are required to have far more experience for each dollar of additional pay — which would itself be unequal.

Check out the full chart below.

(Glassdoor)
(Glassdoor)