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

Algorithmic Monocultures in Hiring

algorithmichiring.github.io

Post reply on HN
Print view View original

71–80 of 93 posts

Re: Algorithmic Monocultures in Hiring

#71

Post by Joel_Mckay »

albedoa wrote: post #69
ErroneousBosh wrote: post #51

Earlier quoted context omitted.

So basically, in order to have a steady stream of money coming in, you need to start with having a steady stream of money coming in? :-D

And often your investments will lose you money, so try to avoid the unprofitable ones, especially when starting out.

Indeed, most kids first big investment is in their schooling.

Sometimes folks have help, sometimes its all debt, and sometimes it pays off eventually.

However, living beyond ones means is almost always unwise. =3

Re: Algorithmic Monocultures in Hiring

#72

Post by sudosteph »

onemoresoop wrote: post #66
tapland wrote: post #45

Earlier quoted context omitted.

I passed a series of those and since I remembered the questions from a relatives autism diagnosis testing I asked what they do since they are effectively filtering for things like that. HR rep said those applicants should probably go see a shrink instead (!!???) and that was the end of me interviewing there. The testing needs to end. The people using these tools don't know how they work, what they are testing and wha…

Wait, they’re filtering out autistic traits or looking for them in candidates?

Filtering out is my guess.

About 20 years ago, I remember getting my hands on an answer key for the personality screener used to work at Target. This was just for a $7/hr cashier position, but it had a very low pass rate. To them, the ideal candidate for them was: always positive and optimistic, preferred being around people than being alone, never complained, frequently sought approval from peers and authorities, always followed every rule no matter what.

So it wasn't explicitly designed against people with disabilities, the rule-following aspect may be more present in autistic people - but for a lot of these, I can't see many people passing if they answered honestly.

Re: Algorithmic Monocultures in Hiring

#73

Post by BigTTYGothGF »

mrkeen wrote: post #60
BigTTYGothGF wrote: post #55

Earlier quoted context omitted.

Are you failing because of the dot test or because of the personality test?

They are smart enough not to say. The usual pipeline is HR then both Alva tests at once (which seems backward to me - why not use the screen as the actual screen, instead of wasting an hour of HR time). All I know is I'm in their pipeline after the HR but not after the Alva tests.

Let me rephrase that. In your own view, with which component of the test do you think you're having trouble getting through? How many times has this happened, and if it's more than twice have you considered some of the various test practice resources?

Re: Algorithmic Monocultures in Hiring

#74

Post by cindyllm »

PaulHoule wrote: post #63
pclmulqdq wrote: post #57

Earlier quoted context omitted.

Not if they are dressed up in a particular way, and not if intelligence is genuinely relevant. I have done a few of these tests for jobs in the US before. They are just bad.

The 2000 census gave enumerator applicants a small multiple choice test that was similar to an IQ and then they hired foremen and line workers working down from the highest scores. I've never worked with as many smart people at a temp job!

[dead]

Re: Algorithmic Monocultures in Hiring

#75

Post by lo_zamoyski »

geraneum wrote: post #47
Eridrus wrote: post #4

It seems like this is measuring algorithms against the disparate impact standard of all demographic groups needing to have the same aggregate results. Which is extra weird because the samples to this are applications, not humans, so this is subject to bias in how people apply to these positions. So if a demographic group is more likely to apply to some jobs they are not qualified for, this paper would say they are be…

> if a demographic group is more likely to apply to some jobs they are not qualified for Can you expand on this? How is a whole demographic group not qualified for jobs on wide spectrum? Is this about certain industries? Certain jobs? Certain groups?

> How is a whole demographic group not qualified for jobs on wide spectrum?

If a demographic groups has, for example, a culture that encourages overconfidence in relation to actual qualifications, then it is reasonable to expect that that demographic group will be relatively over-represented in the application pool relative to its aggregate qualifications.

Similarly, a culture that instills under-confidence relative to actual qualifications can be reasonably expected to be under-represented relative to its aggregate qualification.

Re: Algorithmic Monocultures in Hiring

#76

Post by Eridrus »

BoiledCabbage wrote: post #8
Eridrus wrote: post #4

It seems like this is measuring algorithms against the disparate impact standard of all demographic groups needing to have the same aggregate results. Which is extra weird because the samples to this are applications, not humans, so this is subject to bias in how people apply to these positions. So if a demographic group is more likely to apply to some jobs they are not qualified for, this paper would say they are be…

> So if a demographic group is more likely to apply to some jobs they are not qualified for, this paper would say they are being discriminated against. Your understanding appears to be incorrect. > Our research also found that this pattern does not appear to be the case in other circumstances. We analyzed data from the largest prior study of hiring decisions, which sent 83,000 applications to 108 Fortune 500 firms du…

> If it were what you were asserting, then this behavior and results would persist even without AI being used. Instead when they remove the filter for AI decisions (and AI mono-culture in decisions) the effect is no longer present.

> Our research also found that this pattern does not appear to be the case in other circumstances. We analyzed data from the largest prior study of hiring decisions, which sent 83,000 applications to 108 Fortune 500 firms during the same time period as our study and did not focus on whether AI was used to make decisions. We found that the rate at which applicants were rejected from every firm they applied to in this data was no higher than what you’d expect if each company decided independently of the others.

> If it were what you were asserting, then this behavior and results would persist even without AI being used. Instead when they remove the filter for AI decisions (and AI mono-culture in decisions) the effect is no longer present.

Thanks for this note, I missed this when skimming it. I would love to see their actual analysis here explained more than a single line, but this doesn't say the original study found no adverse impact at the job type level (they seem to say this wasn't analyzed), but rather that firms seemed to look more independent. Which makes sense for the headline, but is not about their notes on harms, which I still think have all the weaknesses I outlined.

Re: Algorithmic Monocultures in Hiring

#77

Post by BigTTYGothGF »

sudosteph wrote: post #72
onemoresoop wrote: post #66

Earlier quoted context omitted.

Wait, they’re filtering out autistic traits or looking for them in candidates?

Filtering out is my guess. About 20 years ago, I remember getting my hands on an answer key for the personality screener used to work at Target. This was just for a $7/hr cashier position, but it had a very low pass rate. To them, the ideal candidate for them was: always positive and optimistic, preferred being around people than being alone, never complained, frequently sought approval from peers and authorities, al…

> I can't see many people passing if they answered honestly.

You're not supposed to answer honestly, you're supposed to answer in such a way as to convince them to hire you.

Re: Algorithmic Monocultures in Hiring

#78

Post by ranger207 »

Joel_Mckay wrote: post #10
dhosek wrote: post #3

This is just one of many reasons why my current job is likely to be my last. I feel like so much of modern life is just irredeemably broken right now.

Could join those in the slow life, invest in passive revenue streams, and keep cash burn rates minimal. You will be fine avoiding debt-based indentured 1980 metropolitan cultures. Learn to play a cheap instrument, garden vegetables, paint miniatures, volunteer at pet shelters, or travel to odd destinations. Play the long game, and remember to have fun. You owe corporate nothing outside what they paid for... and not a…

I want my net worth to be dependent on my skills, not the pile of money I happen to have laying around

Re: Algorithmic Monocultures in Hiring

#79

Post by Eridrus »

geraneum wrote: post #47
Eridrus wrote: post #4

It seems like this is measuring algorithms against the disparate impact standard of all demographic groups needing to have the same aggregate results. Which is extra weird because the samples to this are applications, not humans, so this is subject to bias in how people apply to these positions. So if a demographic group is more likely to apply to some jobs they are not qualified for, this paper would say they are be…

> if a demographic group is more likely to apply to some jobs they are not qualified for Can you expand on this? How is a whole demographic group not qualified for jobs on wide spectrum? Is this about certain industries? Certain jobs? Certain groups?

This is not about being entirely unqualified.

My criticism here is that all you need to get this result is that there are some jobs that some groups apply to more than other groups despite being less qualified.

We know from other literature that men are much more likely to apply for jobs they are unqualified for, there could be other group differences, or there could be a bias of more men from a certain demographic applying for a job type than men from other demographics.

I don't know the answer, but there's a lot of ways for this sort of thing to show up when you look at data with a fine tooth comb and ask why its not perfectly even everywhere.

I'm sure we could have a whole separate argument about the disparate impact standard's validity for society (it is a matter of law ofc), but even if you accept that standard, I think you should be skeptical of the harms noted in this paper.

Re: Algorithmic Monocultures in Hiring

#80

Post by Eridrus »

sudosteph wrote: post #72
onemoresoop wrote: post #66

Earlier quoted context omitted.

Wait, they’re filtering out autistic traits or looking for them in candidates?

Filtering out is my guess. About 20 years ago, I remember getting my hands on an answer key for the personality screener used to work at Target. This was just for a $7/hr cashier position, but it had a very low pass rate. To them, the ideal candidate for them was: always positive and optimistic, preferred being around people than being alone, never complained, frequently sought approval from peers and authorities, al…

You don't see how a customer facing position in a retail chain would reasonably want all these personality traits in their hires as a matter of operating a good business?
Post reply on HN