Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Coursera's Functional Programming - Cheating Discovered

pastebin.com

Post reply on HN
Print view View original

161–170 of 182 posts

Re: Coursera's Functional Programming - Cheating Discovered

#161

Post by pivo »

eranation wrote: post #151
sampo wrote: post #132

Earlier quoted context omitted.

If you have, for example class Rational(x: Int, y: Int) { def numer = x def denom = Y } and then homework is to implement def less(that: Rational) = ??? then you're gonna get 15 000 people answering with def less(that: Rational) = numer * that.denom and another 15 000 with def less(that: Rational) = numer * that.denom and maybe a couple thousand with something like def less(that: Rational) = numer*that.denom (This ex…

yes, but none of them are that trivial... and if they were, then cheating wouldn't make much sense. e.g. the coin change algorithm, I'm sure the distribution of answers was quite unique for anyone who tried solving it alone. same for even simple things like the pascal triangle. But it's a moot argument as we need to do a real research (or ask coursera / Martin Odersky for the data, which I'm sure HN community will be…

Actually I think some (many?) of the answers to the functional sets homework are that trivial in the sense that they're as short (or shorter) the the examples in the parent post. The only longer function was, I think, the one for which they provided a template implementation. So I think even that one would generate false positives somewhat easily.

Re: Coursera's Functional Programming - Cheating Discovered

#162

Post by esrauch »

pohl wrote: post #112
jonknee wrote: post #72

Earlier quoted context omitted.

It's "cheating" though in the sense that it's against a TOS on a website, but tons of things are against a TOS and almost no one reads them. They're pretending to be one, but Coursera is not a college, is not accredited and does not hand out degrees. This is a free online self-study course being taken by hundreds of thousands of people with varying degrees of seriousness. Some people may treat it like a college cours…

Right know there are actual students at EPFL who are taking the Functional Programming Principles in Scala course, using the same web site that random joes like me are using at the same time. So for me it's not an actual college, but it is for those students at EPFL. They have no choice but to enforce academic honestly rules, given that real students at a real institution are involved.

It makes sense for EPFL to use the course lectures and quizzes to help the students, but it makes absolutely no sense to use it for the grading part. It is trivially easy to make a second account, take a quiz, and just answer randomly to be able to see all the correct answers. It's absolutely absurd to not have a totally independent grading system.

Re: Coursera's Functional Programming - Cheating Discovered

#163

Post by michaelt »

scarmig wrote: post #148

Well, this seems like one of those social problems where a technical solution might be available. 1) Figure out the number of students taking course n. 2) For the first assignment, figure out the number of questions m you want. 3) Make a question bank of size M. 4) Satisfy the constraint n 5) You now can uniquely identify uploaders via the set of answers they upload. If you're willing to kick people out of the class…

The problem with question banks, at least in normal academia, is making sure everyone gets a question set of equal difficulty. If one person gets asked for an algorithm to find the length of a linked list and another gets asked for an algorithm to remove a node from an AVL tree, that's not fair on the second person.

You could experimentally calibrate the question bank, seeing which questions trip students up the most, but that requires asking the same questions a bunch of times to different students, which is exactly what using the question bank was supposed to avoid! And if you want to reuse the calibrated data every year, you'll have to ask the same questions every year, which we prefer to avoid. I suppose with a sufficiently large student population an effective calibration might be possible - maybe Coursera will be the ones to do it!

Re: Coursera's Functional Programming - Cheating Discovered

#164

Post by biafra »

PeterisP wrote: post #115
michael_miller wrote: post #66

Earlier quoted context omitted.

I disagree with your assertion. Cheating is not necessarily an act of malice. As an example, a friend of mine (who is from India) helped another student (also Indian) with a programming project in an intro to CS class at my alma mater. My friend was trying to help a friend out, not intentionally trying to break the rules. This took place during the first few weeks of school, mind you, and neither student had any sign…

I am taking this course - it's impossible to not notice these rules, there are annoying confirmations at every single submission or multiple choice test that you are submitting your own individual work only. I.e., submitting a program copied off web is 100% intentional fraud.

To me it is unclear whether it is allowed to talk about solution after the hard dead line. Is that ok? Not publicizing them just talking about them in a closed study group.

Re: Coursera's Functional Programming - Cheating Discovered

#165

Post by biafra »

eranation wrote: post #151
sampo wrote: post #132

Earlier quoted context omitted.

If you have, for example class Rational(x: Int, y: Int) { def numer = x def denom = Y } and then homework is to implement def less(that: Rational) = ??? then you're gonna get 15 000 people answering with def less(that: Rational) = numer * that.denom and another 15 000 with def less(that: Rational) = numer * that.denom and maybe a couple thousand with something like def less(that: Rational) = numer*that.denom (This ex…

yes, but none of them are that trivial... and if they were, then cheating wouldn't make much sense. e.g. the coin change algorithm, I'm sure the distribution of answers was quite unique for anyone who tried solving it alone. same for even simple things like the pascal triangle. But it's a moot argument as we need to do a real research (or ask coursera / Martin Odersky for the data, which I'm sure HN community will be…

The solution for the count change is already in the recommended literature (SICP). And this was mentioned in the forums.

Re: Coursera's Functional Programming - Cheating Discovered

#166

Post by wpietri »

dllthomas wrote: post #124
wpietri wrote: post #87

Earlier quoted context omitted.

As an employer, I take continuing education of any sort as a positive sign when looking at resumes. Software development is a Red Queen field; one has to keep up. However, I do take most certifications as mild negative signs. E.g., SCJP, CSM, PMP. A lot of for-profit certifications are worthless. A notable exception are some of the Cisco certs, which really put people through their paces. Which makes sense, as Cisco…

> Which makes sense, as Cisco benefits more from having expensive Cisco products work well than from certification revenue. I'm not sure about the actual comparative amounts, but Cisco makes a lot of money off of certifications. Of course, in the long run making them actually represent something is best for both sides of the business.

> I'm not sure about the actual comparative amounts, but Cisco makes a lot of money off of certifications.

80% of Cisco's revenue is hardware. And even the service side is something close to $2 billion. I suspect the amount of money they make off of certifications is a tiny slice of that. Even if there are 10k CCIEs each paying $1500 for a yearly test, that's only $15 million, which is nothing to Cisco.

> Of course, in the long run making them actually represent something is best for both sides of the business.

That's not true for many certifiers. Take a look at the CSM certificate. It is meaningless, but popular and therefore profitable.

Re: Coursera's Functional Programming - Cheating Discovered

#167

Post by wpietri »

learc83 wrote: post #103
wpietri wrote: post #87

Earlier quoted context omitted.

As an employer, I take continuing education of any sort as a positive sign when looking at resumes. Software development is a Red Queen field; one has to keep up. However, I do take most certifications as mild negative signs. E.g., SCJP, CSM, PMP. A lot of for-profit certifications are worthless. A notable exception are some of the Cisco certs, which really put people through their paces. Which makes sense, as Cisco…

Considering that the Coursera certificates require you to take a class (usually taught by experts in the respective field) as opposed to cramming for a certification exam, I'd say it's a good bit more valuable. I took the first AI class last Fall, and if you've never been exposed to the material, it was a very good survey of Artificial Intelligence.

I agree, just as long as the certificate actually means you've taken the course and done the work. If one can cheat one's way through the homework and then just cram for the final, the Coursera certs would end up being of low value. I hope not, though. I'd love to have a good way to tell the continuous learners from the people who get stuck.

Re: Coursera's Functional Programming - Cheating Discovered

#168

Post by wpietri »

jonknee wrote: post #97
wpietri wrote: post #76

Earlier quoted context omitted.

If people sign up for a course like this and then turn in work they copied from the Internet, I definitely don't want to hire them. Even if they weren't after the certificate. That's like cheating at solitaire. It's just sad. Also, do you have any evidence that the no-cheating, no-publishing-homework policy is buried in a TOS? If Coursera hasn't made it obvious enough, then sure, people could have made a mistake. But…

Much like I wouldn't hire a Stanford CS grad based on her certificate, I would recommend not hiring a Coursera "grad" based on a certificate either.

Nobody smart hires purely based on the certificate, but the question is what value one assigns to it. Anybody with a degree from a strong CS program is much more likely to get an interview from me. Ditto for anybody with a good online portfolio. I'd like to assign some value to Coursera courses, but we'll see.

Re: Coursera's Functional Programming - Cheating Discovered

#169

Post by pohl »

esrauch wrote: post #162
pohl wrote: post #112

Earlier quoted context omitted.

Right know there are actual students at EPFL who are taking the Functional Programming Principles in Scala course, using the same web site that random joes like me are using at the same time. So for me it's not an actual college, but it is for those students at EPFL. They have no choice but to enforce academic honestly rules, given that real students at a real institution are involved.

It makes sense for EPFL to use the course lectures and quizzes to help the students, but it makes absolutely no sense to use it for the grading part. It is trivially easy to make a second account, take a quiz, and just answer randomly to be able to see all the correct answers. It's absolutely absurd to not have a totally independent grading system.

The quiz questions in the course videos do not contribute to the grade for this course. Only the programming assignments count.

Re: Coursera's Functional Programming - Cheating Discovered

#170

Post by Cope »

eneveu wrote: post #7

When discussing online courses, a recurring theme is that of the "value" of online certifications, versus "real life" certifications. By taking action against cheaters, Coursera is trying to preserve the value of their certification. I wonder how it will affect it. Do people really care about online certifications anyway? For me, the main advantage of the weekly assignments, is that it stops me from procrastinating,…

In my eyes, the only benefit of the certificates is that you have something to show for your time and effort; something that proves you're willing to take the time to better yourself. That being said, it's not much.
Post reply on HN