No Grading, More Learning
May 28, 2010 8:18 AM   Subscribe

"It would take a lot to get me back to a conventional form of grading ever again." Cathy Davidson, an English professor at Duke, teaches a seminar in which final grades are determined by fellow students. She writes about the experience in Inside Higher Ed. (Thoughts by Duke faculty about the philosophy of grading previously on MetaFilter.)
posted by escabeche (58 comments total) 13 users marked this as a favorite
The collaboration that she touts as a primary benefit is not impeded or precluded by teacher-assigned grades. The clarity that she prizes in "how to get an A" is easily available in a teacher-assigned grading (it's called a well written syllabus). So, other than superficially transferring responsibility to the students (I say superficially because the professor still gives apparently considerable "constant commentary) I'm not seeing huge benefit.

If your student's are likely to "cynically" ask "how do I get the prof to give me an A," why wouldn't they ask "how do I get my classmates to give me an A?"
posted by oddman at 8:32 AM on May 28, 2010

"So how did it work? Davidson, the Ruth F. Devarney Professor of English, said that of the 16 students in the course, 15 already have earned an A and she expects the remaining student to soon finish an assignment that will earn an A as well."

That's special.
posted by MarshallPoe at 8:34 AM on May 28, 2010 [1 favorite]

> That's special.

My modest proposal moves one step closer to becoming a reality!
posted by The Card Cheat at 8:40 AM on May 28, 2010 [2 favorites]

Peer assessment isn't that new. I was subjected to it (and participated in it, obviously!) as an undergrad in 1993 or 1994. It was very informative; each student marked two or three other students' work, and you got to see the ideas others had had. The course was HCI (Human Computer Interaction), and the task in question was an interface design, so actually seeing other peoples' attempts was a real part of the learning experience. The mark scheme was clear, the objectives were clear, and the insight into the grading process was fascinating. I think the prof went over the marks in cases where there was serious discrepancy on student grades.

I used a similar scheme when I was teaching intro to cognitive science a few years back - one of the pieces of work was a presentation, and the students had to provide the marks and feedback on each others' work, with grades being given as an average. Again, clear markscheme, clear objectives, student ownership of the process... == big win. I didn't save myself any work, as I sat in the talks and filled in the marking form as well (as a fall-back, which was never needed). But I think the pedagogical win was clear.
posted by handee at 8:40 AM on May 28, 2010 [1 favorite]

I had a conversation last week with my (summer) research supervision about how the cliques in our first-year (law) classes worked: there are two sections of first-year, he was wondering if the cliques were intra-section, if the people in a given clique sat together in class, etc.

What I'm saying, is that cliques exist in school. And there are outcasts and leaders. I have often been that outcast. And I would be terrified, as in afraid to get out of bed scared, to be the outcast in a class like that.
posted by Lemurrhea at 8:40 AM on May 28, 2010 [19 favorites]

A high school teacher of mine tried this once with an assignment he gave us. We were to split into groups, do the assignment, then the class was to award us a mark. And guess what happened? All the popular people gave each other good marks, the "brains" in the class got good marks (because they always had before), but those who struggled or who had bad repuations flunked. To the point where I actually stood up and defended one guy in my group who got a bad mark despite the work he'd put in.

It was instructive to see the extent to which the decision on what grade to award was not driven by merit but by reputation.
posted by LN at 8:42 AM on May 28, 2010 [18 favorites]

"...students had to do all of the work and attend class to earn an A."

No. No no no. Yes "peer pressure" can work. Except I've got ten bucks that outside of a seminar class, peer pressure works more toward covering your classmates' asses than "crowdsourcing" of grades.

"She added an individual comment on every student essay..."

What? How is this even worth noting? You're a professor. Your students are supposed to get feedback from someone who holds their position because they, ostensibly, are more skilled and knowledgeable than they are.

Look, this is clearly an experimental class so an experimental grading structure makes sense. Expanding this to "regular" classes in schools that aren't incredibly competitive? God no. I'm an English major in a not-particularly-selective university in NYC. A few years back, I took a 3rd-level (one under seminar, two under grad-level) Modern Poetry class. On the last day, we were asked to read selections from our term papers. Half the class - roughly fifteen people - made it clear their papers did not contain anything resembling a thesis. These are students who have been taking English classes of rising difficulty for at least two years. Are these the individuals I would want want to depend on to improve my work and further my abilities? No. Never. Yes, peer-reviewing is a great system. I believe everyone may have something to say to help, but to put their input on the same level as the professors? A thousand times no.
posted by griphus at 8:43 AM on May 28, 2010 [3 favorites]

Also, I'd like to cite Muphry's Law (distinct from Murphy's Law) w/r/t all the horrible grammatical errors in my comment.
posted by griphus at 8:49 AM on May 28, 2010 [2 favorites]

So so very many problems! That said, I could imagine using youtube views for grading film projects, assuming you'd some viable system for preventing people from advertising.
posted by jeffburdges at 8:49 AM on May 28, 2010

It's interesting. Without taking part in the class and actually being part of the discussions and critiques, it's a single POV summary. My summary of her summary: students work harder when judged by peers (in part because the judges become the judged the next week), and when given the opportunity to improve their grades.

It sounds like it works because 1) the prof sets the standards and is involved throughout, 2) there is a chance to better the initial score, and 3) the class was tiny. 16 students? I went to a college with small classes, but they were generally twice that size, and classes that small were canceled for lack of student interest.
posted by filthy light thief at 8:51 AM on May 28, 2010

And I would be terrified, as in afraid to get out of bed scared, to be the outcast in a class like that.

That was my initial thought as well. Teachers and professors should feel at least some amount of professional objectivity in grading but fellow students are going to feel that a lot less. This sounds like a formula for groupthink.

Also, in the ideal classroom there are no fellow students, there's just a teacher and a learner at the opposite ends of a log.
posted by DU at 8:52 AM on May 28, 2010 [1 favorite]

Since I already have structured my seminar (it worked brilliantly last year) so that two students lead us in every class, they can now also read all the class blogs (as they used to) and pass judgment on whether the blogs posted by their fellow students are satisfactory. Thumbs up, thumbs down.

Based on the amount of people I went to school with who did the absolute minimum amount of reading required to get a decent grade, my guess is that this system will result in a lot of "tl;dr thumbs up" grades. Even in writer's workshop-type courses I took where literally the entire point of the class was reading each other's work and offering feedback, around half of the students really obviously hadn't actually given the material more than a cursory glance. If an instructor is incompetent at grading and offering feedback, there are systems at most universities to report and deal with those issues, whereas if your grade is determined by disinterested peers who may not even have the skill required to evaluate your work, what is your recourse?
posted by burnmp3s at 8:54 AM on May 28, 2010 [2 favorites]

I can't imagine this works really well in a writing-intensive course. I had a solid two years of writing seminar courses (concurrent with and drawing content from my anthropology/history/culture courses) and we often peer graded. While I can believe that crowd-sourced grading is relatively accurate in placing student work on a curve of quality, I bet it does little to help shift the whole class towards better work unless the group is already fairly high performing. My peers would often pick out little grammar mistakes or suggest minor rewordings, but students who aren't already good writers will not spot a buried thesis, flaws in presenting evidence, etc. That's why you have TAs and profs grade that kind of work. To me this explains why all of her students are getting As- the bar for writing quality is wherever the class mean is, not where she sets it.

I guess there may be some value in her quantity and frequency over quality approach. Writing 1000 words a week is good practice no matter what, and in most cases a single instructor could not keep on top of that much grading. I like that she gives the opportunity to revise. For many people writing more often and getting some (any) feedback will be better than writing one huge paper, but I don't think her approach is preparing students to write dissertations or other long-form challenging work.

But, I think handee's experience in the HCI course shows exactly the kind of context where this approach, blended with prof feedback, would be better than instructor grading alone.
posted by slow graffiti at 8:54 AM on May 28, 2010 [2 favorites]

Also, in the ideal classroom there are no fellow students, there's just a teacher and a learner at the opposite ends of a log.

I disagree, at least within the liberal arts and especially within literature and philosophy. Other students bring up problems I didn't realize I would/may have in comprehending the subject matter. Classes on literature are dull as hell without discussion and interpretation coming from various individuals, moderated and illuminated by the professor.
posted by griphus at 8:57 AM on May 28, 2010 [1 favorite]

slow graffiti: To me this explains why all of her students are getting As- the bar for writing quality is wherever the class mean is, not where she sets it.

I give this answer a grade of B. I would have given it an A, but I only give those to my friends.
posted by three blind mice at 9:02 AM on May 28, 2010 [3 favorites]

I think it would be a great idea if it weren't so goddamn clearcut.
posted by digsrus at 9:08 AM on May 28, 2010

You used to sleep with the professor to get a good grade. Now you have to bang everyone.

Sign me ... up?
posted by Cool Papa Bell at 9:09 AM on May 28, 2010 [6 favorites]

I'm currently avoiding grading by reading MetaFilter....
This system will not work. Students hate to critique fellow students. There is no upside to marking anyone low. Generally, everyone will give everyone else an A, no matter what.
posted by cccorlew at 9:12 AM on May 28, 2010 [3 favorites]

Also, in the ideal classroom there are no fellow students, there's just a teacher and a learner at the opposite ends of a log.

Ouch. You must be Mr. Anderson, from Grade 7. It was an "experimental" school, no doors on the classrooms, limited curriculum, all manner of self-directed study. But every now and then we'd be bad. The collective "we" that is. And the punishment was lectures from Mr. Anderson. He would speak slowly, monotonously. We would shut the fuck up. And at the end of the hour, we'd be tested on pretty much everything he had just said, so you had to pay serious attention.

I can't even remember what the subject was. I just remember the pain of having to sit still and learn from a teacher at the opposite end of a log.
posted by philip-random at 9:17 AM on May 28, 2010 [2 favorites]

Now you have to bang everyone.

I'm sure blowjobs would suffice.
posted by philip-random at 9:18 AM on May 28, 2010 [2 favorites]

> Students hate to critique fellow students.

This. My undergrad was in film studies and any time any of the profs attempted to elicit any sort of feedback or criticism by fellow students it was a complete waste of everyone's time.
posted by The Card Cheat at 9:19 AM on May 28, 2010 [1 favorite]

In the courses I took with this method (like the HCI example I gave above), all work was anonymised, and the class was fairly large (110 students I think). So I truly had no idea whose work I was dealing with. The prof claimed that the admin overhead of assigning work, checking that no fuckwits had accidentally signed their piece and therefore de-anonymised it, collecting and collating the grades and then verifying the cases where there were discrepancies actually took longer than marking the thing himself would have done. But he kept running it - as it was a useful learning experience.

In the courses I've taught with this method the peer assessed portion was a presentation - so no option for anonymity. But the assessments seemed very fair to me, and the comments were insightful and in many cases more detailed than I would have been able to give myself: getting feedback from 20 people on a talk is much more useful than getting feedback from one, however more qualified the one person is. No gaming the system, no giving everyone As. Maybe I just had a couple of years of good classes?
posted by handee at 9:23 AM on May 28, 2010 [2 favorites]

I did a course in contemporary sociological theory in the early days of my BA that was partly peer graded: there was a term paper the prof marked and a participation component (contribution to discussions and two presentations) that the class decided on. so the last day of classes all of us, the prof and the students, sat down and decided who would get what marks.

there was very little structure to the process. it was a very long and awkward hour and a half, with a great many uncomfortable silences in between people suggesting a grade. some people were given a mark by someone else in the class (I think everyone who got an A was given it by someone else), but most people said what mark they thought they deserved. it was a smooth process, for the most part. people would nervously suggest a mark and the class would quietly assent. there was one guy I remember who gave himself a B -- to which the class protested, thinking he deserved a higher mark, but he defended himself. he said, "I gave one good presentation and one that bombed, so this is fair." we argued, the presentation that bombed was on a tricky issue to do with gender and he'd had the courage to try a very unorthodox approach, surely that was worth something. "nope. I took a risk and it didn't pan out. I accept that."

it was a very honest, soul-searching moment for all of us. making it a communal process made it mean a lot more than just getting a mark. it remains one of the best learning experiences and among my fondest memories of my undergraduate days.

(on preview: it was for the participation component and I think that might have had something to do with how well it worked. slow graffiti's points are good; I'm not sure that it would have worked as well for the paper.)
posted by spindle at 9:25 AM on May 28, 2010 [1 favorite]

This seems like it would lead to a Prisoner's Dilemma situation. There's a pressure to cooperate and hand out "A"s so that the others will reciprocate. Defecting early and giving a lesser (even more honest) grade will just lead to your own work being judged more harshly than perhaps it deserves. The fact that the whole class got an "A" seems to back this up.

Interestingly, this idea would be perfect if it was completely anonymous and, better, judged by students studying the same subject but not in that same class. Anonymity would obviate the social pressure within your peer group to automatically hand out good grades, and judging students from another school would hopefully eliminate handing out unfairly poor grades in an attempt to boost one's own ratings (say in a valedictory race or scholarship competitions based on GPA).
posted by Freon at 9:29 AM on May 28, 2010 [1 favorite]

I once had a student try to cheat in my class by duplicating one of my tests - which they had spectacularly and memorably failed - after solutions had been posted, filling it out, and "grading" it themselves, complete with comments beside the problems. Then they dropped me a note to tell me I must have recorded the score wrong in my grade sheet, could I take a look at the test and record the correct grade, please.

I wish I could share the results with you, but I can't do that, so I will share the approximate moral with you instead, even if it doesn't have the same punch: students who are struggling with material themselves do not have a good vantage point for providing accurate assessment of that material. What they have is something more like the worst possible vantage point.
posted by Wolfdog at 9:33 AM on May 28, 2010 [4 favorites]

In our college senior design studio, we worked in groups, and had to offer feedback to the profs in the form of peer review - I forget the exact format. There was one guy in our group, who I loved dearly, but he had a bad habit of running off to Mexico, hundreds of miles away, on Thursdays and not coming back until Monday. He failed. He deserved to.

As a cadet in a military environment we also had to submit peer evaluations at the end of each year. We called them fuck-your-buddy reports. I think I probably fared better in those reports than I deserved, because I carried my own weight and am not an asshole - but I don't think I was a very good cadet, to be honest. On the other hand, maybe that made me better than average. I don't know.

Peer review is probably no worse or better than other methods. Profs might like it because it displaces the onus of responsibility on them.
posted by Xoebe at 9:36 AM on May 28, 2010

A long time ago in a distant university I had a grad seminar with Francis Fergusson, a wonderful scholar. The first class meeting, he would hand out slips of paper and ask that we put our names on a slip and then put the letter grade we expected or hoped to get. In advance, he told us that whatever we put was what we would be given at the end of the term. Slips turned in, he would then say: "Let's now get on to what we are really here for."
posted by Postroad at 9:36 AM on May 28, 2010 [11 favorites]

I also have had horrible experiences with group work and peer editing in high school, so I understand the backlash, but I've also taught a freshmen writing seminar at Duke, and I think that Davidson is absolutely right to say that students (at least, students at Duke) work harder when they know their classmates are going to be reading their work. Students in English classes tend to assume that their written work is going to be read by the professor and only the professor--essentially, a private transaction. Anything you can do to remind students that writing is essentially public in nature (that is, it is meant to be read by people!) encourages them to generate something that matters to them and that they're willing to defend. I've also heard from other colleagues that assigning students to post on blogs or other public internet fora has had great results, for the same reason, I'm betting.
posted by duvatney at 9:40 AM on May 28, 2010 [1 favorite]

I think this scheme confuses peer with cohort.
posted by vapidave at 9:44 AM on May 28, 2010 [1 favorite]

I teach sections of intro sociology and of deviance with more than 400 students in each (more fun than it sounds), so how can Professor Duke's experiment have even the tiniest bit of even theoretical relevance to the reality of MY university classroom?
posted by ethnomethodologist at 9:48 AM on May 28, 2010

What would be very interesting is seeing this idea together with a curve. Right now our grades are belled to a B- (blech). Instead of discussing whether someone deserves an A, discuss how people ranked, with objective criteria only for the outliers / endpoints.

It would definitely get away from the problem of the scenario that happened, in that everyone got A's. But it'd be trickier, too - harder to claim a detached assessment.

Also, then I could set up March Madness-style brackets of my friends.
posted by Lemurrhea at 9:57 AM on May 28, 2010

Some of the other comments have encapsulated this remark, but I think it bears repeating. Absolutely the last thing that I would ever assume, something that I would never assume in the absence of overwhelming countervailing empirical evidence, is that the intellectual quality of the work is what is actually being assessed by fellow students (and that there are no confounding variables). This might work well for creative projects, but the thought of fellow undergraduates to evaluate, say, upper-level term papers in philosophy or sociology (or...or...) raises some questions, to put it mildly.

If I could guarantee that the grading was double-blind I might have fewer reservations. But in that scenario there is less opportunity for getting meaningful feedback or dialogue about your work.

Authority in grading isn't necessarily bad because it's authoritative. Authority can also help (but does not guarantee) consistency, accountability, and an ongoing relationship where the same person sees your progress over time. Any system that cannot convince me that those three factors are present is not one to which I would choose to submit my work unless I had no other options.
posted by mister-o at 10:04 AM on May 28, 2010 [2 favorites]

I am certain that my peers would be too intimidated by the brilliance of my academic output to award it the mark it deserved.
posted by tigrefacile at 10:06 AM on May 28, 2010 [2 favorites]

This. My undergrad was in film studies and any time any of the profs attempted to elicit any sort of feedback or criticism by fellow students it was a complete waste of everyone's time.

Ha, from the other end of the spectrum, our class was savage. So much so, that I had to mentally re-train myself on how to critique art afterwords so as not to start off with "Here is why your work sucks."

That said, this sucks. (lolz?)
posted by cavalier at 10:19 AM on May 28, 2010

so how can Professor Duke's experiment have even the tiniest bit of even theoretical relevance to the reality of MY university classroom?
posted by ethnomethodologist at 12:48 PM on May 28 [+] [!]

I wish I knew more about ethnomethodology to determine if this is funny or not. I think it may be.
posted by griphus at 10:28 AM on May 28, 2010

There's so much wrong with this plan, I don't know where to begin. But I'll start.

Duke is a private university. They're selling you an accredited degree and the promise of future benefits that degree will bring you. Your end of the deal (besides an outlay of cash paid directly to the institution) is an agreement to enter into an educational relationship -- i.e. you have to attend class and work, and you'll be tested and graded, and that grade will be attached to the degree we hand you.

Suddenly, a professor is changing the rules of that educational relationship. "I'm not going to fulfill the deal. I'm going to merely supervise while others fulfill the deal. Your educational relationship is with them."

It'd be like you paid me to make you a sandwich, and I took your money and said, "OK, now the guy behind you in line has to make it for you, and you'll make his. I'll just watch."



posted by Cool Papa Bell at 10:43 AM on May 28, 2010 [9 favorites]

In college I had an art history class where the professor and the TAs made the students do peer review of each other's work, and some portion of the grade was based on the peer review. I refused to do it or to hand over my work to my fellow students, because as I told the professor, a) I doubted the other students were competent to evaluate the work in any meaningful way; b) I wasn't getting paid to read and grade the other students' work, but the professor and the TAs were, and I didn't think I should have to do their job for them.

Yes, I was a bit of ass. But I think I was also right about it on both counts.
posted by jscalzi at 10:48 AM on May 28, 2010

So, instead of sleeping with the professor to get a good grade, now I have to sleep with the entire class?
posted by Slothrup at 10:51 AM on May 28, 2010

Peer grading is a terrible idea, in my opinion. I had to create and listen to dozens of group presentations in school, and they all were completely forgettable. I don't remember a damn thing that I was supposed to have "learned" from my peers, and the grades that the class assigned to its members were completely irrational. A drunken giraffe could have done better.

I don't want to listen to the frat boy next to me drone on about a complex issue and grade me when I drone on about a complex issue. Group work and peer grading are popular, but peer grading is actually much less transparent and fair than traditional grading from a professor.

What's to keep that Type A girl and her friends who want to get into med school or the dude who wants to be a lawyer from cutting you down a few notches as part of the "competition" in class? I'm a big proponent of doing away with grades altogether and going with detailed written performance assessments, but that will probably never happen.
posted by Despondent_Monkey at 10:57 AM on May 28, 2010 [1 favorite]

Obligatory "Duke Sucks" comment.
posted by brand-gnu at 11:02 AM on May 28, 2010

I'll buy that peer grading can be unfair if it's done wrong.

But I'm puzzled by the folks saying, essentially, "I don't want to bother evaluating my peers" — saying it's too hard, too boring, a waste of time, no fun. Lots of people don't want to bother writing papers either. That's hardly an argument for eliminating those.

Evaluating other people's work is a valuable skill. (Evaluating half-assed work done by morons is an especially valuable skill, as anyone who's worked in the lower ranks of management can tell you.) And frankly, I'm willing to be that most college graduates these days will be better served by that skillset than they will by a knack for writing English Lit papers. So as far as I'm concerned, it's completely kosher for a prof to ask students to cultivate that skillset.

So long as it doesn't lead to unfair grading, I'm all for it. And if it does lead to unfair grading, attack it on those grounds, not on the grounds that it's too hard or too boring to learn a new leadership skill.
posted by nebulawindphone at 11:21 AM on May 28, 2010 [1 favorite]

Despondent Monkey: It does happen in some places.
At Sarah Lawrence College Professors give detailed written evaluations each semester. Grades are kept under lock and key in the registrar's office and students are discourged from looking at them. They are available on transcripts for grad school and stuff though.

It keeps the focus on the education, not on the grades. But it requires a high teacher/student ratio and a lot of dedication to the process.
posted by SLC Mom at 12:16 PM on May 28, 2010 [1 favorite]

Cool Papa Bell, it seems from your comment that you think a professor's job is simply to assign grades, not to teach. You might reply that "evaluation is a part of teaching," and that's true--but according to the professor's evaluation of her students, all of them worked harder and produced better work as a result of the classroom dynamic she selected. AND all of them got As. To me, it seems like a pretty good deal indeed.
posted by duvatney at 12:35 PM on May 28, 2010

This seems like it would lead to a Prisoner's Dilemma situation. There's a pressure to cooperate and hand out "A"s so that the others will reciprocate. Defecting early and giving a lesser (even more honest) grade will just lead to your own work being judged more harshly than perhaps it deserves.

Heck, as a former adjunct college teacher in the era of student evaluations actually carrying weight with regard to whether you get hired again the next semester or not, this dynamic happens even when the instructor is the only one grading.
posted by not that girl at 12:40 PM on May 28, 2010

It may be a good thing to keep Cathy Davidson out of the grading business- she was a member of the "Group of 88" that presupposed guilt of the lacrosse team members. It's not hard to imagine that a jock in one of her classes may well get a fairer shake from his or her fellow students.
posted by jenkinsEar at 12:46 PM on May 28, 2010 [1 favorite]

It's perfectly fine to include peer/self-evaluation as a component of grading. The key thing is to have the students evaluate things that they can evaluate effectively and not the things they can't.

A lot of students hate collaborative projects (how they think they'll avoid them for their entire professional careers, I don't know), mainly because they resent getting a shared grade when not all team members contribute at the same level. (I have been on roughly 4000 committees where it's commonplace for some members not to do their fair share when the whole group's results and reputation are on the line, so again, I'm not sure where this wondrous planet is on which people are judged solely for their own efforts.)

Anyhow, when groups have to produce a report or presentation, I usually build in a team evaluation component -- each team member evaluates the team's process as a whole and the contribution of each individual member including her/himself. And then the other half or two-thirds of the grade is my evaluation of the final product, the team process, and individual member contributions, based on work logs, Google Docs revision history, etc.

I've found that they're largely pretty honest about the team/individual strengths and weaknesses as long as you provide a detailed rubric and require reasons for each rating.
posted by FelliniBlank at 1:21 PM on May 28, 2010

jenkinsEar, the Group of 88 did not presuppose guilt of the lacrosse players. The ad merely said that the incident was a "social disaster." No reference was made to the rape allegations. The bulk of the text was made up of quotes from students on the racism and sexism that they found pervasive in Duke's social atmosphere. You can see the original ad here.

The idea that the Group of 88 assumed that the lacrosse players were guilty was part of the right-wing backlash that followed the steadily increasing evidence that no sexual assault took place. Later, the faculty addressed these rumors and referred them to the original ad (the link no longer works here, see above) but this had little effect upon the public's (distorted) perception of their statement.

Sorry to derail, but I don't think this tired (and mostly unchallenged) criticism of Duke's faculty really has a place in a thread about grades.
posted by duvatney at 1:23 PM on May 28, 2010 [2 favorites]

Evaluating other people's work is a valuable skill.

So why not teach them to evaluate work at a professional, rather than rank amateur level? Evaluating undergraduate work is taught, if it is taught at all, incredibly quickly to first year graduate students. Evaluating professional research is an extremely valuable and complicated skill that develops real critical thinking skills and that is never "mastered"; and it is a skill that can be marketable. The former does not help with the latter, but the latter makes the former very straightforward.

College should be at least partly about learning things you probably wouldn't have thought of on your own from people who have put in years of their life to learn about them. It's bad enough that professors in relatively prestigious schools (not talking about the article now, but about personal experience) will now drop your lowest of two midterms in classes that are already easy for anyone who has a pulse and can read above a 10th grade level. The intellectual ROI (bracketing social or economic ROI) of college seems to be on the decline already; please don't cheapen it any more than it already has been.
posted by mister-o at 1:30 PM on May 28, 2010

Cool Papa Bell, it seems from your comment that you think a professor's job is simply to assign grades, not to teach.

No, a professor's job is to produce results. Specifically, a class of well-taught students that, in the end, are graded and evaluated according to their performance. Graded and evaluated by the professor, the expert to whom I assigned the task.

Ask yourself this ... do you think the professor ever runs into a case where he/she thinks the students are flat fucking wrong? Not just a little wrong, but up is down, black is white wrong? If this were a cooking school, and everyone in the class made Twinkies, does the professor overrule them on gastronomic grounds? Come on, man -- everyone loves Twinkies! Who is this silly professor to tell us what's good and what's not good to eat?

but according to the professor's evaluation of her students, all of them worked harder and produced better work

You hire me to clean your carpets. I assign the task to your Schnauzer. I tell you I will teach him to be the best damn carpet-cleaner in the world. When he's done, I have the Golden Retriever evaluate his work. I tell you they both did a bang-up job.

Pay me.

No, fuck that. Give me tenure.
posted by Cool Papa Bell at 1:48 PM on May 28, 2010

This might work well for creative projects, but the thought of fellow undergraduates to evaluate, say, upper-level term papers in philosophy or sociology (or...or...) raises some questions, to put it mildly.

I find it interesting that the discussion here seems limited to creative or "soft science" disciplines. I guess it's because in areas where there's an obvious answer key (e.g. math, chemistry, ...), you either get the right answer or you don't. I was thinking this approach could be useful for letting students see where they (or others) went wrong in solving a given problem, but realized that well-written solutions published by the professor would achieve the same end.

This might be most interesting in an area like computer science. I've always thought of programming as the ultimate marriage of creative and analytic work — there's a correct answer (your program does the right thing) with many ways to get there. I recently had a bad grading experience where the reader for a course I was in took about six weeks to grade a programming assignment. Grading code isn't necessarily the most difficult task, but it takes time. There's a lot to read, and a lot of mental state to construct in order to truly follow the logic and program flow well enough to assign a grade. Who better to evaluate your submission for an assignment than someone else who just spent the last week working on the same thing? Students who did well will benefit from seeing other approaches, and those who struggled will get a chance to see correct answers and play with a functioning version of the program. Correctness could still be judged by the automated systems most professors already use, but I feel like students could be a great resource in evaluating code style and application of best practices.

Or, heck, even if you don't want your style points assigned by your fellow students, a peer review session without the grading component could still be a great thing. Reading other people's code is one of the best ways to better your own knowledge of a language or problem domain.
posted by spitefulcrow at 2:04 PM on May 28, 2010 [1 favorite]

re: spitefulcrow

I agree with everything you said, and the computer science example is apt and illuminating. I did not sufficiently emphasize might, and even then not without theoretical and empirical justification. But even then...if in a creative writing class, my peers think Dan Brown is a literary and stylistic genius, then give me my money back if my grades are based on their evaluations.

For what it's worth, I specifically chose philosophy and sociology and not math and chemistry. The idea that "you either get the right answer or you don't" (and that's that) in a philosophy class is not even worth considering.
posted by mister-o at 2:20 PM on May 28, 2010

But even then...if in a creative writing class, my peers think Dan Brown is a literary and stylistic genius, then give me my money back if my grades are based on their evaluations.

If you are in a class where everyone but you is certifiably stupid, you should find a new class. You are not going to learn anything in such an environment: the students have nothing to teach you, and the professor is going to spend all their time teaching the other students.

In any healthy educational environment, your fellow students should be capable of collaboration and peer-evaluation. They are very powerful pedagogical tools. Otherwise you may as well be at home with some books.
posted by mek at 2:40 PM on May 28, 2010

My father had a somewhat related trick he taught me for studying; when you're in test prep in a difficult class, write sample tests. Get a study group together... each of you should write a test on the material, and everyone takes everyone's tests. Then you grade yourselves, and compare the various tests to see what you all thought was important, and what you collectively thought didn't matter, and make sure you understand the important bits really, really well. When you've got that many minds trying to outguess the teacher like that, you won't often get a question on the real test that's much different than ones you already saw.

It also helps retention a great deal; if you've written a test on something, you'll remember it much longer than just memorizing facts long enough to regurgitate. Nothing teaches a subject like, well, teaching it. To write a good test, you have to have a much deeper understanding, and it will tend to stay with you.

This sounds like a very similar thing, but with a much stronger effect. Each week, in effect, two of the students become the teachers, and do the grading. That's first-rate thinking, and I hope the idea spreads.

One worry: it may take a better teacher than normal to pull it off. And there's the possibility for gamesmanship and collaboration. But, since the teacher appears to still be reading all the papers, that should be fairly clear, and he or she should be able to take measures.

This is the kind of experimentation we need in education. The old patterns work reasonably well, but they haven't substantially changed in centuries. One person standing at the lectern, pontificating, is likely to be inferior to tapping into the brains of the whole room.

The best teachers I've known have believed that they were just students who knew a little more, and this appears to be an expansion of that idea.

I'd love to take a class structured like that. It sounds like a lot of fun.
posted by Malor at 2:47 PM on May 28, 2010 [2 favorites]

Like Sarah Lawrence, the school I went to--Hampshire College--gives evaluations rather than grades. There I generally felt like I got a great experience in the classroom, day to day, and I always felt like I was supposed to try and say something significant and make an effort to understand the material in a deep way rather than just regurgitate notes. I suppose you could argue that Hampshire attracted folks with that sort of attitude in the first place, and I was certainly pushed harder by being around thoughtful, enthusiastic learners, but I think the system has something to do with it as well.

Anyways, dunno about this peer-grading strategy and obviously I agree with the Hampshire/SL/Reed model, but I give her credit for trying something interesting, at least, and acknowledging that grading is problematic. I'd like to see more institutions question grading, but I can see why most wouldn't.

Fundamentally I know grading has no real useful relationship to learning, absolutely speaking; it's an unnecessary add-on. Learning is something you do because you want to understand something and grasp the concepts. Learning happens most successfully when you are motivated and curious. Teaching is a cooperative, collaborative, reciprocal process. True learning does not always have a clearcut division of roles between the teacher and student. Teaching/learning is not necessarily linear, it's a different experience and process for everyone, and grades for some people (teachers and students!) actively stifle their interest rather than pushing them harder. But we need some way to compare people, I guess, so they can fit into industrialized, or something like that.

But fundamentally grades are bullshit.

(For the record I've also taught and also hated giving grades.)
posted by dubitable at 4:45 PM on May 28, 2010

One of the comments on the linked article brought up a good point: how does peer grading account for those students with a disability where they get accommodation? You just tell the whole class about it?
posted by jacalata at 6:37 PM on May 28, 2010

Yeah, we had an instructor do this pretty effectively for some years.

Then the TA union grieved it on the basis that he was having unpaid students do labour TAs are paid to do, thereby devaluing the work and abusing the students. The TA union won.

No one tries this anymore.

Peer assessment is good; peer assessment that replaces paid labour? Recipe for disaster.
posted by Hildegarde at 8:03 PM on May 28, 2010

I'm giving everyone a check-minus for their contributions to this thread.
posted by empath at 9:25 AM on May 29, 2010 [1 favorite]

I don't know what some of you would think of my undergraduate university, if you think that one classes' grades being determined by peer evaluation is the sign of the devil.

Undergrad U strongly promoted student evaluation of student work; sometimes that meant a simple exchange of first drafts for midterms, sometimes that meant evaluating fellow students for their final participation grade, and sometimes that meant evaluating fellow students for their entire grade. It worked pretty well, and I can't say that anyone felt cheated or wanted their money back (then again, Undergrad U was a liberal arts college. You didn't really hear anyone saying 'I pay your salary, give me the grade I want.').

It goes further than peer evaluation, though. At my undergrad there exists an entire alternative college--call it Hippie College because it was founded during the turmoil of the sixties--that survives without grades or majors. Students create their own majors (call them 'emphases') and defend them in front of peers and professors.

Hippie College students are not given grades but rather narrative evaluations (with some exceptions, as in calculus or the like, where professors refuse). They are forced to defend their emphasis when they prepare to graduate (and by extension, they defend their entire time at Undergrad University) in front of a group of their peers and professors. It's a unique feature, and what makes it even more unique is Hippie College co-exists with the more normal denizens of Undergrad University. In fact, I was a 'normal denizen' but since I took classes in Hippie College or with professors who preferred peer evaluation and professor narratives, my transcripts include a page of letter grades and about eight pages of narrative.

I really like this model: sometimes you want a grade out of a class (especially one in, say, math) and sometimes you want a narrative (biology, oddly enough, is way more interesting with the narrative approach to grading). Sometimes you want a peer to evaluate you, and sometimes you want a professor to do so--at my undergrad, you frequently got both. I don't think it's a stretch to say that's there's a lot of parallels to the real world in that model.
posted by librarylis at 2:38 AM on May 30, 2010

« Older Lori Berenson paroled   |   Opera Web Browser: Faster Than A Potato Newer »

This thread has been archived and is closed to new comments