Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

We built voice modulation to mask gender in technical interviews

blog.interviewing.io

Post reply on HN
Print view View original

111–120 of 369 posts

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

#111

Post by Bartweiss »

apl wrote: post #96
Bartweiss wrote: post #91

Earlier quoted context omitted.

If anything, I would imagine it's like the old story about drafting athletes. Given two equally successful point guards, you draft the one with crappy technique because he has more room to grow. If you're hiring for some mid-level orchestra, bad mechanics on a sufficiently-skilled player might be enticing. And as you say, bad mechanics among top performers are basically nonexistent.

That'd be interesting, but once you reach an orchestra you're generally beyond technical training. Players are expected to know how to play; individual training is rare and focuses on style aspects more than technical things.

True enough; I had that in mind when I said mid-level, but even at that point I imagine individual training is largely over.

Perhaps a better example would be a teacher identifying the most promising school/college-age musicians?

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

#112

Post by Jemmeh »

kylestlb wrote: post #44
yanilkr wrote: post #35

Sweet that some one is trying new things with ideas. But Oh boy, Life does not revolve around tech interviews. Make peace with it. An old school psychologist would observe us and say that the people obsessed with tech interviews faced an emotional toll of rejection during early career and are trying to get over it. Google, Facebook and other companies are like parents of the tech kids and the kids are seeking some ki…

Maybe interviewing is over-important because of the frequency in which tech workers/engineers change jobs.

And tied into this, job changes are frequently said to be one of the best ways to get a raise.

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

#113

Post by Tarrosion »

russelluresti wrote: post #93
grb423 wrote: post #55

Is anybody studying the reasons behind the huge gender disparity in roofing, welding or kindergarten teaching? I think there is a tremendous gender bias in those and other fields that is going unstudied, because nobody cares or because those fields aren't as cool or important .

Ehhh. There's only about 360k welders in the U.S. and 3.6 million software engineers. That's an order of magnitude in difference in impact. Additionally, a software engineer probably makes more than a roofer, welder, and kindergarten teacher combined. Part of the reason for studying software engineering is because it's part of a larger trend of women not being as prevalent in high-paying careers. That's also why the…

3.6 million software engineers? That seems really high - do you have a source?

WolframAlpha reports that in May 2016 there were 144.6 million workers in the USA and 87.2 male workers. That'd mean that 2.5% of all workers are software engineers. If you figure that 75% of software engineers are male (some quick searching suggests this is a conservative number), then >3% of all employed males are software engineers. My instinct is that's high, but if you have a source I'll happily become informed.

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

#114

Post by Grae »

rcpt wrote: post #76
grb423 wrote: post #55

Is anybody studying the reasons behind the huge gender disparity in roofing, welding or kindergarten teaching? I think there is a tremendous gender bias in those and other fields that is going unstudied, because nobody cares or because those fields aren't as cool or important .

Surgery has a large gap to but I rarely hear it brought up in these discussions. http://www.ama-assn.org/ama/ama-wire/post/medical-specialtie...

That's because you aren't as steeped in medical culture as you are in tech culture and don't hear about it. It is brought up all the time.

Source: my wife just finished her medical residency.

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

#116

Post by CM30 »

benkuhn wrote: post #15

I would really have liked to see them check whether the gender-blinding actually worked. Even if someone's voice is modulated, they may have gendered behavior patterns that could influence someone's performance ratings. I wouldn't be surprised if I could guess the gender of a voice-modulated person pretty accurately from other cues. To combat this, they could have asked the interviewer during the performance assessme…

This is a good point. There are often differences in how people of different genders (and nationalities, interest groups, classes, educational levels, etc) express themselves regardless of the format, and modulating their voice doesn't really hide that.

For example, I've run quite a few forums before, and I'm pretty confident I (and quite a lot of other forum admins/moderators) could reliably judge someone's gender through their manner of writing and the (non gender related) info in their profiles.

So if you wanted to really remove possible gender 'bias' from the occasion, you'd have to go much, much further. You'd have to give people a way to communicate which reduces their responses to preset answers, that's entirely conducted through text on a screen and make sure all personal information is filtered out as well.

There's always going to be a trade off here. On the one hand, I'm sure the system used in the Luigi's Mansion Dark Moon Scarescraper would completely eliminate any potential gender bias, but that's only because it limits chat to 'over here', 'good job', 'thank you very much', 'help' and 'hey'. On the other hand, an interview in person would give you a lot of information about how well they could do the job (and let you test it), but then you've got psychological biases to deal with.

It's a tough one really.

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

#117

Post by Jemmeh »

jsprogrammer wrote: post #108
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…

The NYT blog post starts out discussing regional dialects, then presents a study of 23 Southern Californians aged 18-22... The only study the post links to is paywalled and the link to a conference paper is dead. Your second link is also paywalled.

I've found most studies to be paywalled in general aside from the free abstract. Not really sure what to do about that.

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

#118

Post by jacobr »

mwcampbell wrote: post #98
BoysenberryPi wrote: post #29

Earlier quoted context omitted.

Because your actual interview will not be text-based and if their goal is to simulate an actual interview scenario where you will be speaking out loud an expected to form coherent thoughts to questions on the fly this comes closer to doing so than text based.

And why shouldn't the actual interview be text-based? Yes, that would be a drastic change, and it would render on-site interviews pointless. But then again, it would eliminate a lot more biases, e.g.based on accent and disabilities.

Many jobs will use mostly written communication, then it doesn't matter much how you perform verbally, and a text-based interview would probably work well, or even better.

If you're expected to cooperate closely and communicate verbally with colleagues, you would want to know how the candidate handles that before hiring them.

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

#119

Post by Bartweiss »

_dark_matter_ wrote: post #110
rcpt wrote: post #76

Earlier quoted context omitted.

Surgery has a large gap to but I rarely hear it brought up in these discussions. http://www.ama-assn.org/ama/ama-wire/post/medical-specialtie...

To be fair, my wife is in medical school, and gender imbalances in specialties is brought up all the time.

Similarly, educators talk a great deal about gender imbalances in teaching (K-8 is female-dominated, physics and economics skew, biology female, and so on).

What I think is interesting is that there are a handful of fields - primarily technology and corporate leadership - where the gender imbalance is a national topic. I'm sure it's largely down to the visibility of those fields, but there are a lot of people discussing sexism in tech whose backgrounds are neither research nor technology. That seems much less true for something like radiology.

It also seems to muddy the waters a great deal. With inside-baseball discussions like medicine, I assume(?) things like roughly when the gender imbalances form are common knowledge. In tech, we seem to be constantly rehashing those matters (see: corporate diversity reports constantly stressing "our applicant pool is lower-diversity than our staff, so please stop reducing this issue to discriminatory hiring").

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

#120

Post by bduerst »

aphextron wrote: post #107
Bartweiss wrote: post #88

Earlier quoted context omitted.

That's a particularly interesting one, because it's not susceptible to the usual charges of "no one cares because it's low pay/status". If you look to radiology rather than surgery, the gender imbalance is enormous. Suddenly we're looking at gender skew in two highly technical, prestigious, lucrative fields, and seeing that they prompt totally different types and amounts of discussion.

It's almost as if men and women have some fundamental differences in psychology that affect their preferred choice of career.

> men and women have some fundamental differences in psychology

This was largely debunked from a biological perspective [1]. Any psychological perceptions you may be referring to are largely cultural and subject to the biases that this product is trying to combat.

[1] http://www.pnas.org/content/112/50/15468.full

Post reply on HN