4 ms·

I agree with you in principle, but it's difficult to draw the line when the hiring process needs to be practical. We can't interview everyone and perfect recall

I agree with you in principle, but it's difficult to draw the line when the hiring process needs to be practical. We can't interview everyone and perfect recall isn't necessary. My experience from dozens of interviews has been that many certifications are a rather strong negative signal, at least for our standard software/data engineering roles. I'm sure there are many exceptions to this, but I'll happily miss out on 1-2 good candidates if that means I can filter out 50 bad ones using the same signal. Of course it's never a binary decision and the complete picture of the resume matters. For example, if someone got a bunch of certs 20 years ago when they started their career that doesn't have any negative impact. But more often, people try to use certs and MOOCs to pad their resume and hope use them to hide the fact that they lack real work experience.

ElemenoPicuares 4y ago
I hear you and have been in your shoes. Hiring is hard, and the steps to hire people ethically are pretty nebulous, which makes it much harder. Especially when the outcome would be the same in many cases, it does seem awfully impractical. But that's the problem with bias in general, though— It's always a matter of practicality. It's been some time since most discrimination deliberately stymied people with a background or makeup the dominant culture considered objectionable. These days it hides in the subconscious mechanisms governing the purportedly rational generalizations we rely on when lacking time and resources to properly investigate something. I don't think it's anybody's fault— it's a bug in our brains and we can't control that. I do think we're ethically obligated to push back when practical requirements compel us to make those generalizations. If we expect police to do it when someone may or may not be pulling a gun out of their pocket, we should certainly expect ourselves to do it when we've got a queue of resumes to process.
- Ruphin 4y ago
  I think it's an interesting point to make, and it's important to understand that this type of selection is a form of bias. But I think the comparison you are making here with law enforcement using bias towards people of color is not exactly right. There is a difference in signaling by choice (what to put on your resume) vs signaling by nature (skin color). I think we agree that there are some things we cannot change by nature, and a selection bias based on these properties is (in every case?) ethically unjust. I do not agree that it's ethically unjust to have a selection bias on things that are behavioral. If someone is coming into an interview naked I'm going to reject that candidate. Am I being biased towards unclothed people if I do that? Should I continue the interview and try to find the candidate's fit for the position and disregard that they are choosing not to wear clothes? I think it's no different for a candidate that chooses to list certain things on their resume. Nobody is forced to list all of their certificates on their resume. It is a choice the candidate makes, and what else do I have to judge a candidates fitness than the choices they make in how they represent themselves and their skills? You can certainly make an argument that listing many certificates is a shallow signal and you should invest more time into getting to know a candidate better before you choose to reject. If someone chooses to use this signal you can make an argument that they will possibly miss a few good candidates. But I don't think you can make an argument that it's ethically unjust to do so.
  - ElemenoPicuares 4y ago
    The whole point is that disadvantaged folks are more likely to use certificates to get a foot in the door and economic hardship in the US is not evenly represented among skin colors— not even close. Folks who don’t natively have the right accent, word usage, cultural reference points and demeanor to seem “credible” and be a “good fit” for internships or jr positions don’t have a lot of options to obtain that credibility. Things like certificates and associates degrees from for-profit colleges might be their only feasible option regardless of how smart or talented they are. Dismissing that entire group out of hand because it saves time is ignoring the economic and cultural realities of who you’re dismissing by doing so. You can separate the cultural implications of nearly any subconscious bias if your analysis is shallow enough.
michaelcampbell 4y ago
This seems circuitous. You pre-suppose that the 50 ones you reject are "bad" because of this signal, then support the use of this signal because you rejected them. And then go further to assume the ratio of good:bad of people with this signal is 1-2:50. I mean, you do you, but it seems that if you were actually interested in the truth you'd have a blind-interview strategy where you interview and judge based solely on the actual interview not knowing about their certificate status a priori. Note: I don't have a lot of certs (I don't even know that the couple I do have are even up to date, nor on my resume), but this kind of response I see quite often and it feels quite smug. I find the process of going through the effort to get a cert quite valuable, as it exposes me to aspects of <subject> that I probably would have not hit just "spending my free time on work-type projects", which is another red flag for me as an interviewee. I've been a dev my whole career spanning 30+ years, so I get the desire for "fire" and "passion" for the craft, but as soon as someone asks about what I do OUTSIDE of work that could be brought to bear for work stuff, I'm out. I can see what kind of company they are representing.
- leaflets2 4y ago
  She wrote: > My experience from dozens of interviews That's not pre-supposing, instead, sounds to me like learning from one's job? > blind-interview strategy I liked that -- how does one make it scale though?
  - ElemenoPicuares 4y ago
    All overgeneralizations start out as conclusions based on observation. Having certificates could not possibly decrease someone’s ability to do a job. Assuming that everybody with certificates is a certain type of person and those types of people make bad workers is clearly not based on rational analysis. Might it generally indicate a some kinds of shortcomings? Maybe. If that merely leads one to investigate, that’s one thing. If that turns into a check in the con column for that candidate, that’s the very definition of bias.
    - leaflets2 4y ago

ElemenoPicuares 4y ago

I hear you and have been in your shoes. Hiring is hard, and the steps to hire people ethically are pretty nebulous, which makes it much harder. Especially when the outcome would be the same in many cases, it does seem awfully impractical. But that's the problem with bias in general, though— It's always a matter of practicality. It's been some time since most discrimination deliberately stymied people with a background or makeup the dominant culture considered objectionable. These days it hides in the subconscious mechanisms governing the purportedly rational generalizations we rely on when lacking time and resources to properly investigate something. I don't think it's anybody's fault— it's a bug in our brains and we can't control that. I do think we're ethically obligated to push back when practical requirements compel us to make those generalizations. If we expect police to do it when someone may or may not be pulling a gun out of their pocket, we should certainly expect ourselves to do it when we've got a queue of resumes to process.

Ruphin 4y ago
I think it's an interesting point to make, and it's important to understand that this type of selection is a form of bias. But I think the comparison you are making here with law enforcement using bias towards people of color is not exactly right. There is a difference in signaling by choice (what to put on your resume) vs signaling by nature (skin color). I think we agree that there are some things we cannot change by nature, and a selection bias based on these properties is (in every case?) ethically unjust. I do not agree that it's ethically unjust to have a selection bias on things that are behavioral. If someone is coming into an interview naked I'm going to reject that candidate. Am I being biased towards unclothed people if I do that? Should I continue the interview and try to find the candidate's fit for the position and disregard that they are choosing not to wear clothes? I think it's no different for a candidate that chooses to list certain things on their resume. Nobody is forced to list all of their certificates on their resume. It is a choice the candidate makes, and what else do I have to judge a candidates fitness than the choices they make in how they represent themselves and their skills? You can certainly make an argument that listing many certificates is a shallow signal and you should invest more time into getting to know a candidate better before you choose to reject. If someone chooses to use this signal you can make an argument that they will possibly miss a few good candidates. But I don't think you can make an argument that it's ethically unjust to do so.
- ElemenoPicuares 4y ago
  The whole point is that disadvantaged folks are more likely to use certificates to get a foot in the door and economic hardship in the US is not evenly represented among skin colors— not even close. Folks who don’t natively have the right accent, word usage, cultural reference points and demeanor to seem “credible” and be a “good fit” for internships or jr positions don’t have a lot of options to obtain that credibility. Things like certificates and associates degrees from for-profit colleges might be their only feasible option regardless of how smart or talented they are. Dismissing that entire group out of hand because it saves time is ignoring the economic and cultural realities of who you’re dismissing by doing so. You can separate the cultural implications of nearly any subconscious bias if your analysis is shallow enough.

michaelcampbell 4y ago

This seems circuitous. You pre-suppose that the 50 ones you reject are "bad" because of this signal, then support the use of this signal because you rejected them. And then go further to assume the ratio of good:bad of people with this signal is 1-2:50. I mean, you do you, but it seems that if you were actually interested in the truth you'd have a blind-interview strategy where you interview and judge based solely on the actual interview not knowing about their certificate status a priori. Note: I don't have a lot of certs (I don't even know that the couple I do have are even up to date, nor on my resume), but this kind of response I see quite often and it feels quite smug. I find the process of going through the effort to get a cert quite valuable, as it exposes me to aspects of <subject> that I probably would have not hit just "spending my free time on work-type projects", which is another red flag for me as an interviewee. I've been a dev my whole career spanning 30+ years, so I get the desire for "fire" and "passion" for the craft, but as soon as someone asks about what I do OUTSIDE of work that could be brought to bear for work stuff, I'm out. I can see what kind of company they are representing.

leaflets2 4y ago
She wrote: > My experience from dozens of interviews That's not pre-supposing, instead, sounds to me like learning from one's job? > blind-interview strategy I liked that -- how does one make it scale though?
- ElemenoPicuares 4y ago
  All overgeneralizations start out as conclusions based on observation. Having certificates could not possibly decrease someone’s ability to do a job. Assuming that everybody with certificates is a certain type of person and those types of people make bad workers is clearly not based on rational analysis. Might it generally indicate a some kinds of shortcomings? Maybe. If that merely leads one to investigate, that’s one thing. If that turns into a check in the con column for that candidate, that’s the very definition of bias.
  - leaflets2 4y ago