Hacker News new | comments | show | ask | jobs | submit login
How do you stop online students cheating? (bbc.co.uk)
24 points by ishkur101 1845 days ago | hide | past | web | 33 comments | favorite



The simple answer to the question, "How do you stop online students cheating?" is that you don't. You discard the idea of "cheating" altogether.

In days not so long past, rote learning was important because, in many professions, there was no way to access the source data within a reasonable amount of time within the arena where that information would be useful. A trip to the library in the middle of the workday was generally not feasible. Pausing to locating and consulting the correct medical textbooks in the middle of an operation was dangerous and could have proven fatal to the patient.

Contrast that with the current state of the world where, in the vast majority of cases, the answers to many questions and the solutions to many problems can be located within minutes or even seconds from a smartphone or tablet. Combine that with widespread wireless broadband access in the developed world and it becomes obvious that not only is rote memorization of many facts, figures, formulae, and so forth not needed, it may actually be counterproductive.

I say counterproductive because, due to the way university-level education works much of what you are learning, especially in technical disciplines, is obsolete by the time you learn it. That's not a swing at formal education, mind you. I'm not asserting that what is learned at a university is not useful. I'm asserting that by the time you memorize that information it is highly likely that some other development has arisen which either invalidates or supersedes what you just recorded.

What I believe would be more productive than online universities worrying about cheating would be to more aggressively time-limit their exams. Doing so would at least validate how quickly students are able to gather and synthesize information to arrive at correct conclusions - a skill set that will serve them well even in today's rapid-access-to-information world.


> Doing so would at least validate how quickly students are able to gather and synthesize information to arrive at correct conclusions - a skill set that will serve them well even in today's rapid-access-to-information world.

That only works if you can prevent the student from simply hiring an expert to take their tests for them.


You don't need to stop the cheating if there is no reason to place value on the outcome in the first place. It seems the deeper problem is that marks have come to have meaning outside of their intended purpose: To provide a gauge to see your own level of understanding. Eliminate that and the purpose of cheating vanishes. If you score 100% on a test, but walk away knowing you have no knowledge of the subject matter, you've only hurt yourself.


This. There would be no reason to cheat in school in general if it was not for the fact, that nobody cares anymore about "knowledge" - grades are the points that get you to better high school, better university, better job.


But you haven't only hurt yourself. You've also (potentially) hurt the company that hires you based on the degree, or whatever qualification, you have - sure, you can say it's their fault for hiring based on it - and you can hurt the establishment that gave you the qualification as people lose faith in them.


That is my point though, if you remove the expectation that the grades mean something, then there is no reason to cheat in the first place. The problem is that we uphold the establishment as having some kind of meaning, not that people are able to game the establishment.


I think it is the wrong question. As Seth Godin says in his talk at TEDxYOUTH https://www.youtube.com/watch?v=sXpbONjV1Jc "if something is worth remembering it is worth looking up." Tests should be open note, open book, all the time. If you can cheat on the test just by googling the question, the right question isn't being asked. It is time we ask our students to start thinking critically and not just have them bubble in memorized facts on a scantron.


Tests are supposed to measure understanding of the subject material.

Therefore tests should be open book iff open book tests are better at measuring understanding than closed book tests.

Are open tests better? I remember from my university days that you can simply look at questions in the book that look sort of similar to the question on the test and apply the necessary substitutions. With open book tests it's remarkably easy to create an answer that looks intelligent about a subject you don't understand at all. With closed book tests this isn't so easy.

Even trivia questions can reveal fundamental gaps in people's understanding. For instance, if a student answers that dinosaurs existed 3000 years ago they must think within a frame of reference where 3000 years is a plausible answer (so no understanding of evolution). So it's a question that is easy to ask, easy to grade, objective, and one that can function as a great litmus test. When all tests are open book you can't ask these questions anymore.

So I'm pretty skeptical about the claim that open book tests are better, period.

As for the argument that people can look things up in books after they graduate, I don't buy that at all. The logical next step would then be to also allow students to ask questions on web forums during exams. Copy question to forum. Get coffee. Copy answer to test paper. After all, when the students get a professional job they may also consult web forums to get their work done.


If it's worth looking up multiple times, it's worth memorization.

The reason is that being able to recall at will have a much lower latency than a google search or tabbing through your iphone or android device.

People disparages memorization as mindless, but that's only if you memorize mindlessly about stuff you don't need to know. You can memorize smart, such as your multiplication table.


Conversely, if you actually find yourself looking something up multiple times, won't you naturally start to remember it?

Our tests hark back to the days when to "look up" something meant going to the library and flipping through an index. Internal recall will remain faster than external look-up for some time yet, but as the balance shifts the current setup will appear increasingly archaic.


This is only true if we don't use technology to improve our mind. Some people like to believe that the computer will become oracle, but some human beings takes advantage of the computer in the opposite direction to improve their ability to recall information.

For example, human memory follows the law spaced repetition and forgetting curve(our ability to retrieve information exponentially decay). We can take advantage of this by remembering or exercising our memory at the last possible to keep our knowledge. That mean we can efficiently review information as needed. Moreover, as we constantly review information, we only have to repeat them later and later in the future. In pyschology, spaced repetition is one of the least taken advantage discovery. Only in the past two decade or so, did we start using computers to speed up the process of remembering and learning.

That's only by taking advantage of human biology. it's possible in the future that we will enhance the speed and computational ability through bioengineering and neuroprosthetics. More over, the propagation of information is limited by the speed of light. So it's likely those with memory at close at hand will have an advantage over people's whose memory is far away, hence longer latency.


I can remember walking into an exam that was open-book open-note, but I didn't realize that before hand. I studied my balls off memorizing all the stuff I thought I would need to know. When it came down to it, I got one of the highest grades in the class which I attribute to not peering through my notes the whole time I just using them to grab a few quotes from each part. I was also the highest person in the class during the test.

My other memories of open-note tests consist of turning notes in as well. I really liked this idea especially on early exams because the professors could see what you were paying attention to and give you better guidance as to where you shift your focus. Note: I went to a small liberal arts college.


> I was also the highest person in the class during the test.

Perhaps it was your substance use or general elevation that lead to your increased performance, not your dedicated effort to memorize the material beforehand?

Though if the test was conducted under a deadline, it stands to reason that you could do better with memorization. As some else noted, the latency is much lower with memorization, but tests do not really mimic the real-world so it is not exactly clear how the lower latency will impact your use of the information outside of a test-like setting.


Allowing a single page of notes can be good, as it forces students to summarise the subject.


My college calculus professor typically allowed use of notes on exams, which he justified as follows.

-- I kinda expect the guy designing a bridge to lookup the correct formulas and not expect that he memorized it correctly. (paraphrased, of course)


Quit worrying about it. All cheaters get is a grade, or certification or diploma. Worry about teaching people. Worry about people learning. Quit measuring people.

People will cheat, and maybe they are great scam artists and will be able to cheat through the rest of their life. Those few and the others who screw themselves by not learning aren't worth being concerned with. Spend (all) your effort on those who want to learn.


Awarding diplomas to unqualified individuals degrades the value of the diploma. Schools can't afford to risk their reputation; they must assess students in some way.


Diplomas are already degraded and degrading.

Schools are worried about reputation and their profit. My point is they should be worried about teaching.


How do you stop offline students from cheating then?

You can't have exams for everything - for example creating or extending a big project is best examined by labs and/or a bigger project. If you have the same lab series for everyone students can simply copy the solution from other students or ask someone for "help" where the help consists of explaining exactly what to do. I know I've been guilty of helping my friends a bit too much.

Organize groups? Some students may coast on the skills of others. We had a programming project assignment where I wrote 95-100% of all code. In the beginning we split it around 50/50 but it ended up with me writing it all anyways. In another course we wrote a game in Java where my friend rougly wrote 1-2 classes or something. I know it's unfair, but I didn't really care - I like to code and I didn't really have to put in any effort anyways.

Cheating may be easier in online courses, but trying die hard to stop all kinds of cheating is bound to fail, cause it's impossible.


Two things which I remember having a significant impact on the amount of 'collaboration':

Marking against the other members of the class. I remember one project where the performance of the code was under test and the fastest 30% got an A, the next fastest 10% got a B etc. There was little incentive to help anyone else beyond the most naive implementation.

Group projects, avoid the problem of collaboration by openly encouraging it. Also has the benefit of improving the real life skills everyone will need. Allowing members of the group to rate other members makes it obvious when someone has simply ridden on the coat tails of the other members of the group.


Those are two things I remember hating every time I saw them implemented. In my experience, members that ride the coat tails of other members of the group always get far more credit than anybody else in the group gave them.


I would have said the same during my original undergraduate work (at 18-19). However, I ended up working full time and taking night classes with other working adults. From then on, there was never a coat tails problem. We would actually have the problem of people wanting to do too much rather than people not wanting to do anything.


Yeah, that is true, but I think that is probably a valuable life lesson!


So was being bullied - my point was that using group projects as a method to prevent cheating is much less trivial than I think you were implying. Encouraging collaboration and communication is a good thing, but in my experience, there is still plenty of cheating.


You can't even stop offline students from cheating, so what hope is there for online students?


You can't. Make people travel to a controlled place for examination if you need the certification, teach using online tools.

The real cost of University(specially in Europe where University is affordable, not like in the US) is living abroad most of the year. Making it 1% offline and 99% online is a good compromise.

Another possibility is asking your students real innovative work and make copying a must!!(organize people in working groups like in real life).


I think it comes down to one statement: Life is open Google, so school should be also.


You can't. Always going to be ways to cheat. Wouldn't be that hard either. Webcam/screen share? Just have another computer hooked up out of view behind your monitor with all the information on it.


You can't. The only way is to have in-person proctored labs and exams. Or they can wait for a rigorous job interview and fail in Real Life.


Another application:

How do you stop remote workers from slacking off?


Paying them in milestones basis.


That helps only if remote workers are able to motivate themselves sufficiently without supervision. Some people fail to work remotely because without the boss to keep them in line they fail to motivate themselves.


I'd say you can't. People are always going to try to game the systems and that's why a certification given to you in person is always going to be more credible than something gained online.

I think ultimately the rigor for actual "testing" is going to have to come from whoever is accepting applications from these types of degree holders, sure you might be able to apply to graduate school with them but you'd probably have to sit an additional test that "standard" degree holders wouldn't or in the case of a job interview there would have to be a bit more rigor in testing the applicant.

Cheating on an online course right now seems so...pointless, the only reason for them at the moment is to enrich yourself and learn something but if people start accepting them in the same way we accept a standard degree we'll see a massive drop in the value of all degrees.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact

Search: