We built voice modulation to mask gender in technical interviews • Hacker News Board

Re: We built voice modulation to mask gender in technical interviews

#311

Post by hiram112 » Thu, Jun 30, 2016, 1:11 AM UTC

etjossem wrote: post #277

hueving wrote: post #211

Earlier quoted context omitted.

The supply of women with related degrees from uni is too low. If only 20% of the candidates qualified on paper are women, then a perfectly blind hiring process would result in 20% women. Expecting something else implies some fundamental misunderstanding of math.

20% is (coincidentally) the percentage of CS degrees awarded to women. They get only 10% of entry-level software jobs. We don't get to blame supply until those numbers match.

I'd be interested in seeing this study.

That is nowhere near what I encountered in my large engineering school. CS101 was probably 33% women, but after a month, we were down to about 15%. By the 4th year, maybe 10% at best.

From what I can see based on many of the millenials I now interview, there is a huge difference in degrees based on school, much more than before. It seems like many low tier schools are pumping out degrees in any field as long as they've paid the inflated tuition. I don't consider a CS degree that didn't require a few semesters of calculus, algorithms, learning C and assembly to be valid. But it appears this is now common for these 'CS lite' degrees.

Re: We built voice modulation to mask gender in technical interviews

#312

Post by pmarreck » Thu, Jun 30, 2016, 1:20 AM UTC

theoh wrote: post #274

xienze wrote: post #147

I honestly feel like this sort of thing is a waste of time. Let's pretend someone invents the perfect "gender bias-free" system: no names, no faces, voices are flawlessly transcribed on-the-fly so you can't pick up on speech patterns. What if it works perfectly and you still end up hiring more men than women? I think exercises like this may end up giving people answers they don't want to hear. So, stop wasting time w…

So you think egalitarianism is a bad idea because your hunch is that existing prejudices are totally justified, and should not be challenged -- it's a waste of time!

I think the point is that you can't assume that in some ideal world where everyone is working harmoniously and fairly, that the ratio will definitely become 50/50.

Why don't we have similar outrage over the fact that most teachers and nurses are women? Why is STEM and IT in particular receiving all this attention? There are already more women than men attending college, the consequences of that alone will push the needle beyond where it would "normally" be (whatever that is, an incalculable value currently)

Re: We built voice modulation to mask gender in technical interviews

#313

Post by pmarreck » Thu, Jun 30, 2016, 1:24 AM UTC

acdha wrote: post #129

Jemmeh wrote: post #92

Earlier quoted context omitted.

Just to be clear, men do vocal fry and uptalk(ending sentences with upward inflection/making it sound like a question) too. Just less often. And not all women use uptalk and vocal fry. Uptalk is more common in women: http://well.blogs.nytimes.com/2013/12/23/overturning-the-myt... Vocal fry is more common in women: "An examination of creaky voice occurring in natural conversations among relatively young educated Ameri…

Vocal fry is more commonly noticed and criticized in women but I haven't gotten the impression that there's definitive evidence that women do it more than men since most of the studies have been pretty small or localized. Mark Liberman has been covering this for years over at Language Log and there appears to be a great deal of ongoing debate within the academic community: http://languagelog.ldc.upenn.edu/nll/index.p…

How can a gendered speech "fashion" (which is what it seems like to me) be criticized? I think it's hot (as a guy)... probably because it's so sex-typed

Re: We built voice modulation to mask gender in technical interviews

#314

Post by junofan » Thu, Jun 30, 2016, 1:32 AM UTC

pmarreck wrote: post #309

etjossem wrote: post #277

Earlier quoted context omitted.

20% is (coincidentally) the percentage of CS degrees awarded to women. They get only 10% of entry-level software jobs. We don't get to blame supply until those numbers match.

If you actually read the article, it seems to indicate that the problem is that women do not handle "attrition events" (such as a job rejection) very well. In fact, they handle it many times worse than men do, apparently. This alone might account for the difference in percentages there, assuming there's some hidden "adversity event" that caused the attrition. Perhaps (speculating here), men handle rejection from jobs…

Actually it may have something to do with biological differences in social stress response on top of conditioning.

I personally find this fascinating--a biological basis carries with it a possibility of pharmacological treatment for those wanting it, and perhaps the design and validation of nonpharmacological approaches for those who don't or third parties seeking to reduce bias.

IIRC, there are a couple experiments in rat models along the lines of: put a male or a female rat in the same cage as an aggressive, territorial rat. The male rats are back to normal within hours-days. The female rats take days-weeks. IIRC some sort of neuronal transcription factor or something is thought involved.

Can track down a review if anyone's interested.

Re: We built voice modulation to mask gender in technical interviews

#315

Post by pmarreck » Thu, Jun 30, 2016, 1:33 AM UTC

weisser wrote: post #174

jachee wrote: post #37

Earlier quoted context omitted.

In the example clips, to me, her vocal fry was a gender tell, even after modulation.

Vocal fry isn't gender specific. Ira Glass (This American Life) has vocal fry and they discuss the gender bias here: http://www.thisamericanlife.org/radio-archives/episode/545/i...

What if Ira Glass isn't a Kinsey Zero? https://en.wikipedia.org/wiki/Kinsey_scale

Re: We built voice modulation to mask gender in technical interviews

#316

Post by tropo » Thu, Jun 30, 2016, 2:40 AM UTC

etjossem wrote: post #301

tropo wrote: post #299

Earlier quoted context omitted.

"pushed away" is so negative. How about "pulled away"? Maybe it isn't politically correct to desire being a stay-at-home mom, but it happens rather often. Sometimes this is an unexpected turn of events, with changing life priorities. Sometimes it is the goal right from the start. I met a woman in college who literally said she was getting her MRS degree. That is, she went to college to meet a suitable husband and bec…

> I met a woman in college who literally said she was getting her MRS degree. Was she taking computer science? If we're going to talk anecdotes here, every computer science major I've ever met has done it either for the sunny career prospects or because they were genuinely interested in the subject matter - and usually both. I do not buy your premise that students are graduating from computer science in order to not…

I can't go back and ask that woman, but my wife nearly did the same and was in fact in computer science.

My wife: was capable in computer science, got through 3 years of a degree before kids interfered, was mildly interested in career prospects but very relieved to realize that I could fully support her, was mildly interested in the subject matter

Had my wife taken another year to become pregnant, she'd have graduated. That would put her in your statistic as a woman who graduated but didn't end up in the field. As it is, she's in the "women who drop out of computer science" group, usually assumed to have been pushed out somehow.

Actually graduating and having a career was kind of a "Plan B" thing, although my wife wasn't quite as deliberate as that other woman.

I met another woman who dropped the career about a year after graduating in computer science from a near-top school. She went on leave to have a baby, then simply stayed home. She didn't need the money, so why be away from the adorable little baby?

These women don't have the values that you evidently have. They aren't being pushed out. They are being pulled away by things that they value more. They don't have to comply with your value system.

Re: We built voice modulation to mask gender in technical interviews

#317

Post by TheSpiceIsLife » Thu, Jun 30, 2016, 3:00 AM UTC

tropo wrote: post #295

dragonwriter wrote: post #286

Earlier quoted context omitted.

> Why do you find that 'argument' hard to buy? I find it insufficient to answer the question of whether sexism is the problem. Computer programming used to have a lot more women -- why did that change? If it was changing interest, what is the source of the gender imbalance in the change in interest? Certainly, the perception that there was gender bias in the field arose while the rate at which women were participatin…

That's backwards. We got rid of the gender bias. Very long ago, computer programming was seen as being like typing and/or secretarial work. It was seen as being unmanly. Compare with nursing and kindergarten teaching. A woman might get assigned to invent COBOL instead of being in charge of an aircraft carrier. Well, we got rid of the sexism. Happy now? Hmmm, guess not.

Your comment amused me a bit, so I wanted to write something in response. I work in metal fabrication, our workshop specialises in structural steel, so we are part of the general construction industry. We call ourselves 'boilermakers', though I've never worked on a boiler.

Anyway, in the 17 years I've been doing this on and off, I have met precisely 1 female boilermaker. And nobody cares.

What I don't follow is why is gender imbalance in other industries seen as a problem but we never see any serious attempt to get more women in to the construction industry.

Why does anyone care there are differences in numbers of men and women in various industries? Things change over time, fashions change. I think approaching these issues from the "we must correct this imbalance" is incorrect reasoning.

And, lastly, if we do manage to level the playing field, what will gender identity become?

A quote form the film Trainspotting: One thousand years from now, there will be no guys and no girls, just wankers.

I'm willing to posit that strong gender roll identity is crucial for a healthy and resilient culture.

Re: We built voice modulation to mask gender in technical interviews

#318

Post by smadge » Thu, Jun 30, 2016, 3:03 AM UTC

omginternets wrote: post #212

smadge wrote: post #210

Earlier quoted context omitted.

Why are you answering nature vs nurture with (I paraphrase) "100% nature and anyone who says otherwise is an irrational charlatan"? This debate is cliched at this point. From the study you linked to: "As suggested by previous literature and the present study, interests stabilize at an early age. Given the essential role of family and other environmental influences in the socialization process of children and their in…

>100% nature and anyone who says otherwise is an irrational charlatan What a strange comment (you don't paraphrase; you caricature)! I'm not saying any of this! Where did you get the impression that I was? I agree with all your points, btw. The existence of these differences doesn't enlighten us as to their causes, proximal or distal. Rather, we have much evidence showing that anatomy and physiology influence cogniti…

I'm sorry I caricatured your position.

I think we agree that the assumption that "removing interviewer/managerial bias will result in a gender ratio of 50%" is false. The variability of sex differences in vocational interests probably accounts for more of the gap than bias in the hiring process, i.e. "mean old white men". I can also accept that physiology influences cognition, although where we disagree is how much of a factor that is in the gender gap, or how relevant it is to any sort of policy decision.

Removing bias in the hiring process wont solve the gender gap entirely, but it will help it.

Telling girls that they can do and be whatever they want wont solve the gender gap entirely, but it will help it.

Changing the masculine and sexist culture in STEM wont completely solve the issue either.

There is basically nothing that can remove the hypothetical physiological factors contributing to gender disparity in STEM fields. So why are we even talking about it? It derails the conversation. There are things that can be done, and we should do them. Until we reach the limit of what is possible according to your hypothetical physiological explanation, why bother even bringing it up?

Re: We built voice modulation to mask gender in technical interviews

#319

Post by nradov » Thu, Jun 30, 2016, 3:32 AM UTC

hueving wrote: post #211

Tyr42 wrote: post #198

Earlier quoted context omitted.

There is evidence to the contrary, where they have run this experiment in the real world. This is for Orchestras. As late as 1970, the top five orchestras in the U.S. had fewer than 5% women. It wasn’t until 1980 that any of these top orchestras had 10% female musicians. But by 1997, they were up to 25% and today some of them are well into the 30s. In the 1970s and 1980s, orchestras began using blind auditions. Candi…

The supply of women with related degrees from uni is too low. If only 20% of the candidates qualified on paper are women, then a perfectly blind hiring process would result in 20% women. Expecting something else implies some fundamental misunderstanding of math.

At Harvey Mudd College 54% of CS graduates were women. https://www.hmc.edu/about-hmc/2016/05/17/class-2016-mileston...

Re: We built voice modulation to mask gender in technical interviews

#320

Post by tomjen3 » Thu, Jun 30, 2016, 3:39 AM UTC

xienze wrote: post #147

I honestly feel like this sort of thing is a waste of time. Let's pretend someone invents the perfect "gender bias-free" system: no names, no faces, voices are flawlessly transcribed on-the-fly so you can't pick up on speech patterns. What if it works perfectly and you still end up hiring more men than women? I think exercises like this may end up giving people answers they don't want to hear. So, stop wasting time w…

If we assume it is important to hire great programmers, then I would want such a system to signal to everybody who don't fit the standard nerd type that they won't waste their time if they apply.