11 May 2017

Experiments with Feedback and Grading in a First-Year Writing Course


It's been a while since I last wrote here about teaching, for a simple reason: I've been teaching the same course, First-Year Writing, for a couple of years now, and haven't really had much to say about it. (Literature grad students at UNH used to be able to get some lit courses to teach after a required year of teaching what we colloquially call 401, but various forces related to lower enrollments made my cohort the last to get any lit courses [when I taught Literary Analysis and then an American lit survey], and so for the past two years I've taught nothing but 401).

For the upcoming year, the university awarded me a Dissertation Year Fellowship, so I will not be teaching. Before all memory of the past few years leaves my mind, here are some reflections...

This academic year, bored to death with my own teaching, I decided to experiment with the course a bit, and those experiments worked out well generally, so perhaps they are worth sharing here.

Most of my experiments are stolen/adapted from other teachers. Last summer, I went back to the work of Peter Elbow, the single greatest influence on my teaching of writing. Elbow's books Everyone Can Write, Writing with Power, Writing without Teachers, and A Community of Writers were hugely influential on my teaching when I first encountered them as a young teacher, and I have returned to parts of each through the years to keep reminding myself of the basic principles of what I do.

While Elbow provides the foundation for what I aim for with writing courses, my recent experiments have primarily been inspired by the experiments of my friend Robin DeRosathe writings of John Warner at Inside Higher Ed and Arthur Chiaravalli's piece "Teachers Going Gradeless", as well as by the examples of some of my friends at UNH who tried out similar things and generously shared their thoughts and materials.

The key changes in my teaching were the use of a "B Contract" and a portfolio system. These have worked so well that I plan to adapt them to as many courses that I teach in the future as I can.

Of course, being an Elbow fan, I'd used portfolios before. I've always liked them in concept, but have often shied away from them because I struggled to find a system where reading and grading them did not overwhelm me in the time between when they were turned in and when grades were due to the registrar. For me, at least, there simply isn't time or energy to do a thorough, thoughtful reading of a portfolio at the end of a term. The B contract along with consistent self-evaluation assignments for the students help with that, and I no longer dread the final grading.

If you want to see how all this plays out in policy and practice, my most recent syllabus is here. (Were I teaching in the fall, I would weed my syllabus. There's a lot of policies we're required to include, but the syllabus needs to be rewritten to be more manageable for students. Even I get lost in its brambles as it is!)

The key problems I identified last summer as ones I wanted to solve this year were: 1.) The research papers are always the weakest of the term, and I generally don't enjoy reading and grading them or have much faith that students have learned a lot from the process; 2.) I needed to spend less time reading and grading student work generally, because I had become like a goldfish growing to fit its environment: I was devoting nearly the same amount of time to this one course as I used to devote to three courses. That's not sustainable with doctoral work to do, and by the end of last year, I was ragged.

I sought out solutions, and the result was a slight improvement in the research papers and a significant improvement in my experience of grading. It's the latter I want to talk about here, since during both terms this year, it was the unqualified success.


I hate grades and I hate grading. I have since I was a high school student and, seeking some explanation for my miseries, read books on education by Paul Goodman and Jonathan Kozol and John Holt. In college, I hardly ever opened my grade reports, and resisted classes with lots of grading. As a high school teacher, I never had the option to de-emphasize grades, because I worked at schools that graded students constantly. (One school issued grade reports every 2 or 3 weeks. It was a ghastly waste of time.) When I started teaching college, I kept trying various grading schemes, never with much happiness. As an adjunct, I didn't feel in a position to be able to experiment a lot, and I had the same reluctance when I began as a grad student. But by this last year, I felt secure in my position, and I vowed to experiment with grading — to truly experiment, not just tinker, as I'd done before.

The effect made me wish I'd done similar things years ago. How much angst I could have saved myself and my students!

First, I should explain the B contract. It guarantees the students a B for their final term grade if they meet certain requirements (attendance, participation, turning in all work on time). At the end of the term, they fill out a Google form with check marks for each of the items, then enter their name to certify that in their opinion they did or did not meet the contract requirements. I take a look at them to make sure I agree, and I have yet to disagree with a student's perception of whether they met those requirements. The contract is as much a psychological tool as a tool for grading. My course is graded quite differently from most others they've experienced, and it de-emphasizes grades in a way that is anxiety-provoking for students who've spent more than a decade being subjected to standardized tests and relentless grading, so the B contract serves a useful purpose in assuring them that it's okay not to be graded all the time, and that if they keep up with the work they're going to do just fine.

Thus, one of my basic principles has become: Don't just grade students less; help them feel less of a need for grades. Most of my students have gone through a type of schooling that makes it difficult for them to feel any intrinsic value to their work, and by reducing grading, we can help them move toward valuing things that matter.

Another element I added (recommended by one of my friends) was a simple system for grading homework and what we called "evaluation drafts" by using numbers: 3, 2, 1 for Excellent, Good, and Inadequate. (For the homework, these translated to grades of 100, 88, and 60; for evaluation drafts, they were purely informative and had no effect on the final grade.) This seemed to be a helpful bridge between students' expectations of constant grading and my desire for them to see beyond grades. I'm ambivalent about the effect, though, as a lot of students did focus on these grades as evaluative, though not nearly as much as they would a traditional grade. It's something I plan to keep thinking about.

Here are some other principles I settled on as I approached this school year:
  1. Establish as strong a firewall as possible between grading and feedback.
  2. Grade what is finished; give feedback on what is drafted.
  3. Make grading as mechanical as possible (e.g. via a contract).
  4. Emphasize self-evaluation.
  5. Teach grammar in the context of writing.
  6. Be open and honest about deadlines and why they exist.
  7. Be lenient when possible. Don't play cop.
Some of these have been principles I've had for most of my career, others are principles I have been attracted to but was never brave enough to put into practice. Here's how they played out...

Establish as strong a firewall as possible between grading and feedback. / Grade what is finished; give feedback on what is drafted.
I've aimed for these principles from my earliest days of teaching, because I've long known that educational research overwhelmingly supports them. Chiaravalli sums up some of that research well, though for me there was no need for the word "surprisingly":
Butler examined 3 types of feedback: scores alone, comments alone, and scores with comments. Her study showed that scores alone made students either complacent or unmotivated depending on how well they did. Scores with comments were just as ineffective in that students focused entirely on the score and ignored the comments. Surprisingly, it was the students who received comments alone that demonstrated the most improvement.
I knew all this — nobody who's spent much time reading research on the teaching of writing could not know it — and I did what I could to keep grades and feedback apart, but I was too cautious to really make it a core principle of my teaching until this year. The insidious culture of grading, grading, grading made me feel like I was doing things wrong when I wanted to do what I knew was, in fact, by far the best way to go.

When I finally did this via the B contract and the portfolio, it had a substantial positive effect on my conversations with students about their work and on their attitude toward revision and toward experimenting with their writing. With this system, more students take risks and discover new writing abilities and new (helpful) attitudes toward writing than did in my classes before. They write more, they revise more, and their work at the end of the term is noticeably better on average. Simply put: This system helps more of my students strengthen their writing skills in the short time that we have together.

I tend toward two types of feedback: I have conversations with students about their work, and I write comments about their evaluation drafts. (I also give students feedback as they're working if they email drafts in progress to me. I tell them if they do this that I will almost always be able to get back to them in 24 hours. Usually, they just want help with a particular aspect of the work or they want to know that they're on the right track. This doesn't take a lot of time for me, and it can make a real difference in the quality of their work.) The only really time-consuming part is the evaluation drafts, since they all come in at once, but there's something about the freedom of not having to assign a grade to these drafts that makes the work feel lighter. I generally read the papers and then write a note to the students about what I think the strengths and weaknesses are, always with an eye toward revision. I rarely mark anything in the text itself, as I prefer to make comments such as "Be careful with its/it's when you proofread," or "There are some misspellings you'll want to look for before you turn in the final draft." If I point out all the proofreading mistakes to students in their text, they'll just change them mechanically, but if I send them back to look for them on their own, they'll engage with their work more fully and learn a lot more. (Or they will ignore me, in which case we'll have something to talk about at portfolio time...) If they can't figure it out, they know they can always ask for clarification.

I always tell the students that I'm happy to give them more feedback if they want it; they just have to ask. This puts the responsibility with them, and it saves me giving feedback that the student will ignore.

UNH is famous for the extensive conferences in its Composition program, but I have to admit that I broke some rules here, and my class has been better for it. As part of the B contract, I require students to make an appointment with me to meet twice during the term to talk about their work. Additionally, particularly during the research paper unit, we have open days (often meeting at the library) where they are free to do their work and I am available as a resource for them. Students show me drafts, ask questions as they brainstorm, get help finding books, etc. Lots of my colleagues would meet with students on separate conference days, sitting through one conference after another for hours at a time (I did this a few times and decided it was bonkers). I think we're supposed to meet with students for each of the three required papers, but I didn't find mandating 3 formal conferences to be useful, given how much feedback my students got and how accessible I am to them throughout the term. I've seen no decline in the quality of my students' work since I moved away from the 3-formal-conferences rule, and if anything I've spent more time cumulatively with students talking about their writing, it's just been more dispersed and some students have sought out more time than others. The formal conferences approach may work well for other people who are better at long conferences than I am, but for me as a teacher, that structure was more of a hindrance than a help.

The separation of feedback from grading also helps me grade quickly at the end of the term. I've seen almost everything in the portfolio at least once, sometimes multiple times, and talked with each student about most of it as well. I tell them repeatedly how I will read their portfolios: First, I look for violations of the basic guidelines (not enough materials, missing sections, etc.), then I read their reflection, and with that reflection in mind, I page through the portfolio. I return to the reflection to determine the grade: I decide whether the grade they propose for themself makes sense to me. If it does, I write very simple feedback (a sentence or two at most) and give them that grade. If I don't agree, I explain why, and, if they've met the basic guidelines, I go with a grade in between my own and their proposed one. With experience, all of this rarely takes more than 10 minutes per portfolio.

Is it fair for me to spend only 10 minutes (or less!) assessing something that is 60% of their final grade and which they've put many hours into creating? It is if I've also put hours into evaluating their work along the way. They'll learn a lot by putting the portfolio together and writing the reflection. That work has value in and of itself. They don't need to work on their writing any more for me, they won't be revising it again, so my job, once the portfolio is done, is to figure out the grade, and I can do that without a whole lot of dillydallying. At the end of the term, that's what they want and it's certainly what I want.

Some terms in the past, I've had almost paralyzing anxiety at the thought of how much student work was coming in and how much work I had to do to assess it and provide feedback. No more. The feedback is extensive, but it is spread through the course, making it more bearable for me and, most importantly, far more useful to the students. Now, when the term ends, I enjoy seeing my students' most polished work, I enjoy seeing them think about themselves as writers, and I enjoy spending only a small amount of time coming up with a final grade.

Make grading as mechanical as possible (e.g. via a contract).
I hate the subjectivity of grading. I'm perfectly comfortable with subjectivity in evaluation, but grades can have such an effect on the graded that I have long tried to reduce my own subjectivity in the grade — I want it to be more "This grade means the student did x, y, and z" rather than "This grade is what I think of this student's work."

(I must pause here to say that I detest participation grades, but have often been required to give them, including when teaching First-Year Writing. Such grades are what Ken O'Connor calls "Practices That Distort Achievement" [and boy do I wish I'd had his book back when I was teaching high school!]. I've tried to mitigate the uselessness of participation grades via the B contract and self-evaluations.)

Obviously, we want grades to be meaningful. I tell students frequently that a grade is a kind of communication, and if it is incoherent, it's not doing what it's supposed to do. We want high grades to mean a student has achieved most or all of the goals of a course and a low grade to mean the student did not meet those goals. I've spent a lot of time over the last nearly-20-years overthinking and overcomplicating this, trying to get the little numbers and letters to say more than they ever could. But grades are shallow evaluations and they mustn't be mistaken for feedback. Let them do the modest bit of work they can do, and don't try to make them stand for all the complexities of a course, because that not only strains the role of the grade, but it devalues the work of teachers and students. A course that can accurately be summed up with a letter or number is a waste of everybody's time.

Emphasize self-evaluation.
This is something else I've done for a long time, but rarely with enough confidence. I must admit: as a student, I hated evaluating my own work. I firmly believed it was the teacher's job and not mine. I wanted that external validation/criticism. I wasn't wrong about that — one of the great values a teacher can provide to a course is informed evaluation of student work — but I was quite wrong that it wasn't my job to evaluate myself.

We all need the ability to assess whether we've done a good job, a great job, a bad job. Like any other skill, evaluating ourselves honestly takes practice. I've come to believe it's one of the most valuable skills we can help our students develop. The hypercritical student needs to be able to step back now and then and say, "Actually, this thing is pretty good," and the hyperhubristic student needs to be able to admit that not everything they do is an unparalleled gift to human culture. If you're always waiting around for somebody else to assess you, you will have a hard time in life.

One key to making self-evaluation work in the classroom is to use it in a variety of ways, and to use it frequently. I at least get students talking about their own assessment of their work as soon as possible, and then add in quick, informal writing assignments to help them think of what they're doing well and what they'd like to improve on. I give them a formal self-evaluation at the end of each unit using Google forms, having them rate their efforts in group work, homework, etc. and asking them to think about what they'd like to improve on. (The average of all these ratings becomes their participation grade.)  Finally, in their portfolio they write an extensive reflection on their writing and the course. The early self-evaluations tend to be facile and cautious, but by the time the students get to the final portfolio reflection, what they write is usually quite thoughtful and perceptive.

Teach Grammar in the Context of Writing
I've known this for a long time and seldom done differently, but I put it here because inevitably when you're a writing teacher people ask, often vehemently, "What about grammar?!"

Again, there's a gigantic pile of research on this. (When I was starting out as a teacher, I found the work of Constance Weaver particularly helpful.) These days, I mostly discuss grammar and style within the context of the proofreading guidelines that I require the students to follow for their evaluation and portfolio drafts, as well as in discussions of ways to write different sorts of sentences. For instance, one of our homework assignments is to look at this selection of sentences and then revise two or more of their own to emulate a few of the models' structure and diction. Such work inevitably leads to discussions of punctuation, independent clauses, etc.

Be open and honest about deadlines and why they exist.
I used to be tough about late work from students. I wouldn't accept any except for absolute emergency situations. If you've got a lot of students, as I've had at various times, work coming in off schedule is a nightmare.

Then I began thinking about how important it is to let students see why policies exist, and I started questioning some of my ideas. At the same time, discussions of late work seemed to have sprouted from the zeitgeist (e.g. Ellen Boucher in The Chronicle). With reflection, I adjusted some policies, and for about a year and a half now, I've distinguished between "soft deadlines" and "hard deadlines" for students, and I'm honest with them about why these exist. 

Most of the deadlines for my classes are "soft": I've imposed them because they fit with the course schedule and they will allow me to evaluate work and return it within a reasonable time. If a lot of students missed such deadlines, it would cause problems; if one or two do, I will hardly notice, though they put themselves in the potentially difficult position of having to keep up with current work while catching up on late work. I tell them this, and say I will give extensions for work due on a soft deadline without any question. I know mine is not their only course, and that their life schedules may be complicated.

Hard deadlines are different. There aren't many — usually just the final due date for the portfolio. I can't give much more than a few hours extension even in the case of an emergency, because there's a deadline for turning grades in. "Short of a death in your immediate family or a meteor crashing through your dorm, don't even bother to ask for an extension for a hard deadline," I tell them.

This seems to work well. Students with difficult schedules, or who are struggling with a particular assignment, feel less panic because of the soft deadlines, and they typically turn work in a bit early for hard deadlines, because they're terrified of missing them. When hard deadlines are few, they are meaningful and feel less oppressive. This move fits with my final principle:

Be lenient when possible. Don't play cop.
One of the aspects of working at a boarding school that I loathed was the need to be a cop. I want less policing in this world, not more; I want to evict the cop from my head.

I've been around enough teachers to know that some of them got into the profession because they are sadists and enjoy wielding power over a group of students. Some teachers get a thrill from policing and punishing. I feel contempt and pity for such people.

Plenty of other people just get exasperated with students and focus on their errors and failures more than their successes. I get it. I, too, get tired. I, too, have just wanted to scream about a students' behavior or about a pile of uninspired papers I had to read or some utterly tone-deaf email. But my exasperation is no help to the student.

(I see even people I like frequently complaining about student work on Facebook or in staff meetings, and every time I think: If I were on a hiring committee, I would never give you a job. Why would any school want to employ someone who apparently hates most of their students and finds nothing but aggravation in teaching?)

When I was a young teacher, I tried hard to be tough and merciless. I couldn't do it. Policing and punishing nauseated me. I got into teaching because I like sharing knowledge and I like seeing what students can do. That's it. That's enough.

I assume good faith on the part of my students, and make clear that this is my assumption. If they say they need an absence to be excused, I excuse it. (Left to myself, I wouldn't have an attendance policy beyond the B contract, but at UNH we have a common attendance policy for all First-Year Writing courses.) If they're struggling with elements in the course, we talk about it. If their behavior doesn't seem productive to me, I address it and also listen carefully to their perspective. If I want the students to behave like responsible adults, I have to treat them like responsible adults, not like toddlers in a crib.

Certainly, such an approach is idealistic and opens me up to occasionally being taken advantage of. So be it. It's far less stressful and far more productive than seeing my students as obstacles or enemies, and it gives me many more opportunities to celebrate my students' accomplishments and experiments.


This academic year was the first one where I worked hard to stick to each of these principles rather than just a few of them, and the result has been a significant improvement in the quality of my students' work overall. Other principles are useful as well, and there are certainly problems I haven't solved (arrrghhh, research papers!), but this foundation is one I plan to build on when I return to teaching a year or more from now. 

Last week, when I left my 401 classroom for the last time, I said to the students, "Thank you for being a great class. This has been fun." I meant it.