Coding Interviews are like testing a Marathon runner by how fast they can run the 100 meter dash. Sure people who struggle in the 100 meters will also probably struggle in a marathon, but the best marathon runners aren't always the fastest sprinters.
I am also not a fan of the "hey code this while I watch you" process. There is no reality to this process, I have written code that many of the people "testing" me would not have ever considered, does that make me some super programmer? NO, it makes me a guy that had no other choice so I learned it and did it. Asking me to write your next brainfuck interpreter or whatever algorithm you want may or may not result in positive results, but either way you will learn very little of what I am capable of in your pre determined time period.
If you want to know what I can offer after 25+ years in this industry, look at what I have done, talk to me for 15 minutes without the entitled attitude that tells me you are a new grad and have no real world experience. That fact is, I don't care if you are a new grad, but I admit I do get excited when I meet new grads because that means I can likely learn things from you, at least as much as you can learn from me. But in the end, don't BS me, I've written production level code for years and (overall) know the reality versus your fiction you are telling yourself.
In the end, I don't mind hard questions or questions that are related to the issue you will have me solve. But to ask me to write an array reversal in Javascript and I do so at a low level and you ding me cause I didn't use some new fangled framework is just an indictment of your own education not my experience.
From the subheading: "Live coding interviews test culture fit under the guise of testing competence."
When I'm interviewing it's not a guise. I am very clearly interviewing you for:
- cultural fit
- how you approach the problem
- and how well you communicate
For us the actual code we ask to write is very simple. Think fizz buzz. I would never dream to ask to, for example, write code to balance a binary tree. I think those questions are crazy. 90% of our team couldn't do that on a white board and that's completely OK because it's not a skill you need day-to-day. If I based decisions on those kind of algorithms I wouldn't have hired most of our team and it would have been our loss not theirs.
With that said, some people couldn't complete our basic challenge so even though it's not intended as a code competency hurdle, it effectively has weeded out some people who clearly would not be able to function in their daily job.
I'm not sure how I'd react if someone outright refused to do it. I guess I'll deal with that if it happens.
"it effectively has weeded out some people who clearly would not be able to function in their daily job"
Did it, or did it weed out people who got so nervous at being asked to perform for an audience on demand that they froze up? I have social anxiety, and while I've been super lucky to get jobs through referrals of other colleagues (who introduced me ahead of time to folks I would be interviewing with), if I had to do this in front of a bunch of strangers judging me, I'd probably appear to be very bad at my job. How I act normally with people whom I know is very different than how I act around people whom I feel are just there to judge me and decide how worthy I am.
You raise a valid concern I don't have a good answer for you.
I've interviewed people with some pretty severe anxiety who have passed it and some who accepted offers. But that is not to say there wasn't someone weeded out that shouldn't have been.
Although for the people who failed the coding part of it I'd say they definitely behaved more overly confident than anxious.
Edit: with that said I have some experience successfully teaching programming to people with disabilities (in a university setting) which probably makes me a fair amount more qualified than many other higher managers. But again, there are no guarantees.
I too find that when programming in front of people my ability to think logically about the problem is hampered by the other half of my brain trying to analyze the interviewers reactions. I am not able to fully focus on the problem because my mind is constantly distracted.
Growing up I played sports all my life and never had a problem in front of crowds. I suspect this is mainly thanks to the cerebellum acting as a co-processor to the pre-frontal cortex of sorts. In an interview however, there is simply way to much ancillary analysis going on to fully focus on the task at hand.
6 comments
[ 3.4 ms ] story [ 28.0 ms ] threadIf you want to know what I can offer after 25+ years in this industry, look at what I have done, talk to me for 15 minutes without the entitled attitude that tells me you are a new grad and have no real world experience. That fact is, I don't care if you are a new grad, but I admit I do get excited when I meet new grads because that means I can likely learn things from you, at least as much as you can learn from me. But in the end, don't BS me, I've written production level code for years and (overall) know the reality versus your fiction you are telling yourself.
In the end, I don't mind hard questions or questions that are related to the issue you will have me solve. But to ask me to write an array reversal in Javascript and I do so at a low level and you ding me cause I didn't use some new fangled framework is just an indictment of your own education not my experience.
When I'm interviewing it's not a guise. I am very clearly interviewing you for:
- cultural fit
- how you approach the problem
- and how well you communicate
For us the actual code we ask to write is very simple. Think fizz buzz. I would never dream to ask to, for example, write code to balance a binary tree. I think those questions are crazy. 90% of our team couldn't do that on a white board and that's completely OK because it's not a skill you need day-to-day. If I based decisions on those kind of algorithms I wouldn't have hired most of our team and it would have been our loss not theirs.
With that said, some people couldn't complete our basic challenge so even though it's not intended as a code competency hurdle, it effectively has weeded out some people who clearly would not be able to function in their daily job.
I'm not sure how I'd react if someone outright refused to do it. I guess I'll deal with that if it happens.
Did it, or did it weed out people who got so nervous at being asked to perform for an audience on demand that they froze up? I have social anxiety, and while I've been super lucky to get jobs through referrals of other colleagues (who introduced me ahead of time to folks I would be interviewing with), if I had to do this in front of a bunch of strangers judging me, I'd probably appear to be very bad at my job. How I act normally with people whom I know is very different than how I act around people whom I feel are just there to judge me and decide how worthy I am.
I've interviewed people with some pretty severe anxiety who have passed it and some who accepted offers. But that is not to say there wasn't someone weeded out that shouldn't have been.
Although for the people who failed the coding part of it I'd say they definitely behaved more overly confident than anxious.
Edit: with that said I have some experience successfully teaching programming to people with disabilities (in a university setting) which probably makes me a fair amount more qualified than many other higher managers. But again, there are no guarantees.
Growing up I played sports all my life and never had a problem in front of crowds. I suspect this is mainly thanks to the cerebellum acting as a co-processor to the pre-frontal cortex of sorts. In an interview however, there is simply way to much ancillary analysis going on to fully focus on the task at hand.