Public: Education Reform
Essay peer-assessment   (+1, -3)  [vote for, against]
members of the class review each other’s essays

An essay is a good way of learning ‘in-depth’ about a subject, while also honing your writing skills and ability to think and argue clearly.

I find it frustrating that the end result of all that effort is only read by the teacher of the course (who’s probably not even interested in your essay, since they know the topic back-the-front).

I also find it worrying that the teacher of the course is solely responsible for marking the essay, and is therefore possible that their bias against your particular argument or writing style could lower your deserved mark.

I suggest that the members of the class review each other’s essays to resolve these problems.

This would have three advantages: reading other student’s essays is a good way of learning the subject; you’d be able to compare yourself to the class standard; and there’d be less chance for teacher’s bias.

It would be too onerous for every student to read and rank the essay of every other class member. Instead each student would be randomly assigned 4 essays to read and rank. Each essay would have the student’s name removed (and replaced with an ID code) so that bias against unpopular students would be minimised.

The students would simply rank (1st, 2nd, 3rd, 4th) the 4 essays. So every essay would be read and ranked 4 times. If an essay consistently gets rated highly, it would be fair to assume it’s a good essay, and if an essay is consistently rated poorly, it's probably a poor essay. An algorithm for ranking the intermediate essays would be easy to develop (I’ll throw in the suggestion of a Bayesian filter, since it’s fairly trendy these days). The algorithm would need to account for anomalous ratings that indicate students trying to exploit the system.

The teacher would then rank and mark all the essays. If there was any discrepancy between the students’ collective ranking and the teacher’s ranking, re-evaluation of those essays might be necessary.
-- xaviergisz, May 25 2006

Professor Uses 'Crowdsourcing' to Grade http://chronicle.co...7538/?utm_source=wc
[xaviergisz, Aug 01 2009]

Bringing Auto-Graders To Student Essays http://news.slashdo...s-to-student-essays
[xaviergisz, Mar 30 2012]

[rcarty], admittedly peer assessment is known (although I've never been involved in it), but does it use this mechanism?
-- xaviergisz, May 25 2006


[jutta], nope, I've never been a teacher of a class that did essays. What's so hard about implementing it?

**edit**

I swear I just responded to a comment (now deleted)by [jutta]...
-- xaviergisz, May 25 2006


Yeah, I read on to the place where you reduce the number of essays to rate to four, and that might be doable for a single person. A whole class full would be too much.

I've never been a teacher, either, but I've had to rate things. Like most people, I can confidently judge end points of the scale, but hearing the finer notes and rating e.g. two mediocre works relative to each other turned out to be very hard, especially if the authors have different styles. It's not like apples and oranges, more like fish stew and grapefruit sorbet.
-- jutta, May 25 2006


so [rcarty], you have mfd this because its a bad mechanism of peer evaluation?

You'll also have to explain how standard peer evaluation that //involve marking rubrics// are superior to this method.
-- xaviergisz, May 25 2006


Can I get a copy of it [rcarty]?
-- methinksnot, May 26 2006


[rcarty], I disagree that the students would collectively vote for amusing yet poorly written essays.

Most students would have spent considerable time and effort writing their essays and would consequently appreciate others who had done the same.

I must admit I had in mind university/post-graduate students to use this system.

I'm a great believer in the 'wisdom of crowds' /democratic processes, and believe that (when set-up correctly) these processes will always give a fairer result than an individual assessment.

I'm still puzzled by what these 'marking rubrics' are, but it sounds like a formulaic way of assessing essays. Essays should be read and assessed 'as a whole' - if you think its a good essay - it is; no amount of dry analysis should dissuade you from this view.
-- xaviergisz, May 26 2006


in what way is this 'essay self-assessment' ?
-- neilp, May 26 2006


[neilp], you're right. I should have titled it 'essay peer-assessment'... too late now though.
-- xaviergisz, May 26 2006


I've re-titled it. Apologies to anyone who clicked on it hoping for a new idea.
-- xaviergisz, May 26 2006


It is a good idea. It is done. It works better for college than high school, I think. I am looking forward to that essay by [rcarty] in l33t talk.
-- bungston, May 26 2006


wouldn't it be fun to give <x> who's an idiot 100's for all his/her essays?... and the reverse to <y> who actually deserves it.

Not saying it doesn't have its strong points given a bunch of people who are actually interested in learning.

[-]
-- FlyingToaster, Aug 02 2009


[FT], I mentioned that the author of each essay is anonymized so this couldn't happen.
-- xaviergisz, Aug 02 2009


This is standard practice in Scottish schools - and I'd imagine all schools in the UK. With less emphasis on ranking the essays in order and more emphasis on highlighting areas of the essays are excellent, which areas could be improved etc though.

Class discussion of a selection of essays negates any mucking around.
-- ChewTheBeef, Aug 02 2009


Names removed wouldn't be able to hide grammar usage and style. Sorry but a properly accreditable teacher is the way to go: he/she has (or should have) an idea as to where the essay fits, not only within the course but within the subject, and not only that but where it crosses over into other subjects.

I'm not saying it wouldn't be useless but ... well yes I am actually.... Great Depression, circle-jerk, etc.
-- FlyingToaster, Aug 02 2009


When the essay is a means of assessing a student's progress in a particular course or curriculum, the grading's best done by someone with knowledge of that course or curriculum. Peer review is more appropriate when 1) the criteria are esthetic (as in creative writing workshops) or 2) the peers are domain experts (as in academic peer review, not that that works perfectly).

//Essays should be read and assessed 'as a whole'// If the criteria are purely esthetic perhaps. And even then, is this appropriate for didactics, as opposed to selection for a prize or publication in a magazine or anthology? Shouldn't the grading system lend itself to constructive criticism? Also, it's really very difficult to grade without a rubric for guidance, and even harder to get a group of graders to agree without one. And harder still to defend the grade when the student disputes it.

Interesting to contemplate a hybrid system. The instructor grades some random sample (possibly all) of essays in the usual way. Students grade each others' essays (again, either a random sample, or all of them). The final grade for each essay is a weighted average of all the grades it received, with the instructor's grade weighted heavily, and the peer grades weighted according to how well each peer grader did on their own essay.

All sorts of interesting emergent symmetry breaking in such a system, and when you've finished studying that behaviour, you can investigate the possibilities for collusion.
-- mouseposture, Mar 31 2012


Mildly baked, asked one my JHS students in Japan who marks their homework and she told me they mark their own...apparently they also get marks towards the year grade on how well they right down whatever it is the teacher writes on the board..damn, how come I never get a job like this?
-- not_morrison_rm, Mar 31 2012


We did this in one college class. It was anonymous. I marked up a paper with helpful advice and gave it what I considered a fair grade. It was returned to the dude sitting in front of me. He took the non-perfect grade as a personal betrayal of some unwritten students code and became enraged. He was saying the most vile things regarding my parentage, intelligence, honor, etc. Of course he had no idea who had graded him, and I certainly wasn't going to tell him. It was a really ugly mix of immaturity and entitlement, and I'm sure that guy didn't do well at his first performance review by someone whose salary he wasn't paying with tuition.
-- GutPunchLullabies, Mar 31 2012


He is now the Regius Professor of Underwater Basket Weaving, at Camford University.

Seriously, he probably did fine on his first APR. He thought that social relationships would determine his asessment, which is only wrong in very artificial situations. As a strategy in the real world, it's about as effective as doing good work.
-- mouseposture, Mar 31 2012


//Peer review is more appropriate when 1)... 2)...//
3) The students actually want to learn. In that case it's great: it may improve a student's understanding of the subject matter.
-- FlyingToaster, Mar 31 2012


Agreed, but that's an argument for having the peers grade essays; it's not a reason the grades they award should actually count for anything.
-- mouseposture, Mar 31 2012



random, halfbakery