This is going to be a super short post, directed to you SBGers. One of the things we talked a lot about this summer at PCMI was feedback! [1]
Here’s my question about feedback and SBG.
Let’s say you give a skill test with 3 or 4 skills on it, and you want to give feedback. Let’s also say that you give a numerical score for each skill. Besides that score, what sort of other feedback do you give on the test?
I’m wondering how explicit you get with identifying the issue, and what the goal(s) of you feedback is/are. Do you…
…circle where the student first started going awry?
… write a sentence (or two! or three! or more!) using words to explain where the work went awry?
… focus on something other than where the error crept in?
… give the total solution, written out for ’em?
… give some broad thoughts, encouragement?
… do something else?
Basically, I am super curious what your feedback looks like — in addition to the numerical score. The reason I ask is this… we talked a lot about feedback, but we didn’t talk about what effective feedback looks like. And for me, I don’t tend to give too much feedback. I’ll do some circling, some encouragement, I might write a sentence or two if the error is easily identifiable, but then I expect the student to figure the rest out on their own, or by working with a colleague.
And at first, as we talked more and more about feedback at PCMI, I felt a little guilty for the paucity of my feedback. But the more I thought about it, the more I thought: I think what I do is actually good for them. I don’t want to do all the work for them — they have to be a little more proactive. But I also wonder what and how others do, to see if there are things to steal to improve!
I wish I had scans of actual assessments to show you and have you give suggestions about how you’d mark them up, but in lieu of that, I’d like to hear how you do it (or would do it).
[1] One thing that struck me was how feedback given with a grade tends to be ignored by students, but feedback given without a grade tends to be really considered.
Continuous Everywhere but Differentiable Nowhere 
Currently, my skill quizzes are always on quarter slips of paper. This works well for paper saving (since I’m the one who has to buy it all) and keeping me in line for only doing bite-sized (let’s call them fun-sized) skills. As I ponder giving better feedback and having students work on their own metacognition and take more responsibility for retakes, I am thinking about going to a half page format where the questions and their answers go on the left column, my comments go in the middle column, and their corrections/comment responses go on the right. I might even only enter in perfect scores and Incomplete into the online gradebook for the first go-round but put nothing on the page itself. Maybe put in a mini column for students to put their predicted score? Hmm, things to post about–thanks for starting the discussion!
The bit about the grades and comments is I think because students have become accustomed to their work being reduced to a single score. That’s the measure they’re expecting at the end of the course so that’s what’s important to them.
Hi Sam,
There are studies that support your observation that, when both scores and feedback are given, the feedback does not help. Ruth Butler did the most famous version of the study.
Bret
I do a mixture of a lot of your bullet points, depending on the mistake. I basically just go through the test and comment, write feedback, circle things (I think usually with notes, but I’ll pay more attention this fall). Then I go back and fill out a score sheet. On the score sheet, I sometimes write a couple of words (or sometimes more) summarizing the problems they had with particular skills, or an overall message. And sometimes I take those notes and enter them into ActiveGrade with my scores.
I think they read more of my feedback on tests now because they correct them for an opportunity to try again (instead of for points back), so they are more invested in figuring out what went wrong and how to do it better.
I agree that with a number grade, they looked at the number and not the feedback. But now even with a standards grade, since it is only yes or no, not a percentage that they can settle with, I think they read a lot more of it. Now a grade that isn’t perfect has a built-in feeling of Do Something rather than a built-in feeling of Darn or Oh Well.
I want to see if I can get some feedback from my kids this year on the feedback I write on assessments. What do they see as most helpful? What do they pay the most attention to? Etc.
This is a really meaty issue (feedback), Sam. Looking forward to this conversation!
When I was last in the classroom a year ago, I ended up not assigning numerical scores on quizzes. I viewed homework as the first practice opportunity – a time for questions, a time to make mistakes, a time to work with classmates, etc. I viewed quizzes as the first “structured feedback opportunity.” I circled where errors were, wrote comments, wrote questions, and then gave students an idea of how close they were to understanding each standard by circling where they were along a lykert scale. See this quiz for an example: https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnxuZXdlbGxmb25kYWZhMjAxMXxneDo1OGI0NWRhNDlhN2ViZjY4
Finally, on the tests, I gave numerical scores for each learning target. Hope this helps!
On tests I try not to give too many hints since I want kids to correct them, I will circle the error if it was arithmetic or if they have a glaring misconception I will write a comment. For what I call investigations (mini projects) I try to record a grade in my gradebook but not on their paper, this never seems to last all year though… This year I have stamps! One gold star and one “spend more time”, I hope to stamp and write comments while only recording a grade for me just in case they don’t resubmit.
On something that was skill based I used a pretty explicit scoring guide/rubric/lykert scale…whatever you want to call it.
5: Perfect
4: Good concept; procedural flaw
3: Conceptual flaw; no procedural flaws
2: Error in concept and procedure
1: No evidence of understanding
I pointed out procedural errors by circling them and left written comments/questions when it came to concept.
I do have a question for Matt, though. Is there a difference between circling where a student is on a lykert scale and giving a numeric score? If we are moving from quantitative to qualitative grading, the numbers simply correspond to levels of quality anyway.
@David
You asked, “Is there a difference between circling where a student is on a lykert scale and giving a numeric score?” I can’t accurately speak for the practice in general, but the lykert scale I used was tied to words “I have no idea…” all the way to “I get it” rather than numbers. It makes more sense when viewing the linked quiz. This practice, for my students at that time, seemed to help them focus on the feedback rather than “what grade did I get?”….especially because it wasn’t going in the grade book no matter where I circled on the scale. Another variation used was asking students (after completing the quiz) to circle in pencil where they felt they were on the lykert scale for each standard. I wrote my comments by each of the problems and then circled in pen where I thought they were along the scale. The gap, if there was one, was a good conversation starter.
Sam,
I have not decided if this would work with HS math students but the most effective feedback I ever received on work came from one of my Profs when I went back to school recently. His attitude about writing assignments was that EVRYTHING was to be re-written at least once. The first time round all he did was put a mark next to a line/passage that needed something to fix it. No indication of whether it was a structural problem with grammar or a logical problem with our argument. I know that I have never analyzed my writing as much as I did then. I wonder if something similar might work with motivated HS students.
I teach a whole master’s course just on giving effective feedback to students, specifically on their problem-solving. Which is a little different than on a skills quiz, granted…
I usually focus on reflection questions, trying to aim for something metacognitive, thought provoking, open-ended, balanced between encouragement and critique, and specific to the learner and task.
The reflection question I have for this group is: we’ve identified lots of kinds of feedback (asking questions, modeling, circling key steps, giving encouragement, describing the error, etc.) and I wonder if there are different moments/kinds of errors/tasks when each one is appropriate. What kinds of tasks does it make sense to just circle? When does a kid need a few words of encouragement? When is telling better than asking? When is asking better than telling?
I’ll mull on these and post my thoughts… Interested to here from others too!
Years ago I observed that students would often ignore feedback and focus on “the grade” which is one of the reasons I switched to SBG. I tried at first not giving a grade at all until the end of the semester, but that didn’t work well. The feedback I got from the students was that–despite other feedback–they didn’t have a good sense of how they were doing in class, and how important it was to put in more effort. In my experience, numeric feedback on specific standards that can be retaken is effective. Students first look at the numeric score, but then they want a better score and ask themselves what needs to be done to achieve it. So they’re interested in other feedback. Common mistakes I just circle and then we have a discussion about them in class the next day. For uncommon mistakes that I know are unlikely to come up in class discussion I will write out a few comments.
On assessments I will generally add questions and comments like “where did this term come from?” or “check your algebra” or “keep trying this method until you’re sure it doesn’t work”.
The metacognitive task for the students comes when they ask to reassess. Like many using SBG, I ask students to describe 1. the source of their error or misunderstanding and 2. what specific steps they’ve taken to solidify their skills.
So, I don’t worry too much if a student focuses on 3/4 initially; he’ll have to read and understand the written feedback and look over his work before he can change the number.
With written feedback I always try to
1. Tell them how they did in reference our criteria (usually the scales/rubrics/whatever you want to call them). I’m fine with using some of the exact wording. “You did fine on graphic motion, but need to improve interpreting them)
2. Give them a next step (this part is CRUCIAL)
In some book I remember seeing an example of a teacher marking up a paper and basically doing all the work for the kid (pointing out every mistake, correct spelling, etc). It stayed in my head that feedback should get them thinking and then doing.
I’m also in favor of not including the score but like Groupact said, they need to know where they stand. So remove the score from the assignment itself and report how they’re doing by standard at a separate time than returning the assignment. Just like how you probably are required to hand out quarterly progress reports. Give them standards reports every so often that are separate from the assignments themselves.
I generally give a quiz like this twice a week and it takes between 5-10 minutes to complete.
I do something similar to what David does on these types of straightforward, short skills quizzes. But, I use just a 3 point scale. My quizzes (of this nature) have 2-3 questions, so it’s just a 6 or 9 point quiz. Each individual one is minor in an overall quarter grade, but together (when given 2x a week), the scores definitely affect quarter grades.
I find that to consistently do skills quizzes twice a week,I need a very straightforward, quick way to grade them so I don’t put any comments on the paper. Like Ashli, I don’t use a full sheet of paper. We have a surplus of 8.5×14 paper which teachers cut down to 8.5×11. There are hundreds of these extra paper tails (strips) laying around which are 8.5×3 and they are perfect for this.
Instead of putting comments on the paper, my feedback is given the next day, during class, during a 10-15 minute period which happens as soon as our warm up is completed (as the warm up is often unrelated to the quiz from the previous day).
As a class, for feedback, we do to things::
1) I cover up the name of the student and place 1 quiz on the document reader. I choose one which has a good, meaty error that will make for good discussion. I label their work (in a bright colored pen), “Step 1” “Step 2” etc. We discuss what the student did well and then discuss in which step did the student first make an error. This conversation takes 5 minutes.
2) After this whole class error analysis, I meet with a small group of students who all had either a similar misunderstanding or sometimes who had a similar thing they did well and I want to talk more about that in some way. It goes something like this:
*Students with a perfect score find their names posted on a poster as the “Divos” and “Divas” for that day. (as a sidenote…the divos and divas are NOT the same people day after day…)
* I call a group of students to meet with me in one part of the room. Sometimes we push desks together, sometimes we don’t (depends on my mood). We always use individual white boards. Meanwhile, the divos and divas’ job is to go around and help the students with whom I am not meeting. Their quiz corrections are due by the end of the time that I meet with my group. If not completed, they are due the following day as homework.
Sometimes I meet with the divos/divas and the rest of the class works together to correct their quizzes.
I mix things around a lot in terms of who i meet with and I find that after the initial leaning of the routine, students don’t complain about being with me or not being with me.
We have block periods, so we’re not as pressed for time as on a traditional schedule.
Oh, and P.S. My name links to a blog which I have just created,but haven’t yet published, so I’m not quite sure what it will look like if you click on it.
I definitely circle where things went wrong. I usually also write more, eg a short explanation or an example of how to solve, but I’ll tend to write more on the first ten or so and then write less as I get through the stack. Not really best practice but it’s what happens.
The one bit of that I’d defend is that once I’ve seen the same error a number of times, I’m less likely to write much for it as I’ll almost certainly bring it up in class next day.
I think the purpose of feedback is to communicate to the student where their problem is, what to do to gain the skills to avoid the problem (or a hint in the right direction), and a realistic idea of how long it’s going to take them. That sounds like a lot, but it’s showing the student the line from A to B and everything they need to travel it. You might write something like “spend 25 minutes on such and such and these kinds of problems will be cake.” It’s a short statement with plenty of information for the student to roll with. If you think they can find the problem themselves, give them a specific activity they should do to find the problem, not just vague encouragement – there’s not enough information in plain encouragement if you want them to get something done.
If I am giving a numerical score (0-4 scale), I rarely give feedback.
I find that most students (even in Algebra 1) are curious enough about their own scores that they will attempt to find out why they got a 2 or a 3 on a problem. That very act is the heart of mathematics: inquiry.
The only time I have doubts about this are the super struggling students who get 0s and 1s. (And are essentially failing all their classes) They need more than just a number, but if they aren’t motivated to put down anything relevant in the first place, will they be motivated to read my comments?
Your footnote at the bottom resonated for me: “One thing that struck me was how feedback given with a grade tends to be ignored by students, but feedback given without a grade tends to be really considered.” I commented a bit about this very subject in my post, “Doing away with grades.” Here’s a relevant extract:
“I also want my students to pay attention when I do go over the test/quiz, but invariably, when I return the assessment, the students who did well don’t bother to pay much attention because they got the answers correct and the students who bombed are usually so distraught that they can’t focus away from the F grade staring at them to gain anything from my going over the test/quiz.”
I’ve tried an approach that provides feedback BEFORE they see the grade, with some success, but your mileage may vary :-)
Paul Hawking
Blog:
The Challenge of Teaching
Latest Post:
Lies about square roots
http://challenge-of-teaching-math.blogspot.com/2011/08/lies-about-square-roots.html
I use Marzano’s approach to formative assessment/SBG which I realize is not what all of you do as far as recording the scores from assessments and giving a sequence of assessments for an objective during normal classtime instead of a retake after school. Anyway, for feedback I pretty much just write the score (0.5 thru 4.0) at the top of the assessment and then the next day when it’s returned to the student we’ll spend time discussing the questions and referring to that objective’s rubric so the student can figure out what their score means (e.g., could do all of score 2.0 items, but had minor errors on score 3.0 items). I like an idea that a poster wrote on somebody’s blog about doing an analysis of the results the night before so as to share with students some statistics about each question as well as to identify common misconceptions and errors. And of course during this discussion of the assessment is when students complete the graphical learning trend analysis and action planning that is a big part of Marzano’s formative approach.
It’s interesting to read about peoples’ philosophies toward recording scores versus giving some other form of feedback. At my school it’s a component of our system to let the student know what score they earned on a particular assessment (and to record it in the gradebook) and to spend time working with this score to plan what needs to be done to increase understanding. Certainly, many teachers (me included) often give additonal feedback in writing on students’ papers, but this can be overwhelming given the huge number and the frequency of assessments we’re doing for SBG. So, we’ve tried to encourage teachers to use the simplified system of the rubric score–and this works okay if the teacher spends the appropriate time and effort in class responding to this score.
I just had a great idea. I’m sure someone else has thought of it before. I am letting their grades drop by summing the last 2 assessments. They have to get 2 fives in a row in order to master a concept. If their grade stays the same, I am thinking about crossing out whatever # score I gave them and just writing feedback. Because I agree with your last statement that they tend to listen if there is no grade attached.