Online study groups: Threat or menace?

A good excuse for some Tom Lehrer (from ‘Bright College Days’):

To excuses we fibbed, to the papers we cribbed, from the genius who lived down the hall.

It’s always been possible to get help (ranging from an explanation at a slower pace than the lecturer to all the answers) from another student. The main limiting factors are the obvious one: people get tired of freeloaders and if you don’t understand the material you’re stumped in exams. i don’t seen how online discussions affect either of those things.

Ah, magistra says the same as I did. Your point about freeloaders is interesting: perhaps online help to many students at the same time makes people more willing to help before freeloader irritation steps in.

And it allows students to find the tutor most likely to help freeloaders, such as this Avenir.

Before the rise of the Internet, a large amount of what computer science students learnt about programming and IT was learnt from the people (students, faculty and tech-admin staff) sitting at nearby machines in the university’s computer lab. You had a problem, you asked the guy (it was always a guy) next to you. After the rise of the Internet, a large amount of what commercial computer people learn about programming and IT is learnt from online discussion forums and (now) blogs and wikis. You have a problem, you search for the answer online or you ask on a forum.

It seems perverse that, after the Internet, computer science students should be prevented from learning in the way they used to learn before the Internet and the way they will learn after they graduate, by asking other people. Indeed, given this mode of learning is perhaps the single most important way to acquire technical information in the commercial IT world, it is worse than perverse to not encourage students to learn in the same way while still at college. How to search and ask online should be a compulsory key skill for CS students, and they should be examined on their ability to acquire information this way.

@magistra and @greatzamfir: the freeloading angle is interesting, precisely because what online systems are fantastic at is *reducing* the costs of freeloading. (I made this argument eight years ago, writing about why P2P systems weren’t collapsing from Tragedy of the Commons issues: http://icanhaz.com/freeloading )

In Avenir’s case, I can think of three factors that might have freaked out the good folk at Ryerson: First is the visibility. As with all social codes, the difference between “everyone does it” and “everyone admits that everyone does it” is the difference between a policy that is easy to enforce and one that is hard to.

The second effect is the beating heart of _Here Comes Everybody_ — by lowering transaction costs, Facebook makes it much easier to advertise, discover, join and benefit from such group work online.

The third one though, which I hadn’t considered ’til now, is that because the conversation isn’t just simpler to join, but also simpler to index, search for, and share,as with the way conversations about High Dynamic Range photography on Flickr became archived textbooks. http://icanhaz.com/flickr_hdr

And given that Avenir was facing *146* charges in his disciplinary hearing — one charge for each member of the group (!) — that one seems plausible. This I think is the answer to zamfir’s ‘is this so different’ question — yes, a single study group that can have 146 members and leave a permanent public record *is* different.

My bias is the same as @mary’s — this change isn’t minor and it isn’t optional, so the academy should just roll with it, but even with that bias, I do think a freeloader-tolerant version of a study group is kind of a big deal.

I agree with the others that homework, especially in chemistry, should not be a significant evaluative tool at the college level, and therefore, students should work together as much as they want.

I agree that this is different than old-school collaboration around the library table, at least in terms of degree. But to give students some sort of assignment where they can just ‘look up answers’ in some indexed online discussion and thus freeload themselves into a good grade on it indicates that the faculty member is failing to devise adequate learning tools. As Mary suggests, freeloaders will rarely put the necessary time into an online (or offline) discussion to actually become part of the collaborative learning process. In the end, if you give a appropriate assignment and evaluative tools, freeloaders take care of themselves.

When solutions are easy but the problems are hard, we are setting our students up for a harsh time when they encounter a world where the solutions are not necessarily easy or elegant. The two main reasons for complicated problems with simple solutions are: 1) to weed out the ones who “get it” from the ones who “don’t get it” and 2) to be easy to grade. 1 is (arguably) selecting for the wrong thing, and 2 is laziness.

In a collaborative world, pretending that cooperation is bad in any discipline is the wrong message to send.

The problem here is that in the STEM disciplines we’ve tried to bribe students into virtue: we know that it will benefit them if they work through problem sets, so we give them some credit towards the course grade if they do so.

And there’s probably some benefit to participating in a Facebook study group… Maybe the way to make this work is to give extra-credit to give “participation” credit for actively participating in on-line groups, just as faculty often do for discussion sections.

I think Uri Treisman’s work on collaborative learning is relevant here. For technical subjects at least, working through problems in small groups appears to be more useful than studying alone, especially for those students who are struggling with the coursework.

Ummm… as far as I can tell, this whole discussion is missing the point. Reread the dean’s comment: “*if* work is to be done individually and students collaborate, that’s cheating”. If, if, if. This isn’t a discussion about the optimal way to assign homework, or whether students should work in groups or not. This is about following the rules. Not all work is to be done individually, but *this* work happened to have been assigned to be done individually.

Students are allowed to work in groups on some assignments, but not this assignment. If students can’t complete this themselves, they have access to the university-approved fora that Avenir lists: mentoring programs and discussion sections, probably office hours for the professor and/or teaching assistant as well. (It’s not clear to me whether independent tutoring would be acceptable here — probably you’d have to work that out with the professor).

Facebook should be treated the same as other study groups, but this assignment was one where study groups were forbidden. I will take issue with the University’s response of focussing on Avenir, though. I don’t think the “head” of a Facebook group should be punished any more than the rest of its members, and if the school doesn’t want to punish everybody then they shouldn’t punish anybody.

One final point — in every class I’ve taken where study groups were allowed, the professor has given us firm rules that there can only be a maximum of, say, four people per group, and that we have to write out a list of everyone we worked with. A Facebook group would almost certainly violate these kinds of rules.

sk —

I’ll play advocate against your devil:

Does education aim for students to learn facts and information (know-what) or does it aim for students to learn to do something (know-how)? At least in Computer Science and the other Engineering disciplines learning know-how is as important, and arguably more so, than learning know-what. In the complex world of technology we now live in know-how almost always requires collaboration. Hence, people who learn know-how also need to learn how to collaborate.

Collaboration is not usually something simple (even asking questions on online forums has its own etiquette), and people have to learn how to collaborate. If university does not teach its students how to collaborate, then it is failing its purpose of preparing students for life outside academia.

OP and 19: of course. But it’s not an either/or. Yes, a good engineer or whatever needs to know how to collaborate and ask other people for information and advice. But he needs to know other stuff as well.

Apart from everything else: what if the “collaborative” freeloader graduates and ends up in a small work group entirely composed of other freeloaders? Who’s going to do the brainwork then?

And the suggestion that collaboration is all you need to learn is lunatic for another reason; it would imply that degrees are fungible. If I can coast through my history degree by being a good collaborator, doesn’t this suggest that I would be just as good at, say, computer science?

Now that I no longer have an opportunity to try this in practice (having left the academy for good), I find myself trotting out an old idea more and more often. I would rather see these spurious “honor systems” die, than continue training young people to be some kind of rugged individualist over-specialized hermits that most of the academic infrastructure promotes.

The utility of a college education, whether an undergraduate “designed” to “demonstrate” “basic skills”, or a Ph.D. “designed” to “demonstrate” the recipients’ “ability to follow through on a project”, is nearly zero for me. I work in research, I hire good and smart people—and increasingly I find myself disqualifying over-educated under-experienced applicants.

It’s not clear to me whether independent tutoring would be acceptable here

It seems to me that this reveals the main problem with a no-unauthorized-help rule: students are often not going to know when they’re in the clear. (Is hiring a tutor different from buying the Cliff Notes? What about buying a senior’s old exam off them? A senior’s old notebook? Their old textbook with underlining and notes in the margins?) We have a strong practical interest in good grades–namely, that our future employers, financial-aid suppliers, and parents all measure us by them. This outweighs the somewhat abstract concern of following exactly the rules that were laid out at the beginning of the semester, and so most students are going to do certain things anyway, on the assumption that they won’t get caught.

Which is a good assumption. Although I’ve personally never used any proscribed forms of assistance on academic work, I’d say about 90% of my friends will admit to having done so (mostly in high school), and to my knowledge none of them has ever been caught. So setting a rule like “no study groups/homework parties/etc.” has nothing like the intended effect–these things are part of the natural order of things in college, like underage drinking. (Which is, if anything, easier to prevent.) To the extent that you make such rules, primarily you’re just delegitimizing yourself as a rule-making authority.

Interesting questions raised here, a few reactions.

My response would be something along the lines of “a”: We need to recognize new tools and social contexts in which students may be doing their learning outside of the classroom and figure out how to proceed in light of all that. I don’t see how “b” can work for someone who really cares about learning on behalf of students. “c” is naïve by thinking that “this must be stopped” could be an option.

As others noted, the situation is different here from what was possible before, not only in magnitude and ease as pointed out by Clay, but also in that anonymity is an option in ways it wasn’t before. That’s less the case on Facebook, but other platforms such as Yahoo! Answers allow it.

While I appreciate that students can learn a lot in group study settings, there are tons of opportunities online where one can get a quick response to a question without putting in any work whatsoever. One example is Yahoo! Answers where people ask for help with homework all the time and often get it.

Regarding the freeloader problem, while it is the case that certain communities won’t tolerate it for long, again, a community like Yahoo! Answers has less of a problem with it. Not only are there thousands of potential respondents (so for everyone to recognize a user as a freeloader would take a long time), they also have the potential of getting something in return: having their answer picked as the right one and thus gaining more points on the system. (The points mainly have a reputational component there, apparently enough to motivate quite a few participants.)

I feel fortunate that the types of classes I tend to teach don’t require simple answers so I don’t have to worry about some of this. That is, the assignments I give are not ones where there is one correct response. Rather, they are assignments that require students to do some original work (e.g., find a relevant article and summarize it where each student has a different project so the relevant article is unlikely to be the same across students). Of course, it’s still possible to depend on others for help, but it’s much less easy to copy that kind of work. But this is not equally easy to do in all types of courses. Plus it definitely places more of a burden on the instructor as checking assignments takes much more effort.

the problem here is graded homework that contributes to a final grade.

Networked collaboration will require a switch to Outcome-based education. The traditional model cannot survive pervasive social networks.

the problem here is graded homework that contributes to a final grade.

Bingo. Without that, there is no freeloader “problem”; the “freeloaders” would only be cheating themselves.

I always told my students that I wouldn’t collect or grade problem sets but that if they didn’t work on them seriously I could pretty well guarantee they’d do poorly on my exams. If they choose to ignore good advice, oh well.

Re: #35. While there’s a lot to that, I’m not sure it’s the whole story. How does one learn to appreciate a style of music that one may not like at first. You can observe someone else listen to it, but you cannot observe the cognitive processes involved in enjoying it. They could try to explain why they like it, but conveying musical experience in non-musical language is notoriously hard, and formal languages for describing music (i.e., notation systems) don’t really convey enjoyment, and don’t seem to be how people come to learn to appreciate music.

Having just completed a post-grad Master’s, can I please put the boot into group learning? All it gave us was trouble with plagiarism and freeloading, plus some useful experience on what it’s like to have to work with idiots (AKA the ‘real’ world of work)

This wasn’t just my opinion; most of the class agreed.

Neon, I’m not talking about writing criticism. I’m talking about styles of music – free jazz, gamelan, modern classical – that I did not like when I first heard them and now do. And it’s not primarily social, as I got into them mostly independent of a peer group of like interest. I’ve never written criticism of such music, though, and am not sure I could.

More broadly, my point is that I don’t think *all* learning is simply social – that mirror neurons are the whole story. I think we also learn by “doing” – by grappling with problems, even in solitude and even without an explicit model. Suppose someone learns through diligence to solve Rubik’s Cube: that have learned something through experimentation and analysis, no? They may have had the time for this as a consequence of being in solitary confinement.

Since even social learning requires attempts to perform rather than simply observe the target behavior I think it fair to say that this component of “doing” is almost always present. And that the attempt to balance the social aspects of learning and the aspects that come from making attempts at the problem directly is a lot of what is at stake in the problem at hand. Not that I expect any of what I just said is a particularly novel or likely controversial assertion; but that is what I mean.

This is factually incorrect. Humans, indeed all primates, do not learn individually. We learn by observing another perform an activity. Our mirror neurons fire and we imitate the other’s performance and repeat it until a corresponding circuit in our brains is created and we can perform the activity on our own.

If this is true, how does humanity do anything new? How do you think the humanity ever learnt to ride a bike?

Humans have the added complication of language so much of this mirroring takes place in a sort of linguistic space but it’s still the same thing. Behaviors are observed, mirrored and then repeated until the appropriate neural circuit can be physically built.

Actually they aren’t. There is far more to learning than mirroring and repeating behaviours, as artificial intelligence researchers have found out. For example, take washing dishes – an activity most of us have learnt. Say we learn to wash dishes by watching our fathers wash dishes at home in the kitchen. But then, we grow up and leave home and must wash dishes for ourselves in a different kitchen. Most people who wash dishes are perfectly capable of washing dishes in a new foreign kitchen, but if we merely imitated dad’s performance we would not be. For example, in my childhood home, the dish detergent was kept on the windowsill in front of the sink. In my first flat, there was no such windowsill and we keep the dish detergent to the side of the sink. We also brought a different detergent. So Dad, to add detergent reached up and placed his right hand around the yellow bottle. I, to add detergent, had to reach to the left and place my hand around the green bottle. If I merely imitated my Dad’s performance, I could not have made that adjustment. But I did, without problem. Another example, some dishes require more scrubbing to get clean than others. And the dishes in a home change over time, and furthemore the dishes we have as adults are often different from the dishes our Dads washed at home. Yet people can wash new dishes.

To really learn to wash dishes requires that the learner forms a conceptual model whereby they can separate out the important ideas from site-specific details. That’s what makes artificial intelligence so difficult, it’s one thing to create a computer that copies behaviour, it’s another thing to make a computer that understands washing dishes and can apply the basic concepts in new situations. But most humans pick that up fine.

Tracy
If this is true, how does humanity do anything new?

Discovery is done through trial and error and inductive reasoning. Riding a bike isn’t that different than walking. You are applying your ability to balance to a new situation. A lot of discoveries are made by applying old tools to new circumstances.

If I merely imitated my Dad’s performance, I could not have made that adjustment.

But we don’t merely imitate, we construct an internal model and we are perfectly capable of adapting learned behaviors to novel situations. I have a mental idea of my apartment that I confirm through my senses. I don’t even need to look for the dish soap, I know exactly where it is. So “get more dish soap” is a trivial problem. I do not know the world and I don’t perceive it directly. All I know are my ideas about the world and I can discuss those ideas just as easily over a telephone or the internet as I can in person. Learning takes place in the realm of ideas. How that happens, solving a hard problem alone or in collaboration with others, doesn’t seem to me to be that important as long as it takes place.

An interesting read. But the thing is, this has been going on for years. Think back to the times when you got help from parents etc on that important piece of coursework during high school. An issue has been made of this only because it is more public and obvious. Your article didn’t mention whether the piece of work in question was contributing to the students’ grade for the course? I’d imagine not, with a score only out of 10? If it doesn’t count, then it’s a massive over-reaction on the part of the uni. However, if it counts as part of students’ grades, then I can understand their position. The silly students should have made the facebook group private! Doh!