I'm glad the "rigorous testing process" was mentioned. The facade of "we only hire superninja-kickass-hyperexperts and we will verify that you are such a programmer before we hire you" attracts a certain type of person, and intimidates another. And it may be that it intimidates more women than men.
But the article seems to glide past that and presents the idea that five years of experience might change someone's willingness to apply. And it probably will help, but that amount of experience is unlikely to change someone's basic perception of their ability to succeed in a high-stakes one-chance prove-yourself situation.
"And it may be that it intimidates more women than men."
I've been looking around for a developer recently, and in looking at the profiles of women vs. men at places like AngelList I noticed a few key differences in how they described themselves. We're talking about people with very similar academic backgrounds here with similar levels of experience.
Women would say things like "I worked on a team that did X," while men -- who obviously did the same kind of work -- would simply say "I did X." Women would highlight the cooperative nature of their work, while men would give the impression that they did it all themselves. In my experience the former is typically more honest, especially if you actually were working on a team or on a team-oriented college project.
Women also tend to highlight team-oriented and non-technical skills more than men in general, which to techie-minded hacker types gives the impression that they are not strong on the technical end.
I also noted my reaction to these differences. Introspecting, I noticed that I tended to assume that "being on a team" meant riding the coattails of the other team members. If I noted this bias in myself, it's probably pretty common.
I decided to try to de-program myself of this bias by thinking of all the counter-examples I've seen.
I do believe in the "10X programmer" -- for the simple reason that there are 10X everything elses. There are 10X athletes, 10X businesspeople/salespeople, 10X artists, 10X anythings. When we look for employees or contractors we are looking for that 10X standout, or at least someone who's going to grow into one. But being really good at something is not mutually exclusive with working on a team or highlighting non-technical skills, and the inverse is also not correlated. I've met plenty of monomaniacal "closet programmers" who really aren't any good. They sit alone stewing in their huge egos and churn out crap. (But then again I've met some who are. There does not seem to be a strong correlation.)
There's probably quite a few really talented female programmers out there who are being passed over because they don't conform to the "whip it out and let me measure it" customs of technical recruitment. I also suspect that this does drive women from the field.
I worked with a 10x developer once, he was awful and I couldn't stand to work with him. Although he could produce an insane amount of code and functionality his presence was like throwing a grenade into the middle of the dev team everyday. He left to take a job at google and the overall team productivity went up significantly. He basically did 10X individually,but took 11x away from the team. So team work is pretty darn important.
I've worked with 10x programmers, and while certainly a lot of their value is simply their ability to write code, they also have a lot of value because they help other people excel. The 10x programmers I've worked with are people who push for code reviews, better tools, proper testing, realistic timelines, good communication with clients, etc.
And in the long run, those programmers made the programmers around them better. A 10x programmer creates a lot of 2x-5x programmers. I'm maybe a 3x programmer and it's because a lot of 10x programmers have taken the time to help me get better.
> How to be a 10x engineer: Incur technical debt fast enough to appear 10x as productive as the ten engineers tasked with cleaning it up.
We have a guy in our onsite team that keeps making our user facing application suck with his changes one day and then fix those every other day. So lots of git commits but limited productivity in the team, and a very tough maintainence task for every developer in the team.
I think you're misunderstanding what "10x developer" is supposed to mean. The impression I've gotten is that they are considered to be 10 times as productive because of the work they don't do. They write code that has less bugs, is more scalable, requires less refactoring, is easier for other developers to understand, etc. You'd never see it in LOC or sprint points completed, but there'd be a noticeable difference of pace between a project of 10x devs and a project with 1x devs.
I wanted to reply to all the comments so I just decided to reply to my own comment. For those telling me I'm misinterpreting what a 10x dev is, I don't think a solid definition of the 10x developer actually exists it's mostly in the eye of the beholder.
I think it's mostly in what other people think of them especially that guy that hands out raises. So it's incredibly subjective.
I say a 10x developer is someone that can create features 10x faster than everyone else, regardless of their other skills.
To be clear my dream rockstart software engineer is not a 10x developer. They are 5x developer, a 5x Team player, 5x architect and 10x Mentor. So maybe we need a dictionary of terms around what people by various nicknames and concepts given to the various kinds of software developers.
I agree everyone has their own definition, but the one you are giving is a pointy-haired definition, where the only metric is LoC and a successful demo of some feature.
Even though it's hard to measure a lot of things like technical debt, architectural benefits, bugs, edge case handling, UX polish, teamwork, etc, it doesn't mean that you can reasonably get away with discounting those effects when you talk about 10x. The 10x is referring to productivity, which is implicitly a net measure (gross productivity is just activity, and no one willingly pays for that).
So yeah, you can be 10x in terms of cranking out working code a lot faster (rare), but if that is blowing other people up then it's not meaningful to say they are more productive, in fact it's extremely harmful to risk putting them on a pedestal like this in front of clueless management. For me, true 10x is more intangible than cranking out functionality, it's more about the ability to reconcile the details with the big picture, and zeroing in on solutions that solve the most problems and create the fewest. How this actually looks in practice varies from project to project, but I think it rarely correlates with sustaining a 200wpm typing speed for 8 hours a day.
impressmyself.co , It probably won't be up until next week sometime though. I have to figure how I want to write the blog post, and then actually do it.
Those exist too. It's possible to be really good at something but to be so awful at everything else that the net effect is zero to negative. But not all "10X developers" are assholes.
That may be a men/women thing, but I actually screen for that when I'm interviewing and significantly prefer to hire people that say "we did X" rather than "I did X". It's my job as an interviewer to figure out which person on that team you were, but if you describe your work as I I I I I and never We We We We, it is pretty telling.
It also goes for managing. If my team fucked something up, I say "I fucked up Y", even if it was a report of mine. If my team did something great, it's a "We did Z" even if I did it alone.
When you put it out that plainly, it sounds hokey and stupid. But trust me, over the long run those tiny things are the difference between a good company culture and a bad company culture.
It's interesting that you say your bias is toward perceiving someone self-reporting as working on a team as "everyone on the team did the work but me".
Does this result from primarily working alone, or have you just been granted license to be more autonomous than most coders I know? A lot of shops both employ pair programming and have a series of pairs working on one project, or have a t the very least a bunch of people working on different aspects of the project.
Of course this could be spun as "I did y for my project, x" but I'm curious as to why "As part of the team working on x project, I did y" sounds less like accomplishing something technical to you.
Say I had a project out of which I'm submitting 3-4 conference papers to various conferences for publication, working with my advisor and another student. Would it be better to say I did it all, or to be honest and say the other student did 1/3, I did 2/3, and my advisor helped shape our vision?
I felt the same as 'api' when he explained that and to me, it came from my bad experiences working on teams, in high school, where 1 or 2 people did all the work and the rest rode the coattails to an easy A.
Of course my experiences in high school shouldn't factor into my judgement of reading judging peoples capabilities (because chances are everyone worked on a team that did X). Something like that might just be a hidden cultural bias - I'm sure everyone has an experience where they forced to worked with others who didn't feel like pulling their weight - and its those negative experiences that shape our notions of what it means to work on a team, rather than the multitude of positive experiences we may have had.
I have often worked alone, but not always. I've never worked in a pair programming shop, and I would probably find it distracting for anything but debugging or syncing up. The trouble is that complex programming tasks usually have the nature where communicating the ideas takes longer than simply coding them. You'd spend all your time syncing state between you and the other person.
But I don't think this explains the bias. I'm not really sure where it comes from but I found it interesting.
I read a book about the development of writing in mesopotamia (http://www.amazon.com/gp/product/0674049683 , but I can't really recommend it as a general-interest book); it mentioned that a particular king had seen his city flourish under his rule and boasted accordingly, in the records we have, of how irrigation canals were dug and roads were built and generally a lot of great stuff happened on his watch.
In a parenthetical, it also noted that, as is typical for the records of the time, they actually attributed all these works to the king personally ("I dug canals in XX regions").
> Women would say things like "I worked on a team that did X," while men -- who obviously did the same kind of work -- would simply say "I did X.
I find this a very interesting fact, but I see a number of other possible explanations for "I do X":
- It is simply shorter. Men being used to more compact communication?
- It is also more boastful - perhaps men being used to market themselves more aggressively than women? Or, put in another words, having and/or presenting bigger egos in their CVs?
- Women being insecure about their personal contribution in X? Maybe the coattail-riding assumption is not always wrong.
In any case, I would withhold judgement before an interview.
There's lots of possibilities. It's really hard to tell whether any of these things are cultural or come from actual gender dimorphism (to use the technical biological term). Ultimately it doesn't matter if they are irrelevant to ability.
What I'm saying is that when we are evaluating peoples' abilities, we should take care to evaluate actual ability rather than unreliable proxies for ability. The latter are prejudices, basically, and they blind us to a lot of talent that might not fit those lazy filters.
I'm neither a "PC liberal" nor a conservative/reactionary on this issue. Liberals pretend differences in ability don't exist, while conservatives like to defend the accuracy of our "gut" prejudices and argue there's nothing wrong. IMHO equality of opportunity comes when people make an effort to look at actual ability as objectively as possible.
It's in your best interest as an employer to do this. Good people are very hard to find, and if you're running some kind of dumb ape-brain filter that excludes ~50% of the population from consideration you are handicapping your recruitment efforts pretty badly. Think like a human being.
I (male, for the record, but gender-specific behavior isn't my point here) used to write/say that I was "part of a team" that did X or "worked on a team" that did Y, but I stopped doing it for a couple reasons, the most prominent being that "worked on a team that..." is usually obvious given the context of the conversation and the scale of the work discussed. In those cases, explicitly mentioning it feels like you're going out of the way to distance yourself from the work and describe yourself as a bystander - someone who just happened to be on the team that did this great work.
Personally, I hate phrases like "I worked on a team that did X" because I can't tell what you did. Where you the lead, a junior, or did you get called in when everything exploded and saved the day?
I'd rather see "I did the Y on the team that did X" as it's more descriptive and leaves less to my imagination.
> " Women would highlight the cooperative nature of their work, while men would highlight that they did it all themselves. In my experience the latter is typically more honest, especially if you actually were working on a team or on a team-oriented college project.
Did you mean to say the former was more honest, i.e. that saying “we did X” is more honest than “I did X?"
It depends on the context of course. If it was your project alone, then you did it. If others contributed non-trivial amounts, it was a team effort. I noted that the males tended to bury that, while the females tended to highlight it.
I really want to hear what the individual contributed, though, and actually couldn't care less what their team did as a whole. I've heard way too many "we"s that were actually "they"s, while the subject was present, but not actually contributing that much. It's really easy for a bad developer to hide in a large team.
EDIT: and it's also easy for a good developer to get quagmired in a bad team! Just because the project as a whole didn't do much doesn't mean the individual is bad. So I still really only want to hear about the individual efforts.
This would be why I spend almost no effort on the interview process. I put people to work right away to force them to prove whether or not they can do the job.
Well, I suppose they could be subcontracting it out, but at that point, I don't care. The work is getting done and at a price I can manage.
Well I think the more honest part is I myself worked on part X and we the team created product Y made out these multiple parts. How big of a part you had does matter.
For example, I made the iOS UI part of this text chat app, and my team members created data models and networking code that I used to make the UI part. Another team member made the android part using the cross platform code and a bunch of others made the server part and so on. Together we made a text chat app pair.
> Women would say things like "I worked on a team that did X," while men -- who obviously did the same kind of work -- would simply say "I did X."
While this anecdote seems to match traditional gender narratives, I'm interested in how you knew the men "obviously did the same kind of work."
You observed that women were more likely than men to emphasize teamwork and not their individual contributions. It seems that there are two hypotheses which result from that:
1. Women are more likely to emphasize team aspects and de-emphasize their contribution (less narcissistic)
2. Women are less likely to individual contribute and more likely to rely on teams for projects (less independent)
"Women would say things like "I worked on a team that did X," while men -- who obviously did the same kind of work -- would simply say "I did X." Women would highlight the cooperative nature of their work, while men would give the impression that they did it all themselves."
Cannot emphasise this quote enough. It is the difference between, "We" and "I", the basis of any community where you put yourself second. It is also the essence of good leadership.
If you want a lone wolf programmer as a new-hire in a small startup, maybe this type isn't for you. As your company grows, the type of programmer described by @api above, the one who works in teams where your livelyhood and codebase is on the line, is the person I'd want to work with and for.
> Women would say things like "I worked on a team that did X," while men -- who obviously did the same kind of work -- would simply say "I did X."
Depending on what X is, and on context, it can be obvious that "worked on X" simply means that they worked on X, but that doesn't mean that others did not also work on it, with them. Especially if "X" is too big and complex for one person to do on ones own. In particular, researchers in CS (maybe coincidentally majority of men) tend to say that they "work on" or "are interested in" X, Y, Z - but that doesn't really mean that they don't collaborate or work with other people on those things. It's not exclusionary at all.
But this interpretation doesn't fit with the usual men-are-egotistical-narcissists-compared-to-women. So pay this comment no mind.
"But this interpretation doesn't fit with the usual men-are-egotistical-narcissists-compared-to-women. So pay this comment no mind."
I'm just reporting the patterns I saw. That's all.
I wouldn't say "egotistical narcissist," just that the culture probably holds out different behavioral expectations for men and women and that gives rise to differences in how people present themselves.
> I'm just reporting the patterns I saw. That's all.
Well let's not kid ourselves. :) If a pattern you happened to observe reflected poorly on women - like if you observed that many more of them were incompetent than men[1] - you wouldn't report on it here. Or else you would just get downvoted or ignored. Can you honestly object to that?
Casually reporting on something which shows women as a group (or subgroup, like in tech) is viewed at best as being in poor taste, or at worst as sexism or misogyny. And there are good arguments for calling it that. But do the same for men and it's just innocuous "observations".
[1] For the sake of the argument: totally hypothetical.
We (www.gapjumpers.me) are trying to solve the gender diversity issue in tech by using blind auditions as a first filter. Our data shows that blind auditions result in greater diversity numbers.
By using a technical challenge as the first step and fast forwarding people to an advanced interview stage, we're eliminating candidate reluctance and increasing transparency (the candidate interacts directly with the hiring manager through our platform).
In our experience, telling the candidate "Show us your skills, impress us, and you'll get an interview" is better than "Tune up your resume to pass every possible filter". Showing your skills early on also removes a lot of stress from the candidate and gives you something solid to talk about during the interview.
Uhm. I was hoping for something much more comprehensible and comprehensive than a series of blurbs and unlabeled pie charts. A statistical comparison of your cohort compared to the background population and relative performance is what interests me.
I want to know how your program does compared to a hypothetical ideal gender-bias-free hiring system. If what you present there is truly all that your data shows, I have some grave concerns about your company.
I wonder if you could also make the matching part of the process blind as well? That is, the article mentions that women might be less likely to apply to places like Fog Creek - but if you matched applicants to employers blindly both ways (blindly to them, of course, you the matcher would not be blind), so the applicants didn't know who they were interviewing for during the interview? Then developers with lower self-esteem but great skills might end up interviewing for high-end places, and not even be stressed during the interview, because they don't know who it's with.
I'd find this very interesting myself, although I'd have concerns about potential employers in some ways. As long as I was able to provide a list of employers I DON'T want to work for up front and have those auto-filtered from my blind interviews I'd be totally open to trying out this process through a recruitment firm. I personally tend to pick my employers broadly on three questions: 1) Are they doing something that's actually innovative? 2) Do they behave ethically in the marketplace towards their employers, vendors, and customers? 3) Is the position located somewhere that I'd be okay with living?
If the answer to any of those is "no", I won't even bother applying/interviewing and I automatically turn down recruiters from that company. #2 in particular is a big one, there are way too many businesses in the world from startups to big corps that behave unethically, and I refuse to be a part of it.
Being able to work with a recruiting platform where I can detail these qualifications up-front and trust that they'll be respected would give me the piece of mind to put myself in a situation where I'd be matched to an employer based on their needs and my qualifications without concern that I'd be getting screwed somehow.
This is an interesting approach. We will probably try it in the future but at the moment feedback says that candidates almost always want more information (more about the company, more about the people they'd be working with, more about the kind of work they'd be doing). But I definitely agree on how it might take the stress out of the process.
The only issue is that hiring and getting hired requires significant commitment on both ends and long term anonymity might not work really well under those conditions.
Our profile capture does not require applicants to include their demographic information, hence the only way we've been able to easily split by gender is by identifying their profile images and names.
Of course, outcomes have shown that racial diversity is also helped along. Our hypothesis is that if you work to remove bias through process there should be a measurable (even if it's small) change in diversity metrics.
It doesn't draw the connection directly, but part of the training program deals with technical interviews and getting the fellows more practice, and hopefully more comfort, with them. My guess is that once people have been through some mock interviews, the intimidation factor goes down quite a bit.
I'm glad the "rigorous testing process" was mentioned. The facade of "we only hire superninja-kickass-hyperexperts and we will verify that you are such a programmer before we hire you" attracts a certain type of person, and intimidates another.
Yup. And I've always suspect that this cant was ultimately something of a foil, and that the actual messages it's intended to send are of a rather different sort. To wit:
Internal: Hoo-ray for you! You made it through the gauntlet! Aren't you just swell and awesome!
External: You know you have to invest in us / buy our products because we only hire the top 0.001 percent! And we have a rigorously testing process (as verified by the magic of survivorship bias) to prove it!
>attracts a certain type of person, and intimidates another. And it may be that it intimidates more women than men.
This is classic horseshoe theory, where extremists on both ends of a spectrum actually wrap around and start to sound the same. In this case they both believe that women are delicate flowers who can't handle the pressure, stress, and competition of the working world like men can.
There's a difference between "lower self-confidence" and "can't handle pressure, stress, and competition". It's also possible that even women who have high self confidence in their abilities are less confident that they will be judged positively by others.
There's plenty of research to the effect that men will apply to a job if they meet 40% of the listed requirements, whereas women won't until they meet 90%. And when you don't meet someone's default expectations of what kind of person a "software developer" (or a "nurse" or a "CEO") is, you often have to work harder to prove that you are one.
But also, why does the working world have to be competitive rather than collaborative?
To me at least, what's intimidating about applying for a job described as Fog Creek's are isn't that they're competitive -- they don't say "you'll be in a knock-down drag-out competition to beat 100 other candidates". It's that the criteria for acceptance are high -- only those who have an extremely high opinion of their own abilities will consider themselves likely not only to meet those criteria but to be judged by others as meeting those criteria. And many people who are very capable are unlikely to judge themselves so favorably. Women in particular are socialized to underrate our own abilities and conditioned to expect our abilities to be underrated by others, especially in STEM fields.
But the article seems to glide past that and presents the idea that five years of experience might change someone's willingness to apply. And it probably will help, but that amount of experience is unlikely to change someone's basic perception of their ability to succeed in a high-stakes one-chance prove-yourself situation.