4 ms·

At Stanford, computer science students are big cheaters

18 points by razorburn 17y ago

shalmanese 17y ago
Alternative title: one Stanford professor cares more about cheating that pretty much therest of the university combined. Cheating is now pretty much rampant at even the most prestigious colleges. The only way to not detect it is to turn an actively blind eye towards it
natrius 17y ago
That's not what the article says at all. In computer science, it's easier to detect cheating, which is part, if not most of the reason why it has a notably high proportion of honor code violations. The only people I knew who got caught cheating on assignments weren't computer science majors. Please just copy the headline from the article next time.
- jorgeortiz85 17y ago
  +1. Quoth the article: There is a higher incidence of cheating people that are not CS majors — the ones that are new to the field and potentially find themselves being in a situation where they can’t handle the assignment.
- makmanalp 17y ago
  Also, it says "About 1,000 students will have taken the course by the end of the year, with more than 75 percent of students passing through a computer science introductory course before the end of their time at Stanford." which makes it much more normal for there being a high number of cheaters. A lot of people are taking the course, so there are a lot more cheaters although the proportion might be the same. (I don't know if this is actually the case.)
dockd 17y ago
In college, you are punished for using others source. After college, you are rewarded. Perhaps there is a better way to teach than giving an entire class the exact same assignment. I wonder if you might find the same results in the basic math classes, where non-math majors just have to "tough it out". I would think it's harder to catch copying because the work can't be easily analyzed by a computer.
- dbz 17y ago
  Of course in college you are punished! For a whole boat load of students this is their first time programming (a bad time to start imho). We want them to learn. Not to hack something together! =]
- dasil003 17y ago
  Yes, but in the real world you are constantly faced with novel (albeit simple) problems, whereas in college you're doing the same trivial thing tens of thousands have done before you. If you can't put forth the effort to figure things out on the canned problems how will you ever become competent to ship working software?
- swolchok 17y ago
  This argument gets trotted out in all sorts of internet discussions re: CS education, and it's bogus. You have to be able to stand on your own before you can contribute meaningfully to a group. More importantly, individualized assignments don't scale to classes of one or two hundred students.
tsally 17y ago
CS is not like an English paper where you can just turn it in and there’s not a way to say if it’s a good paper or not... at 3 a.m, you can be looking a program and definitively say that it’s not going to work. The ignorance I encounter among fellow CS people in this area never ceases to amaze. A strange phenomena in a group of otherwise intelligent individuals. Trust me when I say that you can evaluate a paper the same way you evaluate a program. In a functional sense you (the writer) can definitively say whether a paper will work or not [1]. Most people who make statements like the one I've quoted simply never have been taught the appropriate skills. [1] Creative writing is a different story.
- abscondment 17y ago
  Most papers aren't graded with some pass/fail grammar and punctuation rubric. I had many CS assignments for which no credit would be given if the automated submission site couldn't compile the work.
  - tsally 17y ago
    A program that doesn't compile is not the same as a paper with a typo in it. A program that doesn't compile is like a paper that never got printed; there is no finished form for the grader to interact with. Regardless of discipline, no credit should be given if the grader has nothing to work with.
  - emily37 17y ago
    I think the original poster might be referring to the strength of a paper's arguments rather than grammar. But I basically agree with you... most Stanford CS projects are graded almost entirely by automated tests (which test the output of the program, not just whether it compiles or not). When there are automated tests, you can predict how your program will be graded with much higher accuracy than you can predict the grade your paper will get, even if you are very good at evaluating papers.
    - tsally 17y ago
      When there are automated tests, you can predict how your program will be graded with much higher accuracy than you can predict the grade your paper will get, even if you are very good at evaluating papers. Of course. This is the same as saying that a CS student can better understand the reasoning of a program than the reasoning of a human being. As a rule programs are, in general, more predictable. I will disagree with the idea that it is impossible to achieve high accuracy in predicting grades on paper. It's anecdotal evidence only, but I've never been off by more than half a letter grade on any paper in the last three years of taking English classes (it's my minor). That's 6 classes with 5 different professors. So probably 20 to 30 papers? Usually I get the exact grade right.
dbz 17y ago
I understand that most of the cheating is done in the first cs course, but why on God's earth would you sign up for a major notoriously difficult if cheating can be a problem and you are the cheating type. I mean- If you have to cheat during the CS course, I can very easily see you never be a good programmer. A CS course isn't even like the real world programming jobs.
- emily37 17y ago
  I would guess that most cheaters don't know they're the cheating type until the moment they cheat. Also, one of the points made in the article is that most people attribute the honor code violations largely to non-CS majors. The mindset is probably that they never want to be a programmer, and all they want is a good grade and the general ed requirement out of the way.
- noilly 17y ago
  I go to stanford, and intro CS is required or strongly recommended in most majors (all engineering, hard science, and even social sciences)... so a pretty large fraction of the student body goes through the class- most prospective cs majors have already placed into higher level classes.
grinich 17y ago
I still stand by the belief that it is the professors' responsibility to create interesting and unique problems that cannot be solved easily with out-of-the-box libraries. Seems very basic for the amount of money I spend to attend college. Another solution is to use an unfamiliar language, such as Scheme.
- emily37 17y ago
  I think the main problem is students copying each others' code, not students copying library code. I don't think I've ever had a CS assignment at Stanford where I could have gone out and find library code to solve the problem.
  - grinich 17y ago
    Like code from previous terms? I didn't mean that using a library is bad— in fact, learning how to effectively combine and structure your code to work with preexisting libraries is a skill that should be enforced in every CS department. To clarify, I was referring to finding an exact solution to the problem (SCRAM state machine for example) and just plugging it in as the solution.
    - emily37 17y ago
      Yep, code from previous terms (friends who have taken the class before) or even from the same term (friends who are in the same class who have already finished the assignment). My impression is that the vast majority of honor code violations in the CS department are of one of those two types.
      - grinich 17y ago
        Do Stanford profs have tools to automatically check for copies? Or is it something the graders just notice.
        sparky 17y ago
losethos 17y ago
I wish there were a organization one could get a originality certificate from. I wrote the LoseThos operating system with it's compiler and I am forever arguing with scoffers. I want a Stanford teacher to certify my code. http://www.losethos.com http://www.losethos.com I wouldn't be surprised if teachers were a little annoyed with me for posting code their students might plagarise. I wonder what the ethics of posting such code are.

18 points by razorburn 17y ago

Alternative title: one Stanford professor cares more about cheating that pretty much therest of the university combined. Cheating is now pretty much rampant at even the most prestigious colleges. The only way to not detect it is to turn an actively blind eye towards it

natrius 17y ago

That's not what the article says at all. In computer science, it's easier to detect cheating, which is part, if not most of the reason why it has a notably high proportion of honor code violations. The only people I knew who got caught cheating on assignments weren't computer science majors. Please just copy the headline from the article next time.

jorgeortiz85 17y ago
+1. Quoth the article: There is a higher incidence of cheating people that are not CS majors — the ones that are new to the field and potentially find themselves being in a situation where they can’t handle the assignment.
makmanalp 17y ago
Also, it says "About 1,000 students will have taken the course by the end of the year, with more than 75 percent of students passing through a computer science introductory course before the end of their time at Stanford." which makes it much more normal for there being a high number of cheaters. A lot of people are taking the course, so there are a lot more cheaters although the proportion might be the same. (I don't know if this is actually the case.)

dockd 17y ago

In college, you are punished for using others source. After college, you are rewarded. Perhaps there is a better way to teach than giving an entire class the exact same assignment. I wonder if you might find the same results in the basic math classes, where non-math majors just have to "tough it out". I would think it's harder to catch copying because the work can't be easily analyzed by a computer.

dbz 17y ago
Of course in college you are punished! For a whole boat load of students this is their first time programming (a bad time to start imho). We want them to learn. Not to hack something together! =]
dasil003 17y ago
Yes, but in the real world you are constantly faced with novel (albeit simple) problems, whereas in college you're doing the same trivial thing tens of thousands have done before you. If you can't put forth the effort to figure things out on the canned problems how will you ever become competent to ship working software?
swolchok 17y ago
This argument gets trotted out in all sorts of internet discussions re: CS education, and it's bogus. You have to be able to stand on your own before you can contribute meaningfully to a group. More importantly, individualized assignments don't scale to classes of one or two hundred students.

tsally 17y ago

CS is not like an English paper where you can just turn it in and there’s not a way to say if it’s a good paper or not... at 3 a.m, you can be looking a program and definitively say that it’s not going to work. The ignorance I encounter among fellow CS people in this area never ceases to amaze. A strange phenomena in a group of otherwise intelligent individuals. Trust me when I say that you can evaluate a paper the same way you evaluate a program. In a functional sense you (the writer) can definitively say whether a paper will work or not [1]. Most people who make statements like the one I've quoted simply never have been taught the appropriate skills. [1] Creative writing is a different story.

abscondment 17y ago
Most papers aren't graded with some pass/fail grammar and punctuation rubric. I had many CS assignments for which no credit would be given if the automated submission site couldn't compile the work.
- tsally 17y ago
  A program that doesn't compile is not the same as a paper with a typo in it. A program that doesn't compile is like a paper that never got printed; there is no finished form for the grader to interact with. Regardless of discipline, no credit should be given if the grader has nothing to work with.
- emily37 17y ago
  I think the original poster might be referring to the strength of a paper's arguments rather than grammar. But I basically agree with you... most Stanford CS projects are graded almost entirely by automated tests (which test the output of the program, not just whether it compiles or not). When there are automated tests, you can predict how your program will be graded with much higher accuracy than you can predict the grade your paper will get, even if you are very good at evaluating papers.
  - tsally 17y ago
    When there are automated tests, you can predict how your program will be graded with much higher accuracy than you can predict the grade your paper will get, even if you are very good at evaluating papers. Of course. This is the same as saying that a CS student can better understand the reasoning of a program than the reasoning of a human being. As a rule programs are, in general, more predictable. I will disagree with the idea that it is impossible to achieve high accuracy in predicting grades on paper. It's anecdotal evidence only, but I've never been off by more than half a letter grade on any paper in the last three years of taking English classes (it's my minor). That's 6 classes with 5 different professors. So probably 20 to 30 papers? Usually I get the exact grade right.

dbz 17y ago

I understand that most of the cheating is done in the first cs course, but why on God's earth would you sign up for a major notoriously difficult if cheating can be a problem and you are the cheating type. I mean- If you have to cheat during the CS course, I can very easily see you never be a good programmer. A CS course isn't even like the real world programming jobs.

emily37 17y ago
I would guess that most cheaters don't know they're the cheating type until the moment they cheat. Also, one of the points made in the article is that most people attribute the honor code violations largely to non-CS majors. The mindset is probably that they never want to be a programmer, and all they want is a good grade and the general ed requirement out of the way.
noilly 17y ago
I go to stanford, and intro CS is required or strongly recommended in most majors (all engineering, hard science, and even social sciences)... so a pretty large fraction of the student body goes through the class- most prospective cs majors have already placed into higher level classes.

grinich 17y ago

I still stand by the belief that it is the professors' responsibility to create interesting and unique problems that cannot be solved easily with out-of-the-box libraries. Seems very basic for the amount of money I spend to attend college. Another solution is to use an unfamiliar language, such as Scheme.

emily37 17y ago
I think the main problem is students copying each others' code, not students copying library code. I don't think I've ever had a CS assignment at Stanford where I could have gone out and find library code to solve the problem.
- grinich 17y ago
  Like code from previous terms? I didn't mean that using a library is bad— in fact, learning how to effectively combine and structure your code to work with preexisting libraries is a skill that should be enforced in every CS department. To clarify, I was referring to finding an exact solution to the problem (SCRAM state machine for example) and just plugging it in as the solution.
  - emily37 17y ago
    Yep, code from previous terms (friends who have taken the class before) or even from the same term (friends who are in the same class who have already finished the assignment). My impression is that the vast majority of honor code violations in the CS department are of one of those two types.
    - grinich 17y ago
      Do Stanford profs have tools to automatically check for copies? Or is it something the graders just notice.
      - sparky 17y ago

losethos 17y ago

I wish there were a organization one could get a originality certificate from. I wrote the LoseThos operating system with it's compiler and I am forever arguing with scoffers. I want a Stanford teacher to certify my code. http://www.losethos.com http://www.losethos.com I wouldn't be surprised if teachers were a little annoyed with me for posting code their students might plagarise. I wonder what the ethics of posting such code are.