4 ms·

What do you suggest instead of leetcode/hackerrank questions for companies to evaluate your coding skills during the interview?

andrew_ 6y ago
Use of "coding skills" in the question leads me to believe that your experience thus far has been more geared towards producing code than it has been solving problems and exercising critical thinking. Perhaps that's at the root of the follow up "what else, then?" question. It's 2021 and leetcode is a red flag of lazy hiring practices without imagination, and reflects poorly on an organization.
- jasonpeacock 6y ago
  Why is it an either-or choice? I want a candidate who can problem solve, apply critical thinking, and write code. I never said that coding was the only part of the interview - I asked how would you collect data about the candidate's coding skills.
flashgordon 6y ago
Well let me reverse this question. Why are leet code questions any more effective at evaluating your ability to succeed in a job that is most likely managing yet another another crud service than say pair debugging, or api schema development or code reviewing or even discussing previous oncall/escalation scenarios followed by fizz-buzz?
- jasonpeacock 6y ago
  Coding questions evaluate your ability to code, which is a significant part of most software engineering roles. There are other questions I ask to evaluate other skills, but I wanted to know how to collect data about coding skills without asking such leetcode questions.
  - flashgordon 6y ago
    There are coding questions and there are "coding" questions. If what you say is true nobody needs to "practise" leetcode intensively for 3 months before applying for a SWE role (or have a daily routine that involves leetcoding). Yet our interview philosophy is geared towards those who are actively able to workout before doing a tech interview. If all you want to know is to see if someone "can" code why not ask someone to write something they are familiar with and branch from there? You not only get coding but also an evaluation of maturity. Even if somebody memorized and regurgitated you can easily catch it rather than putting forward a contrived LC problem. PS - I am actually reasonably good at LC but only seasonally as I spend 3 months working out before a job search (thanks to an amazing wife who lets me ignore my kids as I lock myself in my study). The moment i start a job my LC skills drop pitifully and yet I am able to do well at my job (if reviews are anything to go by). I wonder how?

andrew_ 6y ago

Use of "coding skills" in the question leads me to believe that your experience thus far has been more geared towards producing code than it has been solving problems and exercising critical thinking. Perhaps that's at the root of the follow up "what else, then?" question. It's 2021 and leetcode is a red flag of lazy hiring practices without imagination, and reflects poorly on an organization.

jasonpeacock 6y ago
Why is it an either-or choice? I want a candidate who can problem solve, apply critical thinking, and write code. I never said that coding was the only part of the interview - I asked how would you collect data about the candidate's coding skills.

flashgordon 6y ago

Well let me reverse this question. Why are leet code questions any more effective at evaluating your ability to succeed in a job that is most likely managing yet another another crud service than say pair debugging, or api schema development or code reviewing or even discussing previous oncall/escalation scenarios followed by fizz-buzz?

jasonpeacock 6y ago
Coding questions evaluate your ability to code, which is a significant part of most software engineering roles. There are other questions I ask to evaluate other skills, but I wanted to know how to collect data about coding skills without asking such leetcode questions.
- flashgordon 6y ago
  There are coding questions and there are "coding" questions. If what you say is true nobody needs to "practise" leetcode intensively for 3 months before applying for a SWE role (or have a daily routine that involves leetcoding). Yet our interview philosophy is geared towards those who are actively able to workout before doing a tech interview. If all you want to know is to see if someone "can" code why not ask someone to write something they are familiar with and branch from there? You not only get coding but also an evaluation of maturity. Even if somebody memorized and regurgitated you can easily catch it rather than putting forward a contrived LC problem. PS - I am actually reasonably good at LC but only seasonally as I spend 3 months working out before a job search (thanks to an amazing wife who lets me ignore my kids as I lock myself in my study). The moment i start a job my LC skills drop pitifully and yet I am able to do well at my job (if reviews are anything to go by). I wonder how?