Honestly, I really like whiteboarding. I liked doing programming contests too, which are like whiteboarding except harder and you spend more time at them. There are times when I have gotten together with other like-minded software engineers and given each other whiteboard programming problems for fun. (Also to test whether problems are good interview problems.)
A lot of people like whiteboarding problems and like interviewing with whiteboarding problems. Those are usually also the people who pass whiteboarding interviews. Just because there are a lot of complaints about them, don't assume that represents everyone.
It’s probably the most strongly correlated with liking and being able to do the job signal I’ve ever seen.
First, you’re not going to get your resume in front of me without a college degree, CS degree or equivalent experience is a bar that recruiters apply well before I’m involved.
But as far as I can tell, Google CodeJam is an individual sport (I may be mistaken about that, but Google’s page won’t load in my browser, so I could only read the Wikipedia page about it).
The key thing about the ACM contest is it requires BOTH effective cooperation with a teammate AND the desire to solve programming puzzles.
Now, just to be clear, I’ve fired people who passed this bar; being a smart puzzle solver who’s social enough to collaborate is a great start, but it turns out lots of people just aren’t interested in shipping software. I still think ACM team participation is a strong signal, it just doesn’t cover the fact that being an adult means 90% of every job is boring, and not everyone is ready to do the boring stuff when they’re fresh out of college.
I think the right answer here, to the extent that there is a right answer, is to give applicants a choice.
Consistent processes are also important for avoiding bias. For example, take-home projects are more likely to make sense for young candidates who aren’t raising families and have free time after work. If you have most of your young candidates doing take-home projects and the older candidates doing whiteboarding, you will end up evaluating candidates differently based on age.
If they have github, I'll ask them to walk me through a few files, asking their reasons for doing X, why not Y - what does this do?
If they prefer take-home, I'll give them something to create/improve.
Want to do something now? knock yourself out - with or without me watching.
I don't need to sort the candidate into an ordered list, just filter out the "over confident".
The questions I want answering is: Can they actually code? Does their skill match their experience? Is their skill level similar to the codebase?
Are you saying in general - or just for Windows-based software (since you mentioned needing to set up Visual Studio)?
The latter I can understand; but the former makes me want to ask, "Who applies for an SWE position who doesn't write code at home?"
I've been an SWE professionally since 1991 when I was 18; I got my first computer in 1984, and have been "coding at home" ever since. I'm not saying "to be an SWE you need a similar experience" (I know that's untrue) - I'm just relating that I've been coding for a long time, and "at home" is very much a part of it.
Then again, I've heard of people who absolutely hated coding, but were extremely good at it (and I assume made good money doing it), but when they went home, they didn't even want to think about such a thing. Maybe you are similar in that regard?
I guess I am just curious and a bit fascinated by your comment...
Professional baseball players rest after games. They don't finish up and then immediately start practicing again. And soldiers and police get down-time.
Just because I'm doing something I enjoy and that I want to do, doesn't mean I should do it all of my waking hours. I'm not doing this to code, I'm doing this because my employer is paying me to build cool shit and I like doing that. I don't lift weights 24/7, I don't run 24/7, I don't hike 24/7, and I don't program 24/7.
Frankly your attitude is one of the reasons why we don't have any diversity in this profession at most self-described tech companies, and it also underlies a lot of the seeming ageism today. All of these self-described non-conformists expect everyone in the profession to conform to their attitude and it's sickening.
That's where the person who knows how to set up a VirtualBox VM running Windows and VS would have a large advantage; when they brought in the result, showing their MBP running VS in a virtualized manner, and explained what they had to do - that's a bit more of an impression than someone who says "they can't do it because MBP".
Also, I'm nowhere near ops jobs, so I don't think those people interviewing me would care much about those kind of logistics (anyways, they wouldn't want to know the details). It wouldn't score any "points" beyond just getting the task done at all.
When Im looking for a job, I make it pretty clear what Im ok and not ok with, which will dictate which companies Ill interview at or not. If we were in a dotcom crash it would be a different story, but right now its definitely something people can do.
Isn't that the point of a resume? What am I missing?
Resume review -> take home exercise -> phone interview -> onsite interview -> offer.
If in your case, you demand allocating more time for that step, consider adding additional filter (requiring 10-15 mins) after the resume review. And then redirect [part of] resources (money) from the onsite part to pay for that take home task.
Another red flag I see in your process - asking for allocating time before even talking to a human being - that would be another reason for an immediate rejection. Consider moving your phone part before the take home task.
When there is something before the phone interview, it is a coding assessment; these are generally different from a take home assignment in that they are generally timed and automatically scored, e.g. you hit start and have two hours and then a computer scores you. With take home exercises, I've always had engineers at the company read my code and discuss it with me.
But to be fair the problems were not too hard, the interview was set-up properly (problem files on place and a good code editor at hand, no tricks or gotchas)
And the worse of course is having to write fizzbuzz in a whiteboard, I just have to roll my eyes then
Doing 10 hours of work to apply for every random "we make another business PHP/crud website" company is not worth it.
Your company isn't Google, don't interview like it is. I am not going to write custom code for you, but I am willing to flash my credentials.
Some people even suggest that advanced degrees are an anti-indicator of ability:
http://blog.alinelerner.com/how-different-is-a-b-s-in-comput...
And mostly they aren't wrong.
Too true, I've started to recruit non-CS people and teaching them how to code. The state of CS education in the US is insanely inconsistent. But I'm lucky, I have lots of time to train people; most companies can't handle months of startup time.
That's why companies should allow laptops.
> Take-home assignments are unpaid work
True, they could be optional. You may not get the job if you don't do it. Honestly, if you can't spend one or two hours on a take home assignment during a 7 day period, you certainly have different problems than applying at this company.
> Pair programming during interviews makes people uncomfortable
I agree, pair programming during an interview is mostly a no go. Not because of the issues you mentioned but because most interviewers are totally unqualified to do this properly (unfortunately, most interviewers are also unqualified in general to conduct interviews). It takes a lot of skill to conduct pair programming interviews and draw the correct conclusions.
> People who currently have good jobs are unwilling to consider contract-to-hire
In US, there is practically no difference between those three. You can be fired any day anyway. But yes, if you have a secure job that is good, the bar for switching is high, but so what? Why even switch? It's your choice then... Luxury problem!
> Credentials, including graduate degrees, are no guarantee that someone is a good engineer
I would go as far as saying your credentials and degrees are horrendously irrelevant in predicting your skills as an engineer. The only thing you might be able to conclude is that a GPA 4 M.I.T. graduate probably can pick up missing knowledge quickly, as opposed to a no degree dish washer who learned "how to code" in a 6 week bootcamp.
On that note I have experienced first hand, Berkeley and Georgia Tech CS majors, who were absolutely incompetent on the job and just had no idea what the hell they were doing, with little improvement in sight...
That's an extremely privileged thing to say. And what a terrible way to start out your working relationship. The very first thing you ask of me is to do unpaid work in my free time? IF you're willing to do that, how else will abuse me in the future?
Also, why should they spend time on your homework before they know if you're a company they would even want to work for?
I think the reason why take-home assignments tend to feel more unreasonable is the asymmetry. If I attend an interview, it's costing a pile of my time, but it's also costing a similar pile of time for multiple people at the company where I'm interviewing, which gives me some reason to think that my application's going to be considered seriously and therefore that the effort I'm putting in isn't going to be totally wasted.
Whereas a company can hand out a take-home assignment with no effort at all, so the fact that I've been given one doesn't indicate that the prospective employer sees me as a serious prospect or that there aren't hundreds of other candidates in the same position as me, so it's no guarantee that my chances of getting hired aren't miserably low, so there's more risk that I'm wasting my time.
Although perhaps an escrow system would work. The company pays me a large sum (say $10,000), which I then pay back if they call me in for an interview. That way I know they are serious, and it actually forces them to consider the cost to themselves.
I'd suggest that an MSEE or MSCS degree from a good school tends to be pretty good training for working in the industry. The projects you complete are usually challenging, relevant, and intentionally educational.
For example, an MSEE might learn how design analog amplifiers and filters, program a DSP, lay out VLSI circuits for an ASIC, implement a RISC CPU in verilog, implement an internet router on an FPGA, and implement Wi-Fi on a software radio. I'd be happy to hire (or work with) someone who can do all that!
I'm sure some programs teach this stuff already. But I've hired people with graduate degrees in CS from well-known schools who didn't even use source control for their group projects in grad school -- they just emailed the source files to each other.
0. You seem to have misunderstood my electrical engineering example.
1. Why would an MSEE not qualify you to be hired as a junior electrical engineer?
2. Why would a PhD in the appropriate area not qualify you to be hired as a junior research scientist?
> In US, there is practically no difference between those three. You can be fired any day anyway. But yes, if you have a secure job that is good, the bar for switching is high, but so what? Why even switch? It's your choice then... Luxury problem!
It seemed weird to me that we are one of the only industries (well I guess engineering in general) who gets a contract that is contingent on each day until it is completed. At any given point you could be let go and yet if I contract someone to work on my home to legal work you setup a contract that states you owe them the money unless you are willing to go to court to contest their work.
When I last did a contract-to-hire it was basically "you can be let go for any reason, from slow output to your boss having a headache." Unless you get a sweet deal its difficult to pull a job with a set start / end time with guaranteed payout regardless of employment. I busted my hump at that last job so instead of doing my 6 months for the hire option they offered me the job before then 2nd month was over.
- Nobody likes whiteboarding
- Take-home assignments are unpaid work, and candidates who are raising children don't have time to do homework after they get home from work
- Pair programming during interviews makes people uncomfortable because of the pace or because it's not how they are used to working
- People who currently have good jobs are unwilling to consider contract-to-hire or "tryout" periods
- Credentials, including graduate degrees, are no guarantee that someone is a good engineer
The current interviewing situation in the industry sucks but all of the alternatives are worse in important ways.