7 ms·

I come from a city university, and in my experience many of the master students there (information system or computer science program) are not up to standard wh

I come from a city university, and in my experience many of the master students there (information system or computer science program) are not up to standard when they leave the program (i.e. they can't really program, let alone with theories). I think this is the problem with my school, but when I had this conversation with an old friend he felt that his master classes focused on theory, very little room for programming. He also had the feeling that in PhD, unless one's research requires programming, theory is emphasized over engineering. Perhaps this is an edge case... I don't know. Then a long time ago, someone claimed to have PhD failed my interview. I asked this candidate to pick a language and write a function to determine whether a string is palindrome or not. He couldn't even start with a function declaration (I don't even program Java but I know how to write a function declaration). I was disappointed, but I gave him another chance. This time I wrote the declaration on the board, and I asked him to write code to determine whether a number is even or odd. He flunked again. Our recruiter never followed up to request evidence for his PhD degree because he failed the interview anyway. Even though I still believe he lied, but if he didn't, how did this university, a university with good reputation, gave him his degree? (btw, every single time I bring this up on HN, I get downvote, I never understand. So challenge me if you think I am lying.)

umanwizard 9y ago
I assume you get downvoted because theoretical CS research really doesn't need to involve coding, so it's not as odd as you seem to think.
- brians 9y ago
  Dijkstra famously compared this to astronomers who have no interest in telescopes or observing. I love theory, would happily retire to it—but I can’t see how anyone could have the right imagination calibration without lots of practice.
  - dagw 9y ago
    Dijkstra famously compared this to astronomers who have no interest in telescopes or observing That certainly isn't my interpretation of the quote (unless you're thinking of a different quote than the one I'm familiar with) The quote I know is: "Computer Science is no more about computers than astronomy is about telescopes."
    - wiz21c 9y ago
      yep, the fact that computer sciences spend a lot of time on pen/paper/board and brings proofs without actually program them doesn't mean that CS researchers are not interested in code. Code is their ultimate goal, but it's an implementation detail :-)
  - kaeluka 9y ago
    No offence, but you sound like you haven't worked in or close to such a field. CS theory very often doesn't require coding at all, and coding would be barely helpful. edit: this is not a value statement.
  - Chinjut 9y ago
    You have definitely gotten Dijkstra's intent with that famous quote completely backwards…
pm90 9y ago
A PhD does not guarantee good programming skills. I think you are mistaken in assuming that and hence get downvoted. I have personally found most people who finish/almost finish their PhD's to be very good in describing problems, approaching solutions from different paradigms and articulating their thinking in writing. If you're trying to hire a PhD for a programming job, in most cases you WILL be disappointed. That said usually they are fast learners and can quickly pick up good programming skill pretty quickly. (Its not as hard as people imagine it to be). I was planning on doing a PhD but left after my Masters precisely because I wanted to improve my programming skills, which I knew the PhD didn't care much about.
- yeukhon 9y ago
  I was disappointed that he couldn't even come up with a solution for odd or even. I get it for some low-ranked schools, but if for example someone from UC Berkeley can't do any decent programming (I don't expect everyone write a transcompiler because I have no idea how to myself) I'd be very surprised.
  - AlphaSite 9y ago
    Transpilers are easy, algorithms are hard.
xor1 9y ago
I know someone currently enrolled in a CS Master's program. I forgot the term, but they have to spend an extra year taking undergrad classes, since their undergrad wasn't CS. Their friends do the vast majority of their homework and projects for them, and these assignments + open-book take-home quizzes/tests (that they also get help with) count for enough of their grade that they can tank the stuff that you can't cheat on and still pass all of their classes. I hope that they'll be exposed once they get to post-grad next year, but I doubt it. I don't know anything about being a master's student, but I'm assuming they'll just have their hands held by professors, and get carried through everything by their fellow students. They won't even need to reach out to people outside of the program. Anyways, now I know how people finish a CS degree without actually being able to code...
arvinsim 9y ago
I guess it is really hard to believe that a PhD graduate could not even figure out if a number is odd-even. It is usually taught at the first year in a CS bachelor. Maybe they are from a different bachelor?
WhitneyLand 9y ago
Your interview questions sound like bullshit. What are these interviews for, how did you arrive at those questions, and what leads you to believe they have a strong correlation with success?
- troupe 9y ago
  The interview questions don't sound particularly rigorous, but asking someone to pick a programming language and do something simple with it sounds like a good way to make sure that they at least have actually done programmed before. If the job involved any type of programming, this would seem to strongly correlate with success. What do you think would be better?
  - WhitneyLand 9y ago
    It sounds like a good way? Yes, I agree intuitively it sounds reasonable, but that alone is not considered a best practice. Why do we all work so hard to go by good data in our real work, but when it comes to interviewing someone we’re encouraged to just wing it, or use our best judgement? Careers can be affected with this stuff, and bad hiring decisions in either direction can hurt an organization. There is research done on hiring but the learnings are slow to benefit the practice. General intelligence is one of the strongest indicators, but not gotcha trick puzzle questions. Work samples are predictive, but again, approach is everything. Track record of past success (when that applies) is important. This is not my field so I struggle like everyone else to keep up with what’s right. It just burns me that such litttle rigor is typically applied to something so important. Maybe it’s because interviewing is just one of those topics that seem easy, like, nutrition - most people think they know a lot more about it than they really do.
    - dagw 9y ago
      not gotcha trick puzzle questions. In what world is testing if a number is even or odd a gotcha trick puzzle question?
      - 9y ago
syllogism 9y ago
You get downvoted because your story is about as rational as someone getting angry at you because you did something in a dream. Compare: "He told me he was the king of Nigeria. I didn't believe him, but if he is, how come the king of Nigeria is white?" By far the simplest explanation is that the candidate lied about having a PhD. You have no reason to believe this isn't the case. And yet you're holding onto this story as though the candidate didn't lie, and bringing it up in that context.

umanwizard 9y ago

I assume you get downvoted because theoretical CS research really doesn't need to involve coding, so it's not as odd as you seem to think.

brians 9y ago
Dijkstra famously compared this to astronomers who have no interest in telescopes or observing. I love theory, would happily retire to it—but I can’t see how anyone could have the right imagination calibration without lots of practice.
- dagw 9y ago
  Dijkstra famously compared this to astronomers who have no interest in telescopes or observing That certainly isn't my interpretation of the quote (unless you're thinking of a different quote than the one I'm familiar with) The quote I know is: "Computer Science is no more about computers than astronomy is about telescopes."
  - wiz21c 9y ago
    yep, the fact that computer sciences spend a lot of time on pen/paper/board and brings proofs without actually program them doesn't mean that CS researchers are not interested in code. Code is their ultimate goal, but it's an implementation detail :-)
- kaeluka 9y ago
  No offence, but you sound like you haven't worked in or close to such a field. CS theory very often doesn't require coding at all, and coding would be barely helpful. edit: this is not a value statement.
- Chinjut 9y ago
  You have definitely gotten Dijkstra's intent with that famous quote completely backwards…

pm90 9y ago

A PhD does not guarantee good programming skills. I think you are mistaken in assuming that and hence get downvoted. I have personally found most people who finish/almost finish their PhD's to be very good in describing problems, approaching solutions from different paradigms and articulating their thinking in writing. If you're trying to hire a PhD for a programming job, in most cases you WILL be disappointed. That said usually they are fast learners and can quickly pick up good programming skill pretty quickly. (Its not as hard as people imagine it to be). I was planning on doing a PhD but left after my Masters precisely because I wanted to improve my programming skills, which I knew the PhD didn't care much about.

yeukhon 9y ago
I was disappointed that he couldn't even come up with a solution for odd or even. I get it for some low-ranked schools, but if for example someone from UC Berkeley can't do any decent programming (I don't expect everyone write a transcompiler because I have no idea how to myself) I'd be very surprised.
- AlphaSite 9y ago
  Transpilers are easy, algorithms are hard.

xor1 9y ago

I know someone currently enrolled in a CS Master's program. I forgot the term, but they have to spend an extra year taking undergrad classes, since their undergrad wasn't CS. Their friends do the vast majority of their homework and projects for them, and these assignments + open-book take-home quizzes/tests (that they also get help with) count for enough of their grade that they can tank the stuff that you can't cheat on and still pass all of their classes. I hope that they'll be exposed once they get to post-grad next year, but I doubt it. I don't know anything about being a master's student, but I'm assuming they'll just have their hands held by professors, and get carried through everything by their fellow students. They won't even need to reach out to people outside of the program. Anyways, now I know how people finish a CS degree without actually being able to code...

arvinsim 9y ago

I guess it is really hard to believe that a PhD graduate could not even figure out if a number is odd-even. It is usually taught at the first year in a CS bachelor. Maybe they are from a different bachelor?

WhitneyLand 9y ago

Your interview questions sound like bullshit. What are these interviews for, how did you arrive at those questions, and what leads you to believe they have a strong correlation with success?

troupe 9y ago
The interview questions don't sound particularly rigorous, but asking someone to pick a programming language and do something simple with it sounds like a good way to make sure that they at least have actually done programmed before. If the job involved any type of programming, this would seem to strongly correlate with success. What do you think would be better?
- WhitneyLand 9y ago
  It sounds like a good way? Yes, I agree intuitively it sounds reasonable, but that alone is not considered a best practice. Why do we all work so hard to go by good data in our real work, but when it comes to interviewing someone we’re encouraged to just wing it, or use our best judgement? Careers can be affected with this stuff, and bad hiring decisions in either direction can hurt an organization. There is research done on hiring but the learnings are slow to benefit the practice. General intelligence is one of the strongest indicators, but not gotcha trick puzzle questions. Work samples are predictive, but again, approach is everything. Track record of past success (when that applies) is important. This is not my field so I struggle like everyone else to keep up with what’s right. It just burns me that such litttle rigor is typically applied to something so important. Maybe it’s because interviewing is just one of those topics that seem easy, like, nutrition - most people think they know a lot more about it than they really do.
  - dagw 9y ago
    not gotcha trick puzzle questions. In what world is testing if a number is even or odd a gotcha trick puzzle question?
    - 9y ago

syllogism 9y ago

You get downvoted because your story is about as rational as someone getting angry at you because you did something in a dream. Compare: "He told me he was the king of Nigeria. I didn't believe him, but if he is, how come the king of Nigeria is white?" By far the simplest explanation is that the candidate lied about having a PhD. You have no reason to believe this isn't the case. And yet you're holding onto this story as though the candidate didn't lie, and bringing it up in that context.