You can’t fix diversity in tech without fixing the technical interview • Hacker News Board

Re: You can’t fix diversity in tech without fixing the technical interview

#491

Post by geebee » Fri, Nov 04, 2016, 4:46 PM UTC

Ar-Curunir wrote: post #490

geebee wrote: post #469

Earlier quoted context omitted.

I had to write my own tree traversal once. If I didn't know about trees, I wouldn't have had the basic context to do this. But I did have to spend a few hours looking up and refreshing my knowledge of tree traversal. If I'd taken a technical interview exam on tree traversal prior to that four hours or refreshing my memory, I almost certainly would have failed. And yet I within the day I was able to write the code.

That's at least partially because you knew where to look; you'd already learnt the material at some point, so you knew what to be looking for.

Agreed - my point is that I would be rejected from a job in the morning for an inability to write code that I'd be able to write by that afternoon.

They aren't testing whether you can look it up and do it, they're testing whether you have it all loaded into short term memory, on the spot.

Like a lot of people, I'm tired of having to reload it all and essentially retake my data structures exam. A lot of us just don't want to interview anymore. I know that if I need to, I can write a BFS or find all permutations of a set. If the opportunity is good enough, sure, I'll study up and get ready to do this at a whiteboard, but it's boring and unpleasant at this point, and I might not get or want the job, so at this point, I rarely bother.

Tech interviews are a big part of why tech companies are experiencing a "shortage" of applicants. They're hardly the only reason, but I'm pretty convinced they are a reason.

Re: You can’t fix diversity in tech without fixing the technical interview

#492

Post by wpietri » Fri, Nov 04, 2016, 5:18 PM UTC

moduspol wrote: post #484

wpietri wrote: post #483

Earlier quoted context omitted.

> Your comment I responded to asked for a contrary assertion. No. No it did not. I asked oldmanjay if, having falsely criticized me for asserting things with evidence, he wanted to make a contrary assertion with evidence. This was a rhetorical device to draw an interlocutor out. I neither asked for nor wanted an Internet random to offer half-baked, unevidenced pro-racist, pro-sexist waffle, thanks. > I really think i…

> I neither asked for nor wanted an Internet random to offer half-baked, unevidenced pro-racist, pro-sexist waffle, thanks. There's nothing pro-racist or pro-sexist about what I said, but that's the world we live in. Making an argument for hiring the best person for the job is called racist and sexist. This is the cause for anonymity--people like you demonize those you disagree with. > Yes, because I never before wou…

> Your [sic] frustrated response at someone who would have the audacity to challenge your view reinforce the claim that it doesn't happen enough.

The problem here is that you're not challenging my views. You're just repeating a generic pro status quo argument, which comes up every time anybody suggests that there might be some better approach than the (historically racist, sexist) status quo. I have seen it a million times.

I'm not frustrated by the argument. I'm frustrated by the zillion buttinskis who aren't willing to take an actual stand, who try to claim they are being apolitical, but somehow only choose to be "apolitical" when jumping in to object to antiracist or antisexist changes to the status quo.

I'm frustrated by the cowardice. It's not clear that you believe in anything but yourself. You'll argue to protect a status quo that benefits you. But you won't dare do it under your own name, because that might not benefit you. You don't want to limit your "hiring options", while you work to retain the limit on hiring options for women and black people.

> the people like you have been successful in demonizing anyone who disagrees with you

I am not interested in demonizing people. But I will happily repudiate the notion that those who aren't white dudes aren't really people, aren't deserving of the same respect, consideration, and privileges that accrue to white men.

Re: You can’t fix diversity in tech without fixing the technical interview

#493

Post by moduspol » Fri, Nov 04, 2016, 6:46 PM UTC

wpietri wrote: post #492

moduspol wrote: post #484

Earlier quoted context omitted.

> I neither asked for nor wanted an Internet random to offer half-baked, unevidenced pro-racist, pro-sexist waffle, thanks. There's nothing pro-racist or pro-sexist about what I said, but that's the world we live in. Making an argument for hiring the best person for the job is called racist and sexist. This is the cause for anonymity--people like you demonize those you disagree with. > Yes, because I never before wou…

> Your [sic] frustrated response at someone who would have the audacity to challenge your view reinforce the claim that it doesn't happen enough. The problem here is that you're not challenging my views. You're just repeating a generic pro status quo argument, which comes up every time anybody suggests that there might be some better approach than the (historically racist, sexist) status quo. I have seen it a million…

There's no cowardice involved, and the sooner you realize that, the sooner you'll be on a path to a rational discussion. The argument's validity isn't based on who says it, and continuing to harp on this point really shows the hoops you're willing to jump through to avoid making a rational point.

> You don't want to limit your "hiring options", while you work to retain the limit on hiring options for women and black people.

I'd hire women and black people for every single position if they're the best for the job. It's amazing how somehow Asian males have a significant presence and we're all the better for it! Somehow they pulled it off without your help, and I'm sure they're proud they didn't need to play the race card or invoke white guilt to get where they are. They're there because they were determined to be the best for their position, and they earned every bit of it.

But yeah, maybe I'm just racist because I don't doubt women or black people's ability to do the same.

If you want to talk about making it easier to account for white privilege, I'm all ears. But equating hiring people based on skillset to doing so based on genitalia or skin color is dishonest. It's condescending to even think we'd buy it. Make an honest argument for hiring someone based on gender / race and we'll take it seriously.

> But I will happily repudiate the notion that those who aren't white dudes aren't really people, aren't deserving of the same respect, consideration, and privileges that accrue to white men.

Nobody said this, and the gap between "people who think we should hire the best person for the job" and "people who think those who aren't white males aren't really people" is colossal.

Re: You can’t fix diversity in tech without fixing the technical interview

#494

Post by derefr » Fri, Nov 04, 2016, 7:56 PM UTC

kasey_junk wrote: post #200

throwaway420 wrote: post #148

I'm waiting with great anticipation for the politically correct outcry to solve the problem of not enough women being represented in the mining, construction, sanitation, logging, or deep sea fishing industries, or for there be a popular movement to get more men into teaching or nursing where they're also dramatically underrepresented there. Let's see if there's any intellectual honesty here, or maybe there's somethi…

Personally I'm happy to make an argument that given a lack of evidence that gender plays a negative roll, we have a moral imperative to eliminate unfair and discriminatory barriers to entry because its the right thing to do . But lets say that isn't a compelling argument for you. Our industry has a disastrous success rate. Something is wrong. Most software is not getting written correctly, on time, or on budget. On t…

> Most software is not getting written correctly, on time, or on budget.

Oh, sure it is. It's just that the business-people and/or clients driving the projects want "fast and crap" rather than "slow and well-made", because "fast and crap" still makes money.

Engineers (as in, people who have an engineering mindset, in any profession) will naturally fight back against this, which is where the time and budget creep comes from: theoretically, all software could be made "on time" and "on budget" and would be exactly as half-baked as the business-people were expecting it to be (and they would still be able to sell it, believe me.) But instead, engineers want to make something good, despite only having the time- and money-budget to make crap. Thus late nights and burn-out; thus unmaintainable hacks to get features "working for real"; etc.

The problem is fundamental to the collision between these mindsets. Remove either, and the problem goes away. Software can be made using "pure engineering" (see NASA) and succeed. Software can be made using "pure business concern" (see bottom-dollar outsourcing shops) and 'succeed' (at least in the sense it's measured by.) But there is no long-term healthy place on the spectrum existing between these two points.

Re: You can’t fix diversity in tech without fixing the technical interview

#495

Post by ubernostrum » Fri, Nov 04, 2016, 8:39 PM UTC

Lio wrote: post #440

gspetr wrote: post #429

Earlier quoted context omitted.

I think the issue here is that society sets a taboo on certain views, no matter if they make sense or not. In a sense we have gone back to believing that the Earth is flat when we deny uncomfortable conclusions based on society's (implicit) pressure.

I hope that's not the case here. My reluctance to give my own conclusions about paper is that I don't have the background to interpret the observations properly. I didn't want my posting of it to be taken as a statement of a poorly informed opinion one way or another, rather than me feeling pressured by society not to discuss it. I personally don't have anything objective to contribute to discussion that wouldn't be…

The problem as always is that, every time a field or industry has tried to argue "well, (insert underrepresented group here) just aren't interested in/aren't as good at what we do", no matter how many attempts they make to justify it with weird evolutionary arguments, they usually end up being proven wrong when someone finds a way to remove bias from the entry process of that field or industry.

Seriously, go read about stuff like the blind auditions for orchestras (and applications of the same technique to other fields) and how they blew up any notion of meritocratic processes and gender/racial imbalance being purely due to lack of interest or ability from the groups being excluded.

Re: You can’t fix diversity in tech without fixing the technical interview

#496

Post by belorn » Fri, Nov 04, 2016, 10:03 PM UTC

geebee wrote: post #489

belorn wrote: post #480

Earlier quoted context omitted.

Yes I can. But first I would like to say that I could too just have requested citation from the parent comment and be done. Finding papers you read a month/year/years ago take times, and links goes sometimes dead, and in a online forum like this it is generally appreciated if the requester has done some initial attempts themselves to search online for sources. Anyway, it was a polite request and I happened to save th…

"I could too just have requested citation from the parent comment and be done" Been done with what? Yes, I know, sometimes people say "citation please" as a way of refuting without reasoning, but other times, people are actually interested in reading the citation. This is one of those cases. Furthermore, this is not one of those cases would something could be easily established through a quick google search, this sou…

Thanks for showing that was the case, and Im glad that source was helpful. Chapter that begins at Page 27 is especially interesting since it directly address potential reasons, and the aspect of being in minority vs majority.

Re: You can’t fix diversity in tech without fixing the technical interview

#497

Post by gozur88 » Sat, Nov 05, 2016, 3:47 AM UTC

divbit wrote: post #13

Animats wrote: post #5

The big result there is this: "Poor performances in technical interviewing happen to most people, even people who are generally very strong. However, when we looked at our data, we discovered that after a poor performance, women are 7 times more likely to stop practicing than men." On the hiring side, "For the specific case of an online job posting, on average, 1,000 individuals will see a job post, 200 will begin th…

That is a pretty huge statistic, - what also pops out to me is: "stop practicing ". This seems to imply that these technical interviews are skills that you don't really gain on the job. I know that personally would 9 times out of 10 spend time on a cool / beneficial project rather than making time to practice trivia problems.

Getting a job and actually performing the job are two entirely different skillsets. There are a whole lot of people who bounce from one three month gig to the next as the people who hire them realize it was a mistake.

Re: You can’t fix diversity in tech without fixing the technical interview

#498

Post by gspetr » Sat, Nov 05, 2016, 12:53 PM UTC

ubernostrum wrote: post #495

Lio wrote: post #440

Earlier quoted context omitted.

I hope that's not the case here. My reluctance to give my own conclusions about paper is that I don't have the background to interpret the observations properly. I didn't want my posting of it to be taken as a statement of a poorly informed opinion one way or another, rather than me feeling pressured by society not to discuss it. I personally don't have anything objective to contribute to discussion that wouldn't be…

The problem as always is that, every time a field or industry has tried to argue "well, (insert underrepresented group here) just aren't interested in / aren't as good at what we do", no matter how many attempts they make to justify it with weird evolutionary arguments, they usually end up being proven wrong when someone finds a way to remove bias from the entry process of that field or industry. Seriously, go read a…

This is an interesting analogy and there is evidence that it works, as long as you somehow negate the gatekeeper: Biases from interviewers.

Does it still hold when there is no gatekeeper?

I'm not seeing that many minority or women as entrepreneurs. Can you hypothesize why is that?

For example if the oft cited gender pay gap is really as wide as 23%, where are the all female companies with female CEOs exploiting the market inefficiencies?

Re: You can’t fix diversity in tech without fixing the technical interview

#499

Post by clarry » Sat, Nov 05, 2016, 3:07 PM UTC

tonyarkles wrote: post #486

clarry wrote: post #363

Earlier quoted context omitted.

Do you have any online resource that could tell me what these foundations of CS are?

This, I think, is what is called "the Curse of Knowledge". I've taught 2nd year (Assembly) and 3rd year (Intro to Operating Systems Concepts) CS courses quite successfully, but I would have a helluva time teaching a 1st year course. Like I mentioned in the sibling thread, I've been programming C for twenty years, and learned BASIC on a Vic-20 four yearsish earlier than that (at age 8). I learned all of that from libr…

Right. Well, my story is that I'm selft taught (started programming when I was 6 or 7) and thus have been programming as a hobby for nearly twenty years. I decided not to go study CS because I suspected the amount of new and interesting material would be low compared to the time, money and effort it'd take to graduate. Instead, I went on to learn something completely new.

So one of my problems is that I don't actually know what I'm missing out on.

Re: You can’t fix diversity in tech without fixing the technical interview

#500

Post by eveningcoffee » Sat, Nov 05, 2016, 6:27 PM UTC

tonyarkles wrote: post #485

geebee wrote: post #467

Earlier quoted context omitted.

Interesting, but your case may really get at what may go wrong with the algorithm exam approach to interviewing. I would expect a developer with a good understanding of fundamentals to see the difference in using a hash vs an array, and I would hope that developer at one point implemented a hash as an exercise. I haven't needed to implement my own hashing table or methods on a real world project, as far as I can reme…

To address the last point, there's a tradeoff, as always. There's two competing forces with a hash function: speed and distribution. We want the hash to compute as quickly as possible, but want as close to equal probability of each bucket being chosen. Assuming you're looking for a 32-bit hash value, my lazy whiteboard hash function would probably be something like: - for strings, xor the bytes a word at a time. - fo…

for ints, just use the 32-bit value

I assume that this should be value modulo m, where m is actual size of the backing array.

for strings, xor the bytes a word at a time.

Well, for strings the same, that is, final output modulo m.