For a teacher beginning the process of Standards Based Grading (SBG), one of the biggest mental and practical road blocks is the answer to the question, “How am I going to take all these standard based scores and turn them into grades for students?” Here is what my grade book looks like at any given time:
|From Blog Photos|
All my grades are on a five point standard-based scale (5=Outstanding, 4=Good, 3= Competent, 2=Approaching Competency, and 1=Unacceptable), where a 3 represents a satisfactory passing grade for the end of the year (whereas a two is satisfactory wt the beginning of the year). Therefore, a 60% is a solid grade in my class, and cannot correspond to a 60 on the transcript, which is a failing grade. Likewise, it is extremely rare that a student will be excellent across all my standards, so a 90% or above is essentially impossible. Yet every nine weeks, I have to provide my students with a numerical percentage grade which goes on their transcripts that must correspond to a traditional grade scale.
In an ideal world, we would just give students and parents direct, unmediated feedback in terms of how students preform on various standards so students can have an honest assessment of their strengths and areas where further growth is required. The unfortunate reality though, is that the overwhelming majority of teachers must report a letter or a number grade multiple times throughout the year.
The solution I will propose here can work in one of two ways:
- Teachers only report grades based on standards on some finite scale throughout the marking period, and convert standards to a final numerical grade at the end of the marking period (which is what I do in my classroom)
- Teachers convert each convert each individual assignment to a traditional number or letter grade, which allows for teachers to try out SBG without completely changing everything they do.
For teachers who are new to SBG or who might want to start by just getting their feet wet, I would recommend starting with the latter option, though I think the former is a more logical implementation of SBG as a system of assessment and reporting. There are other solutions to this problem out there which do the job simply and effectively, more abstractly, or that aim to ensure motivation and high expectations across the board, all for the first situation, but I’ve yet to see anything that can work for both.
The Bump and Space system below was shared with me when I was a student teacher by a 30+ year teacher from Classical High School in Providence, RI, whose names escapes me, sadly. He used this to convert rubric grades from essays to letter grades in his gradebook. While at first glance it will seem rather unscientific and arbitrary, he said he had never had a student complain in the years he used it so long as he took the time to explain it to students. I have had the same experience. Further, and more important, I have found that in the end it has always given me a fair and true assessment of where students preformed relative to each other. Here is the process, followed by examples:
- Determine the total number of standards based points all students earned for the marking period or assignment
- Determine an appropriate level of rounding or grouping of scores
- Remove outliers (which are nearly always students who have numerous missing assignments because of attendance issues)
- Create a frequency bar graph all students/ scores
- Look at the bar graph and identify the bumps and the spaces
- The bumps will tell you the number of points for an A, B, etc.
- The spaces will tell you where the cutoff should be between an A-/B+, B-/C+, etc. If there are not obvious spaces, look for smaller bumps
- Identify the range of scores for a letter grade
- If necessary and/or desired, decide within each range what grades get plusses and minuses. Also if necessary, convert from letter grade to numerical grade
- Preform a common-sense check to make sure that your grades make sense given students’ performance on the task
And that’s it. Having done this for seven years, it still sounds ridiculous to me, but trust me, it works. Here are two examples:
Example 1: Individual Assignment Grades
- This writing assignment was assessed on a 1-5 scale for 5 standards. Therefore, possible scores ranged from 5-25.
- Step 2 is usually unnecessary on an individual assignment.
- The outliers in this case are the zeros, which were removed.
- Here’s the graph I ended up with:
From Blog Photos
- I see bumps 10, 15, and 22 points.
- I decide to make 10 my C, 15 my B, and 22 my A. (Note: this assignment was at the beginning of the year, so I was okay with students in the 2 range. By the end of the year, I would have made the decision to consider the 10 a D or F).
- Because there was not a wide distribution creating many spaces, I look at smaller bumps for the scores for plus and minus grades. I decide on the following thresholds: A/B = 19 and B/C = 12.5, C-/D+=9 (we do not have a D+ grade).Identify the range of scores for a letter grade
- I end up with the following chart of scores: 0-7=F, 8-9=D, 10-12=C, 13-19=B, 20-25=A
- (I don’t have plusses/minuses)
- I plug these in with names and a common sense check pans out.
Example 2: Marking Period Grades
- From a marking period with 500 total points during the third quarter of the year, I export my point totals from my grade book to a spreadsheet
- I round to the tens place
- Students who earned less than 100 points, meaning they averaged below a 1 which is the lowest possible score for a completed task, are removed. Those students earn F’s.
- Here’s the graph I ended up with:
From Blog Photos
- I look at the graph and see there are bumps at 240 points, 330-350 points, 400-410 points. I note there are not many spaces in this situation.
- Based on the bumps, I decide to make 240 points my D, 270 points my C, 340 points by B, and 405 points my A.
- Because there was not a wide distribution creating many spaces, I look at smaller bumps for the scores for plus and minus grades. I decide on the following thresholds: A/B = 375 and B/C = 300, C-/D+=255 (we do not have a D+ grade).
- I end up with the following chart of scores: 0-239=F, 240-254=D, 255-299=C, 300-374=B, 375-500=A
- (I don’t have plusses/minuses)
- When I put the grades back with names, a common sense check tells me this system works.
Does this actually work?
Yes. It sounds very counterintuitive, or at least very unscientific, but in the end I have found it always gives me a fair ordering of students. Granted, I wish I didn’t have to give grades that only serve to make it easy for others to compare students, but that is my reality.
Is it fair for students who earned different number of points to have the same grade on their transcript?
I haven’t taken any statistical courses since AP Stats in HS, so I wont get into the technical argument here (though I’m sure it would be easy to make), but i don’t think there is any teacher who would claim that there grading system is so accurate and specific that the difference between a handful of points is actually meaningful. By placing the grade cutoffs in the spaces, it ensures that no one ever misses a higher grade by one or two points.
Why can’t you just simplify this and do something based purely on reporting how students did on each standard like all the teachers you linked to earlier?
Two reasons. First, there is a cumulative whole of writing that is more than the sum of its component parts. No one would break down War and Peace into its various components and only talk about those; we do this for students because it is helpful for them to see where they need to improve. But at the same time, we need to help students understand that the collective whole of their writing has meaning beyond the individual parts. Similarly, students need to understand that if spelling is poor, it lessens the overall affect of the piece as a whole. The second reason is that in the history classroom, I need to asses both students’ skills and content knowledge. Whereas I will assess certain skills multiple times throughout the year using a true SBG method, the unfortunate reality of teaching a survey history course is that students will not have opportunities return to certain content area; I simply have to move on to teach evereything necessary for the state exams.
Are all standards created equally?
I do not believe they are, and I think this would probably be true for most humanities teachers. While i do grade every piece of writing I assign for spelling and grammar, in my history class, I emphasize this less in my grades because i do not provide spelling and grammar instruction. This system makes it really easy to weight certain standards.