Specifications Grading #5: End of Term Report and Reflections

Well, my first experience with specifications grading is almost over, and with the semester drawing to a close, it’s time to reflect on the experiment. Find my first entry on specs grading here, and previous entries on this experiment can be found here, here, here, ad here. But now, here are my top 5 take-aways and lessons learned from specifications grading:


#5: Specifications Grading is more work up front, but much less at the end and moving forward

If you’ve been following my posts throughout the semester, you know that I put a ton of work into converting my course to specifications grading. I ended up creating specs for over 40 assignments, most of them small concept-based modules that I carved from longer problem sets. In some cases I added to these assignments, but in most cases this was a simple conversion. But for each one, I had to write a purpose statement (so students understood the point of the module), very clear questions or tasks, and then criteria for evaluation. I think this clarity and transparency is a good thing (see #2), but it did take a lot of time.

So did creating the syllabus. I had to make decisions about what assignments went into each grade bundle, how tokens could be earned and spent, whether a module would be required or optional—the conversion took a lot longer than I had planned, and having to revise each assignment (only some of which I did at the very beginning) was a time suck during the term.

However, specs grading is saving me time now that we are in high grading season at the end of term. Since I had very clear criteria for assignments, and everything is graded as pass/fail with no partial credit, it takes me much less time to get through an assignment. I don’t have to write a lot of comments, either, as I can simply indicate which spec they did not meet, and invite them to spend a token to make another attempt and resubmit. The other day I graded about 30 pieces of student work stretched across 8 different assignments, and it took only about an hour or so.

Better yet, now that all the specifications are done, I won’t have the front-ended work next semester. After making a few simple revisions, I can post every single assignment on day one. Since this is a class I teach regularly, this means that I put in a lot of work one semester but will have an easier time grading from now on. Totally worth it.

#4 Students like it.

I did a post-course debrief with my students last week, and they all agreed that they preferred specifications grading to more traditional systems once they got the hang of it. They admitted that despite my spending more than an hour reviewing the syllabus, they found the system confusing at first. But it only took a couple of weeks for them to catch on and appreciate the differences. Here are some of the things they liked:

  1. According to their feedback, they found specs grading to be more intuitive. The grade was based on what they were able to do, rather than on points won or lost.
  2. They liked that there was no guesswork about their grade. They knew what they had to do to get the grade they wanted, and didn’t have to worry about my specific grading quirks on what constitutes A, B or C level work.
  3. Some of them claimed that specs grading pushed them further, since they knew to get ANY credit on an assignment, it had to be B level work—they couldn’t just skip questions or half-answer a question, or the rest of the work they put into an assignment would be meaningless.
  4. Others pointed out that they liked the fact that all their assignments and the criteria for each grade was transparently laid out in the syllabus, in chart form, giving them a clear checklist on what they needed to do to get the grade they wanted. They didn’t need me to update them on their grade, because they could track whether or not they had received satisfactory level scores on the required assignments in their chosen bundle.
  5. Finally, they liked that I put the grade in their hands, and that there was no judgment from me on what grade they aimed at earning. One student noted how odd it was that I was essentially giving them permission to get a C in my class—but appreciated that they were CHOOSING to get a C, rather than getting one because they fell short in one respect or another.

The only negative feedback I got was the early confusion due to the unfamiliarity of the system, and an issue with the assignment deadlines. The class consists of over 30 required and optional homework modules, and originally I put all the deadlines in the syllabus. Since we fell behind schedule, I kept having to change the deadlines and this caused confusion—some days they would have three or four modules due, and other days, none. They recommended having modules due on set dates—say, every other Friday—rather than on individual days throughout the term. This would help them with scheduling their time and remembering due dates. I agreed to make this change for next semester.

All else being equal, I want students to LIKE my class. As long as I can preserve academic integrity and equality, I’m open to trying unique systems that students can support.

#3 Tokens are better than extra credit

I’ve always hated extra credit as a motivator. Some students focus on extra credit more than the ACTUAL credit for the course; others don’t care about it at all since they don’t need it to do well. I’m also not completely comfortable with the idea of giving a student a better grade than they earned because they attended an event or performed well in a game or simulation. To this, I add a comment from a student: that they never know how extra credit will affect their grade until the very end of the course, so it’s not a strong motivator for them.

Tokens are SO much better. First, they eliminate the need for extra credit. All the kinds of things that previously earned extra credit—attending events, winning games, coordinating group work—now earns tokens. This is a more meaningful award—students know exactly what they can do with their tokens, whereas extra credit is not as tangible a reward.

Tokens also eliminate participation grades, which I’ve also never quite liked. I give tokens when students say smart things, or succeed at something they’ve struggled with, or when a shy student speaks up, or when a group works well together and stays on task. Rather than directly affecting grades, tokens give students flexibility in terms of the benefits they confer. They can use them to redo assignments, or earn extensions on assignments, or forgive absences.

Another benefit is that it almost entirely eliminates the need for students to ask me for extensions or excuses. Most of us have the experience of students begging us for clemency regarding a missed class or assignment, or requesting an extension due to extenuating circumstances. Tokens reduce this drastically. I’m no longer in a position where I have to judge whether an excuse is legitimate or good enough to warrant extra time (something I’ve never thought was my job), nor do I have concerns about fairness. Students simply email me they are using a token for a 24 hour extension or to excuse an absence, with no questions asked.

Here is a summary of my favorite things about tokens:

  1. I don’t have to award or track extra credit.
  2. I don’t have to track participation.
  3. I don’t have to judge student requests for extensions or excused absences
  4. Students can’t earn them without coming to class, so it can also eliminate the need to track attendance if you want.
  5. Gives responsibility and flexibility for students to decide how to spend them
  6. Tangible reward that incentivizes good work and professionalism.

My students told me that they LOVED the tokens. I’m going to use them in every class from now on, even if I don’t use specifications grading itself.

#2 Increased transparency is valuable

I recently attended the POD Network annual conference and was fortunate enough to sit in on a session on the TILT project led by Mary-ann Winkelmes of the University of Nevada, Las Vegas. TILT—or Transparency in Learning and Teaching—aims at increasing student success by improving the transparency of assignments. The research shows that students in the study improved on a host of indicators (including feelings of belonging, academic confidence, and skill development) that led to real results on academic performance and retention. These results are strongest for underserved students—particularly first generation, low-income, and minority students.

I didn’t know it at the time, but the work I was doing on specs grading tapped into this transparency movement. The basic idea is that assignments should clearly state the purpose of the tasks, outline the specific tasks students must complete, and lay out the criteria for success, with annotated examples of work for students to evaluated.

My assignments don’t go quite far enough to qualify for TILT, but they were certainly an exercise in increasing transparency for my students.   I had to be very clear on exactly what I expected and what students had to do to earn a passing grade on the assignment. This required me to revise my assignments specifically to improve the questions I asked—such as saying ‘provide two specific, APA-cited examples’ instead of just saying ‘explain’.

Transparency is my new obsession, so expect to see more on this topic in the future. For now, I’m really pleased with how much insight this experiment with specs grading has given me in how NON-transparent my assignments were before. It is definitely affecting how I craft assignments in all my classes—even those I haven’t adapted for specifications grading.

#1 Student feedback, instructor reflection, and flexibility are absolutely essential.

 As chronicled in ALPS, I had to make a LOT of changes to the syllabus throughout the term in response to student feedback and my own sense of where I had made mistakes. For example, I had to reduce the number of required homework modules from 21 to 17, as otherwise a single missed assignment resulted in a student completing failing the course. I also changed the rule to allow students to count optional assignments toward their total without having to attempt every required assignment (the original rule). Moving forward, I’ll simply require a certain number of assignments, and let them pick which ones to do. As mentioned earlier, I’ll also be changing how I do the deadlines in the class, having set deadlines of every two weeks, on Sunday nights, with the modules due depending on what we covered during that time period. That will let me grade them before class on Tuesday, so I can review with all the students the collective points of success and weakness.

The point is, I didn’t just set the syllabus and follow it, but consulted regularly with the students to ensure that they were not disadvantaged by this first experiment with specs grading. Part of my purpose in trying specs grading was to decrease anxiety for the students by putting their grade in their hands and being very transparent in what they needed to do to earn their chosen grade. Some of the early decisions I made INCREASED their anxiety, so I had to be responsive to that while still preserving the integrity of the course.  This flexibility and constant feedback from students, combined with my own assessment, ended up being essential to the success of specs grading in my course.

So there you have it. I would judge this first experiment with specs grading a success. Now that I understand the system better—its benefits AND challenges—I’m ready to see if it has any impact on learning, student anxiety, and grades. I’ll be putting in an IRB so I can gather data on my next attempt, so be on the lookout for some real results in the future.

In the meantime, if you or someone you know is trying specs grading, consider writing for ALPS! We want to hear all about it.