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

11–20 of 369 posts

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

#11

Post by StavrosK »

vonmoltke wrote: post #7
anon2016 wrote: post #4

So it masks gender by making everyone sound "like a dude" according to the videos. That's not masking gender. Make everyone sound like a robot and the blog post title will be more accurate.

The point of the experiment was for the interviewer to think the interviewer was a particular gender to see if they as it based on that impression. Making everyone sound like a robot does not accomplish that goal, so I don't see how it makes the title more accurate.

It makes everyone sound the same. Changing only women's voices is a tell-tale sign. If you get a guy who sounds "computery", or "off", it's a girl.

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

#12

Post by lotharbot »

anon2016 wrote: post #4

So it masks gender by making everyone sound "like a dude" according to the videos. That's not masking gender. Make everyone sound like a robot and the blog post title will be more accurate.

The text of the article clarifies -- they made some women sound processed-male, some men sound processed-female, some sound processed without a gender shift, and some were unprocessed. So if you got a processed-sounding voice, you had no way to know the gender of the person you were talking to.

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

#13

Post by StavrosK »

minimaxir wrote: post #10

> Contrary to what we expected (and probably contrary to what you expected as well!), masking gender had no effect on interview performance The main issue with gender and hiring is unconscious bias on the interviewer's end , which voice modulation does make sense as a tool to avoid bias. Per the grading criteria, they are judging on skill, problem solving, and communication. I'm not sure how voice modulation would af…

The interviewer is the one grading the interviewee.

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

#14

Post by lotharbot »

StavrosK wrote: post #11
vonmoltke wrote: post #7

Earlier quoted context omitted.

The point of the experiment was for the interviewer to think the interviewer was a particular gender to see if they as it based on that impression. Making everyone sound like a robot does not accomplish that goal, so I don't see how it makes the title more accurate.

It makes everyone sound the same. Changing only women's voices is a tell-tale sign. If you get a guy who sounds "computery", or "off", it's a girl.

They didn't only change women's voices. They also changed some men to sound like women. And they made some voices sound processed without pitch changes.

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

#15

Post by benkuhn »

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 assessment whether they thought the applicant was definitely male/probably male/unsure/probably female/definitely female. Then you could use voice modulation as an instrument for perceived gender and get a better estimate of the true effect of perceived gender when controlling for actual gender.

Hope they do something like that for the next round of experiments!

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

#18

Post by gravypod »

Why not just use text? There are many speech characteristics of gender that will never be found and masked by modulation.

Simple placements of words, parts of speech referring to the self, and other things are all much better clues to gender then voice frequency and pitch. For example, when speaking to some people who I know who are either naturally very high or low pitch, or are trans-gendered, there is a lot more information that can be contextually extracted from the content of their speech.

Really, just do text-based interviews. It will allow people to revise what they are thinking and really allow them to mask their gender.

Post reply on HN