My previous post ended on a high note: a list of the benefits I believe came from my decision to try specifications grading. Most of these things are hard to be absolutely sure about, but that’s true of most questions about pedagogy, especially when your only sample is a single iteration of a course. Still, I have taught first-year writing classes almost every year since 1995, so I give some weight to my own impression that on balance, this change in strategy had some good results for them and for me. It wasn’t all good, though, and especially as I do intend to try specifications grading again, I am already thinking a lot about what needs to change in the second version of the course policies and requirements.
The first problem I need to fix is one that I tried very hard to prepare for and fend off, but (perhaps inevitably) without complete success: unfamiliarity. As far as I could tell, specifications grading was a brand new concept to everyone in the class, and though some students seemed to grasp its procedures and implications very quickly, others clearly struggled–some of them right to the very end of the term. There were clear indications that some students just did not understand the mechanics of it: they did not realize (or not until it was almost too late) that their final grade was going to depend, not on doing well on an individual essay or exam, but on how consistently they did or did not do satisfactory work across the entire term.
I am confident that my explanations were clear and explicit and abundant–but the very care with which I laid out the assessment model probably backfired in some cases, as it meant there was a lot of information to take in. I tried (so hard!) to clarify and simplify, including having ‘quick reference’ tables that summed up exactly what each bundle required and what the standards were for a satisfactory assessment. I reiterated essential points in announcements and assignment instructions and in the FAQ discussion board … but I think the difference between what we were doing and what they were used to or assumed about grades and grading was just too much for some students to process. Specifications grading was an unknown unknown for them: for whatever combination of reasons (other challenges related to online learning or with information management, or the same kind of indifference or inattention to the syllabus that we always see some of, no matter what the mode of course delivery) some students not only didn’t realize it was not business as usual but they didn’t even think to check or ask. They were a minority of the class, and it was equally obvious that some students very swiftly grasped the system and made their plans and did their work accordingly. Still, next time I will be more aware of how hard the conceptual leap can be and work even harder to leave nobody behind.
On a related note, by the end of the term it was clear to me that the specific set of requirements and bundles I had devised was much harder for many students to makes sense of and (even more important, perhaps) to keep track of than I’d expected. Again, I took a lot of measures to make it all legible and easy to follow, including my handy reference tables. I am still actually quite puzzled about why so many students completed such a random array of components, such that their final tallies were not consistent with any particular bundle. As the end of term approached I did an audit of what everyone had done to that point and there were a surprising number of cases in which, if I had not changed the bundle requirements, the students would not have been able to earn even a passing grade, because they had completed a high number of one kind of components but completely dropped the ball on others.
I did revise the requirements, mostly to reduce them, although I had some misgivings about doing so. The process of revision also helped me think about ways of totting things up that might actually work better: requiring an overall number of journal entries for the term rather than a consistent number per week, for example (although, without getting too much into the weeds of my plans and planning, one goal I had and don’t want to give up on is ensuring steady effort, not a crush of last-minute contributions). Perhaps my bundles just had too many moving parts: again, a significant number of students didn’t seem to have any problems, but confusion was widespread enough that one of my key take-aways is that I need to simplify things.
(As an aside, though, one thing I did like about the bundles system with its many moving parts, however, is that it was possible to sort out particular cases where a student was far short of passing the course or just short of a better bundle by counting up the components the student was missing and devising a make-up plan. This felt much more constructive than any options we usually have for a student who has done poorly on a midterm or otherwise compromised their standing under a regular assessment scheme.)
Finally, and I’m not sure if this is a problem or it just feels like one, far more students than is typical ended up with A or A+ grades in the course. Clearly, knowing that they could earn an A if they just put in the work was highly motivating for a lot of students–and also stress-inducing, in some cases, because having set their sights on that goal, students could not let go of it, even if they were finding the work more difficult or time-consuming than they’d expected. Heading into the semester, I expected a lot of students to decide (eventually, if not immediately) against the MOST bundle, on the grounds that all they wanted was their writing requirement credit and their other courses were higher priorities. Over the years I have encountered that attitude a lot in first-year classes, often quite explicitly. Now I wonder if some of that hasn’t been the side-effect of students (especially those in STEM programs who do not think of themselves as “good” at English) simply writing off the course, especially if they get below an A on their first paper: resigning themselves to it or telling a story about it that makes it matter less. This time, told they could end up with an A if they just did X, Y, and Z, students who might not have bothered before decided to go for it…which is great! As I said in my last post, they did the work and that paid off.
And yet I can’t shake a lingering concern that the number of A+ grades I filed tells me my standards were not high enough: my requirements, or the specifications, or (more likely) the enforcement of the specifications, was lax. It’s true we were not as fussy as we originally intended to be. This was partly about grading fatigue (we had to process a lot of discussion posts!) and also about the pressure we were constantly under (arguably rightly) to give students a break during a term that was already difficult for them. I did end up filing grades across the whole range of possibilities from F to A+: there were just significantly more As and A+s than I’m used to seeing. Should I–could I–have made A and A+ grades harder to get? My guess is that increasing the requirements for the largest bundle would not have deterred most of those who went for it on the terms I set, and why would I want to deter them anyway? I think my lingering unease comes from the same doubts my colleagues expressed about quantity vs. quality and what exactly we think letter grades are supposed to mean or signal. Specifications grading is designed to sideline judgments about “excellence,” but the habit of thinking in those subjective qualitative terms is deeply ingrained in most academics, even though we know perfectly well that grades, publications, grants, awards, and tenure-track jobs depend more on other factors (including luck and privilege) than on intrinsic and objective kinds of excellence.
Summing up my own “lessons learned” from my first attempt at specifications grading, then:
- you can’t explain too clearly or too often how the system works, and even then some students will not grasp the difference it needs to make to their work habits, so I need to anticipate problems and prepare interventions;
- relatedly, I need to make my bundles simpler and more flexible: they should (probably) have fewer moving parts overall, and I should have more flexible ways of adding them up;
- and I need come to terms with the potential for what will look like grade inflation–whether this means raising (or being stricter about) standards, putting the very top grade a bit further out of reach, or accepting (maybe even encouraging!) a larger than usual number of As and A+s.
If you’ve used specifications (or contract) grading, I’d love to know if these are concerns you too have had, and if so, how you have handled them. It’s especially hard to know how big a part the transition to online teaching and learning played in the successes and failures of specifications grading for me (and for my students) last term. It might have been exactly the wrong time to try it! I’m teaching the same course again next fall, and right now we don’t know if it will be in person or online. Whichever way it goes, I have lots of time between now and then to rethink, revise, and regroup.
Novel Readings 
I tried grading essays on “mastery and completion” this fall, with students who fell short of either getting a chance to revise (only four out of twenty fell short on “mastery” on one assignment). This worked well for a small class at a small college where students are very motivated and also very grade-conscious.
If the student didn’t revise, I usually offered them a B (85) for leaving the essay at the place it was. A few of them took that; they didn’t have the energy to revise. I tried to say that was okay. Sometimes good enough should do.
Everything I’ve read about contract or specifications grading says there can be problems with it for mostly privileged students who have grown up learning to write to please teachers, but I thought it worked better than my previous more evaluative models–for one thing, it spurred my very motivated students to think about what would actually please them, personally. This was a writing class, though, not a literature class.
LikeLike
That’s really interesting, Jeanne! The opportunity to revise work we assessed as unsatisfactory was a key part of my overall scheme too: it wasn’t an automatic option for low-stakes assignments like discussion posts but we did use it as a way to encourage fixing easily mastered errors (like apostrophe errors), which is actually really hard to do ordinarily. A few essays did not meet the mark and those students had the chance to do repairs and resubmit
Moving away from a focus on “getting an A” to a focus on making the assignment one that interests them really is the main payoff for students in this kind of system, I think.
LikeLike
Thank you for this and the previous entry. They are both very helpful as I prepare my own “first attempt.” While most of what you share confirms what I learned in Nilson’s “Specifications Grading” (2015), your comment about “lingering unease…[about] what exactly we think letter grades are supposed to mean or signal” especially resonates for me. I suspect this is going to be the highest hurdle to clear when I seek approval from my associate dean to use this assessment method. I am really frustrated with the practice of passing students who have mastered only 60% of the course content. Generally speaking, her in Japan students are not required to submit transcripts when applying for jobs. As a result, there is no indication of the extent of mastery. As educators, I feel we owe it to society to graduate students who have mastered at least the core content of our courses. (Who wants a physician who has mastered only 60% of medical school?!) I’m hoping specifications grading promotes excellence.
LikeLike
I think there are definitely tensions between any kind of ‘completion’ grading (whether it’s ungrading, contract grading, or specifications grading) and the aim of mastery, of either content or skills, or of “excellence”.
I tried to set my standards for credit higher than a bare pass, but there’s no question that diligence earned students high final grades who would not have gotten them under a different assessment system. In some ways I am totally on board with that, not least because it meant they kept practising and trying and writing when an early C- or D on a paper might have meant they either dropped the class or just assumed it wasn’t going to be worth the effort to try to improve. Also this removed the ethical quandary of how to mark work by students working in English as another language who are forced to take these classes but end up struggling and working incredibly hard often for very low grades (which is a systemic issue, or an institutional one, not their individual problem).
I have used this system so far only in a first-year writing requirement course. The stakes are quite different than in medical training! One of the hardest things (in my experience anyway) to get novice writers to do is write a lot, to buy into the process of it – that you can’t expect to do well or improve as a writer if all you do is sit down the night before a deadline and crank out something to turn in the next day. To get their A in this class, they had to write something on topic every single week, and they also read lots of writing by their classmates. I honestly believe that just by doing this they learned something. It seems like a system very well suited to good writing pedagogy, in other words – I don’t know if it can serve the goals of every class (including upper level literature classes).
I also think it would work much better, or that I would at least try it in a wider range of courses, if class sizes were much smaller than ours ordinarily are, because for the feedback to be meaningful and for there to be enough follow-up to push to excellence, it would take a large investment of time per student!
LikeLike