4 ms·

Sites like HackerRank and TripleByte do a great job of telling me whether a candidate has been studying to pass interview questions on HackerRank and TripleByte

by yodon 7y ago

Sites like HackerRank and TripleByte do a great job of telling me whether a candidate has been studying to pass interview questions on HackerRank and TripleByte. They do very little to tell me whether, in the words of Joel Spolsky, they are "smart and gets things done."

mykowebhn 7y ago
You have me quite interested. What else do sites like HackerRank & TripleByte tell you?
- yodon 7y ago
  That assessing programmers is like assessing teachers - a much harder problem to solve well than naive commentators expect, complete with detailed measurements of the shape of the top of each iceberg.
duxup 7y ago
I'm sort of terrible at terminology and trivia. I finally got an interview with a take home assignment. They asked questions about my work and how I found solutions, most my explanations we're something like: "I just looked some things up and didn't like that approach because it seemed like it wasn't very flexible if things changed, but did like this one so I used it..." I was later told I was not the top candidate, but was the only one who produced an assignment that checked all their boxes and more and showed I could work independently and produce. First job aquired.
wolfgang42 7y ago
I'm going through the Triplebyte process right now (I passed their interview last month and have just started the bit where I talk with companies this week) and that is not at all what they seemed to be testing. Their famous quiz seemed to mostly be checking if you have any idea what you're doing[1], and their two-hour interview[2] spent less than ten minutes on data structures. I muddled through that bit by saying things that sounded plausible (with lots of disclaimers that that was what I was doing and that I had no experience with low-level code) and still passed it with flying colors. Having gone through this process I would say that having some experience and being able to write practical code is probably the best way to prepare, and studying interview questions would probably be fairly ineffective. [1] One of the incorrect answers was to use a factory function that passed the data to an assembly line interface which put it in a package class and shipped it to the user interface. [2] https://triplebyte.com/interview_guide https://triplebyte.com/interview_guide
- yodon 7y ago
  Thanks for this - I'm actually more familiar with more pure coder test approaches like HackerRank and just mentioned TripleByte in the same breath because they've been advertising their coding test heavily in parts of SF I see (proving that at least some ad dollars do work to some extent).

by yodon 7y ago

mykowebhn 7y ago

You have me quite interested. What else do sites like HackerRank & TripleByte tell you?

yodon 7y ago
That assessing programmers is like assessing teachers - a much harder problem to solve well than naive commentators expect, complete with detailed measurements of the shape of the top of each iceberg.

duxup 7y ago

I'm sort of terrible at terminology and trivia. I finally got an interview with a take home assignment. They asked questions about my work and how I found solutions, most my explanations we're something like: "I just looked some things up and didn't like that approach because it seemed like it wasn't very flexible if things changed, but did like this one so I used it..." I was later told I was not the top candidate, but was the only one who produced an assignment that checked all their boxes and more and showed I could work independently and produce. First job aquired.

wolfgang42 7y ago

I'm going through the Triplebyte process right now (I passed their interview last month and have just started the bit where I talk with companies this week) and that is not at all what they seemed to be testing. Their famous quiz seemed to mostly be checking if you have any idea what you're doing[1], and their two-hour interview[2] spent less than ten minutes on data structures. I muddled through that bit by saying things that sounded plausible (with lots of disclaimers that that was what I was doing and that I had no experience with low-level code) and still passed it with flying colors. Having gone through this process I would say that having some experience and being able to write practical code is probably the best way to prepare, and studying interview questions would probably be fairly ineffective. [1] One of the incorrect answers was to use a factory function that passed the data to an assembly line interface which put it in a package class and shipped it to the user interface. [2] https://triplebyte.com/interview_guide https://triplebyte.com/interview_guide

yodon 7y ago
Thanks for this - I'm actually more familiar with more pure coder test approaches like HackerRank and just mentioned TripleByte in the same breath because they've been advertising their coding test heavily in parts of SF I see (proving that at least some ad dollars do work to some extent).