Twitter | Search | |
Tarah M. Wheeler
Switch nothing but name & ethnicity in LinkedIn profile & recruiter response rate goes from 40% to 2% with n=30. As a scientist, I believe this should be expanded into a peer-reviewed study w anonymized data being provided publicly on a per-company basis beginning with the F100.
Reply Retweet Like More
Rachel Tobac Aug 2
Replying to @tarah
This is a fantastic study. Would love to see this replicated at scale.
Reply Retweet Like
turbogrrl Aug 2
Replying to @tarah
it appears gender *and* ethnicity were changed. i suspect just changing ethnicity would show a much smaller delta in response rate.
Reply Retweet Like
SpeedRussr Aug 2
Replying to @tarah
I could support this kind of research. So much of the decision process could be based on unconscious bias. Anonymize the data, and see what the response rate looks like. ALSO: I wonder if there is a notable difference in communication style that could impact decisions?
Reply Retweet Like
Tarah M. Wheeler Aug 2
Replying to @turbogrrl
That’s why this is interesting and should be replicated with multiple variables. It’s the study that can tell just what it’s really like in a job hunt w characteristics only visible from a picture. More data is good, and I’d like to think companies would want to know this.
Reply Retweet Like
Jake Williams Aug 2
Only has that data, but I strongly suspect that the former leads to the latter.
Reply Retweet Like
Tarah M. Wheeler Aug 2
Replying to @v3rtig0
This is why this would be so powerful. The communication style wouldn’t change at all; only the picture and name.
Reply Retweet Like
zxq9 Aug 2
Yes. And heuristics are strong. And statistically accurate, despite being individually wrong on an arbitrary scale in outlying cases. But OF COURSE we cover for lack of data with algorithmic assumptions. Or are you anti-success?
Reply Retweet Like
Baka Karasu 烏 Aug 2
Replying to @tarah
*sigh* Just serves to further validate my default position of always lying about PII / demographic data in any profile anywhere.
Reply Retweet Like
Emanuil Tolev Aug 2
Replying to @tarah
Wonder what it'd take to actually do. I know of a few recruiters but they're more boutique, probably don't go contacting people on LinkedIn all that much. Maybe some reasonable recruitment agencies used by F100. Which discipline would it fall into on the science side, psychology?
Reply Retweet Like
SpeedRussr Aug 2
Replying to @tarah
I was wondering if there could be a bias under the surface based on communication styles that may be comfortable for the reviewer. Does that make sense? Just thinking aloud, here. :-) See you soon, I hope.
Reply Retweet Like
Just another IT dude Aug 2
Replying to @tarah
Friends of mine have done smaller scale tests by hiding their military service as well. So many places make judgements based on some form of stereotype.
Reply Retweet Like
Rainbow Admin 🌈🖥☁️ Aug 2
Replying to @tarah
Agreed. Would like to see more data across roles per company. Publicizing this information is a great start for change. Wonder what the response rate is on gender alone.
Reply Retweet Like
Fallenour Aug 2
Replying to @tarah
I would point out on this the average recruiter would recognize the similarity in information, and would respond to the drive/social trap. This creates a damned if you do, damned if you dont requirement for them, and is a bad example of testing for standards.
Reply Retweet Like
Fallenour Aug 2
Replying to @tarah
Also, additional items to identify is quality of outreach, which wasnt shared or verified. Example, if I say "hey" vs a substantiated introductory. This wasnt verified as equal in terms of method or verbage.
Reply Retweet Like
🚮 Aug 2
Replying to @tarah
I changed my name by a single letter and I got exponentially more responses.
Reply Retweet Like
SpeedRussr Aug 2
Replying to @tarah
I should be clear.... and "additional" bias. Not a replacement. I just realized how that reads.
Reply Retweet Like
ILoveBitcoin Aug 2
Replying to @tarah
I actually know most of my contacts IRL, so the sex change might cause some distress. I’m game!
Reply Retweet Like