Teaching
A simulation leaves behind a room full of things that look gradable and mostly are not. This page sets out what can be graded fairly, what the grade should weigh, how to write the comment, and why the safest answer is to grade the preparation rather than the performance.
Grade the preparation, not the performance
The temptation is to grade the room: who spoke well, who held a coalition together, whose resolution carried. There is a documented reason not to. Engel and colleagues found that women participated less relative to their numbers in a Model UN simulation, that the gap cost them grades, that they enjoyed it less and felt less included, and that a short explicit intervention meant to close it made it worse. Coughlin found women in the majority at a Model UN conference while men gave most of the speeches and shaped the resolutions. A participation grade, imposed on a room with a participation gap that size, converts the gap into a transcript.
The written preparation is the opposite case. Every student writes it alone, under the same conditions, before the room can reward whoever is most comfortable speaking in it. It also carries the learning you actually claimed: a student who can say what their office may and may not do, and how the body decides, has learned the thing the session was for.
This is not an argument for leaving the session ungraded. It is an argument for grading what the session produced, which is a document, rather than how a student conducted themselves while it was being produced.
Four things worth grading
| What you collect | What it actually tests | Weight, and the catch |
|---|---|---|
| A position paper, due before the session four to five pages, or two if the session is one period | Whether the student understands the office rather than the person: its powers, its limits, who it has to answer to at home, and how the body it sits in decides. | The anchor. Ten to fifteen percent of a course grade is enough; it does not need to be the biggest assignment to be the fairest one. The catch is that it has to be due before the room meets, or it becomes a report on what happened. |
| A one-page negotiating brief the student writes themselves | Whether they can turn the position paper into something usable: an opening ask, a line they will not cross, and the two delegations they intend to talk to first. | Cheap to grade, high in diagnostic value, and the single best predictor of whether a student will have anything to say. Mark it credit or no credit; a number here buys nothing. |
| The document the room produced resolution, committee report, signed ruling, conclusions, a decision on the application | Whether the body did the thing the body does. Not who won it. | Grade it as a group product or not at all. Grading it per student reintroduces the participation problem by the back door, because the students whose language survived into the final text are the ones who were loudest. |
| A debrief essay, written after one to two pages, within a week | Whether the student can explain the outcome by naming an incentive rather than a personality. This is the one that tests transfer. | The highest-value piece of writing in the whole exercise, and the one most often cut for time. Cut an agenda item instead. |
A rubric that travels
This is the rubric used on the position paper in a college EU simulation, and it transfers to any body with the names changed. Each part is scored out of 100 and weighted; the weighted total is rounded half up, and the arithmetic is shown to the student.
| Part | Weight | What reaches the top of the band |
|---|---|---|
| The position | 45% | Procedural and institutional specificity above everything else. Citing the governing document by article rather than by name. Naming the forum and saying how it decides. Reaching the working groups and the committee of permanent representatives beneath the headline body, which only the top of a set does. |
| The person, or the party | 15% | Biography used to predict a negotiating posture, not recited. Where the individual is obscure, the national party and the European or caucus group profiled instead, which is usually the more informative answer anyway. |
| The judgment on the file | 15% | A position on this proposal, in the officeholder’s voice, with a reason. Not the student’s own view, and not a tour of the whole agenda. |
| Quality of writing | 25% | Structure signposted to the parts asked for, sources attached in the body to the claims they support, and the subject’s own name spelled correctly. |
The weighting is the argument. Nearly half the grade sits on the institutional part, because that is where the thinking is and that is what the simulation was supposed to buy. Writing carries a quarter, which is enough to matter and not enough to turn the grade into an English mark.
Inside that largest part, one question does most of the discriminating, and it is worth choosing it deliberately. In an EU simulation it is whether the student treats the Union as intergovernmental or supranational, and the top of the band is not the label but the reason: a small member state has a real interest in a strong Commission, which is an answer; “intergovernmental” on its own is not. Pick the equivalent question for your own body. In a Security Council it is what the veto does to the drafting rather than to the vote. In a congressional committee it is why the rule is voted first. In a planning board it is whether the criteria in the ordinance bind the board or merely guide it. One question, answered with a reason, separates the A from the B more reliably than any quantity of detail, and a student who never answers it should not clear the top band however good the rest is.
The two assignments, written out, with this rubric at the end of each
The Country Profile · The Alter Ego Paper
Two four-to-five page papers, assigned several weeks apart, that between them do all the preparation a student needs: the country profile answers what the government wants, and the alter ego paper answers what this student can do about it. Each handout gives the three parts with what the top of the band looks like, the deductions that recur, a section on adapting it to a Model UN, a Model Congress or a planning board, and the weighted rubric above at the end. Free to adapt.
Where the points actually go
These are the recurring deductions from recent sets of position papers, in rough order of how often they appear. They are worth putting in the assignment prompt, because almost all of them are cheap for a student to avoid once named.
- Citations that exist on the reference page and not in the body. The most common single deduction. A paper can carry nine primary sources and attach none of them to a claim.
- Nobody states the decision rule. Across fourteen papers in one set, almost no student named the voting rule their own body uses. Put it in the prompt as a required sentence.
- No section headings for the parts the assignment asked for. The most common structural miss, and the easiest to fix.
- The head-of-state and head-of-government category error, or its equivalent in another body: a national minister described as not being an EU decision-maker when in Council that minister is a co-legislator. It appeared in three of ten papers in one set. Where the same paper gets it right elsewhere, grade it as a slip rather than a misunderstanding.
- Institutional description that never connects to the role. A correct paragraph about the legislature, sitting in a paper about a minister, doing no work.
- The student’s own view where the officeholder’s was asked for, and the whole agenda listed in place of a position on the student’s own file.
- Weak sources where primary ones are free. A personality section built on two videos, when the ministry, the constitution and the official transcript are all online.
- The subject’s name misspelled. For a paper about one named individual it is the first thing a reader notices.
The student who prepared hard and got steamrolled
This is the case that makes people want a participation grade, and it is the strongest argument against having one. A student reads the ordinance, finds the controlling clause, comes in ready, and then sits behind three people who will not stop talking. If the grade rewards floor time, she is marked down for somebody else’s behavior.
Grading the preparation protects her automatically, which is most of why it is the recommendation on this page. Three structural moves protect her during the session itself.
Assign the roles rather than letting students choose them, so you own the distribution of the powerful ones. Baranowski and Weir found that students given high-power roles reported more understanding and found the exercise more worthwhile than students in low-power roles, which makes a one-time allocation of power a grading problem and not just a design choice. Rotate the chair and the floor-managing roles in anything longer than a single day. And structure the speaking rather than encouraging it: a named question put to every role, a speaking order instead of a raised hand, a ninety-second limit, and a required response to a prior speaker. A turn then arrives by structure rather than by being won.
One more rule, and it matters more than it sounds. Where your own design may have put the student in an impossible position, withhold the deduction and ask her about it before you score it. If the role sheet and the agenda disagreed, or the file she was given was not the file the room took up, that is your error arriving on her paper.
Writing the comment
One to two marginal comments a page, each anchored to the passage it is about, most of them a short question: what is that, how, why, what data supports that. Then exactly one closing comment that lays out the grade and explains it, with the part scores and the weighted total shown. Two habits are worth copying: comment volume runs opposite to the grade, so the strongest papers get the fewest marks and the closing comment gets longer as the grade falls; and praise is a rating rather than a description, so the adjective and the number have to agree. “An outstanding paper” above an 84 undoes the rubric.
Before you deliver a grade, check that the part score matches your own comment. A closing comment saying the paper is unsourced, sitting above a 92 for writing, is not defensible, and a second read catches that kind of internal inconsistency far more often than it catches a difference of taste.
Never bend a part score to make the total land on the number you had in mind. That falsifies the rubric record, which is the thing that makes the grade explainable six weeks later. If the weighted total comes out at 87 and you felt 88, the answer is 87. Lateness is a flat deduction from the computed total, taken from the total and never from a part score, and taken once; if your course already has an automatic late policy, do not also flag the submission, or the student pays twice for one day.
What not to grade
Effort, enthusiasm, costume and props. They are visible, they feel like information, and they measure confidence rather than preparation.
Who won. A simulation worth running is one where the vote is genuinely allowed to fail, and a grade attached to the outcome quietly removes that. The student on the losing side of a well-argued question has usually learned more.
Prose, line by line. Grammar is what the writing part of the rubric is for, plus at most one sentence telling the student to copy edit it. Line-editing a paper takes an hour and changes nothing.
The student. Every criticism is an instruction about the document: you need to, I would, this should. Nothing in a comment should be about the kind of person who wrote it.
And one caution about the thing everyone reaches for first. Self-rated gains are the least reliable finding in this literature, because students revise their opinion of their former selves downward and inflate the apparent improvement. A survey asking students how much they learned is worth collecting and is not worth grading. The Debrief has the exercises that produce gradable writing instead, and What the Research Says sets out what the evidence does and does not support.
Baranowski and Weir 2010, JPSE 6(3). Coughlin 2013, JPSE 9(3). Engel, Pallas and Lambert 2021, JPSE 17(4). Kartal 2021, JPSE 17(sup1).
The rubric, the deduction patterns and the comment conventions here come from grading position papers for a college EU simulation. No student work, name or grade appears on this page; the patterns are stated as frequencies. If you grade these differently, or better, please say so.
