3 ms·

Putting aside the irony of this article sitting next to the "Statistics Done Wrong" article, a cumulative average to detect a gender/race wage gap is absolutel

by zalzane 13y ago

Putting aside the irony of this article sitting next to the "Statistics Done Wrong" article, a cumulative average to detect a gender/race wage gap is absolutely insane. Here's a short list of factors that could contribute to the results the article in OP has: -Lower number of years working in the field due to taking time off for maternity leave, or going back to work after raising a child. I suspect this is probably the biggest single contributor to the pseudo-gap in the article's charts. -Concentration of male/female developers in different subfields. Ex: men may be more inclined to work in security penetration, which may pay higher than other disciplines -Male/female college degree/other pedigree rates I'm no statistician, but I think these factors are more than important enough would have to be taken into account when trying to figure out the size of the gender wage gap. Don't forget to leave a comment telling me why I'm wrong when you downvote my comment for being unproductive.

reeses 13y ago
On the surface, the points you cite are inflammatory and often used a priori to justify paying women less. It's a bit irresponsible after that slide not to put the rest of the results into the context of gender. Are women represented proportionally by age? The last table in the article ("What Matters Most") would blow the lid off of many of the arguments. Do people who "could find seven attributes of their job more meaningful than base salary" in general make less than others? Do women and men respond differently to those questions? What is the disparity between men and women who value base pay most? The only value I can extract from this survey is that a) it's BS that female compensation stagnated while male compensation rose (was there a mini-baby boom during which a lot of women took unpaid maternity leave?) and b) Dr Dobbs is getting less and less relevant, given that they consider "Company-paid smartphone/wireless email" to be a form of compensation.
dcre 13y ago
I upvoted you, but I do disagree. Overall I see your points as explanations for/additional evidence of a more general gender gap — not as reasons to think it's a statistical illusion. One point you didn't make, though it is hinted at by your first point, is that if the proportion of women in the field is increasing over time, women's average age will be lower than men's (assuming people tend to enter the field from college and not from other fields, which seems reasonable), which would distort the salary numbers in a genuinely misleading way.

by zalzane 13y ago

reeses 13y ago

On the surface, the points you cite are inflammatory and often used a priori to justify paying women less. It's a bit irresponsible after that slide not to put the rest of the results into the context of gender. Are women represented proportionally by age? The last table in the article ("What Matters Most") would blow the lid off of many of the arguments. Do people who "could find seven attributes of their job more meaningful than base salary" in general make less than others? Do women and men respond differently to those questions? What is the disparity between men and women who value base pay most? The only value I can extract from this survey is that a) it's BS that female compensation stagnated while male compensation rose (was there a mini-baby boom during which a lot of women took unpaid maternity leave?) and b) Dr Dobbs is getting less and less relevant, given that they consider "Company-paid smartphone/wireless email" to be a form of compensation.

dcre 13y ago

I upvoted you, but I do disagree. Overall I see your points as explanations for/additional evidence of a more general gender gap — not as reasons to think it's a statistical illusion. One point you didn't make, though it is hinted at by your first point, is that if the proportion of women in the field is increasing over time, women's average age will be lower than men's (assuming people tend to enter the field from college and not from other fields, which seems reasonable), which would distort the salary numbers in a genuinely misleading way.