Report Card Grades
When I was a principal it was difficult to develop report cards. What drove me crazy was that I didn’t know what the grades meant. I reviewed report cards. I talked with teachers about the semester and what went into the grades, but I could never confidently say that the grade truly represented what a student learned in class.
In trying to make some sense of it all I brought it up at a professional development session. I brought an article on standards based grading and another one on effective grading practices, and afterwards I asked the staff to list the most important elements of a good grading policy. We eventually came to a consensus and agreed on a new and improved grading policy.
Except, it still didn’t work.
At the end of a semester I looked back at the report cards and I still didn’t know where they came from or what they meant, because there was no information other than a number. I sat in on parent teacher conferences and I still didn’t know for sure we were informing families with valuable information, and by the way I had some phenomenal teachers. I had teachers who could tell you everything about their students. They could connect personally with families, and families felt like the meetings were very positive.
For me there was still something missing.
Standardized Test Data
New York State has very similar issues when they report test grades, and so I set out to understand how students earned the test scores they received.
One thing that always irritated me, even as a teacher, is the way educators analyze the test results and how they speak about students afterwards. My personal pet peeve is the phrase “moving the twos to threes, and the threes to fours.” It gets under my skin, because I know that no one knows what that means. No one knows how a student earned a score of 2 and they certainly don’t know how a student got to a level 3. It’s not the same idea as a teacher reviewing an essay her student just handed in and figuring out how that student will score higher on the rubric next time.
New York State scores may score the grade 3-8 exams on a 1 to 4 scale, but there is a lot of math that is used to get there. First, there is the raw score, which is the actual number of points a student earned on the test. Second, that raw score is converted into a scale score. I’m sure there is a sound statistical reason for this, but I don’t know what it is. Finally, the scale score is converted into proficiency level. You could also go one step further and get the proficiency rating which will tell you how close a student is to the next level.
Below is an example of this conversion on the 2016 NY State grade 3 math exam.
To earn a level 2 on the exam a student could get as little as 43% of the test correct. The label the state and most schools assign to level 2 is approaching standards. To earn a level 3 a student must earn 66% percent of the points to be labeled as meeting standards. Educators are very accustomed to grading students by percentages. When we think in terms of percentages a 43% would not be approaching anything: it would be failing. We would see major deficiencies in student understanding, but converting the raw score to a level 2 and labeled “approaching standards” we believe the student is much closer than they actually are. This whole process seems to be done to confuse everyone about how poorly students are actually doing.
Using the Scores in a Meaningful Way
As I discussed in my last post, why data can’t drive instruction, a number can’t help us plan our lessons. We are told however that these tests should drive school goals, and they give schools feedback on student performance; however, it wasn’t until 2015, two years after the first common core tests, that educators could even see the test questions. When we did get to see the test questions after the test we didn’t get all of them. We certainly never get to see student work.
How can we use the test data effectively if the true results are so hidden in the numbers and we never get to see student answers?
Schools need better insight into the tests, or any assessment, to be able to use it effectively. NY State does make individual student reports, and they’re colorful, but still not helpful in explaining a student’s true performance.
The truth is that in their current form, these tests are evaluative. They are a report card grade. They are not designed to help schools learn about student performance. They are a measuring tool. They are helpful way for outsiders to make judgments about a school’s proficiency. That might still be valuable, but it’s not to give students, families, or schools feedback.
It takes a lot of work to help schools analyze test results, mostly because the useful information is in an excel spreadsheet. Excel is a powerful and useful tool, but it is not a great communication tool for educators.
I always start with the June Instructional Reports and below is a small part of a sample report.
It gives a school information on how students did on each standard covered in the assessment. There is some useful information in here, but there is no need to train teachers to be able to analyze this spreadsheet. I try to keep a goal of no more than 7 minutes looking at spreadsheets with educators.
Remember, data can’t drive instruction, so this doesn’t do us much good yet. We have to at least look at the test to start making sense of what the numbers are saying.
Aligning with the Assessment
Let’s hypothetically say that a school’s data sheet says that students had trouble on standard 3.NF.2b, which is about fractions on a number line, and they wanted to look into it further: they would go back to the test questions. In 2015, there were two released test questions for 3.NF.2b.
So now we have some numbers about performance and we have two questions. I’ve led professional developments that focus on connecting the instructional reports to tests questions, and as soon as we get to this point many teachers are sure they know exactly why students didn’t do well.
Subscribe to get our latest content by email.
I still have no idea. Not for sure, but they often do.
Schools, I should add, also get separate performance for multiple choice and short response questions. The vast majority of schools will notice that their students did much better on the first question than the second.
At this point in the analysis we have hit a dead end. All our next steps are speculative. Some hypotheses could be that the curriculum didn’t assess this standard in this way. Another could be that the topic was covered early in the year and it wasn’t scaffolded back into later units.
The reason we don’t have definitive answers is because we don’t get to see the student work. We don’t know how many students weren’t able to start the second question. We don’t know how many could label the number line, but could calculate the difference. We don’t know a lot about what was on student answer sheets. We do get information on the answer choices students picked from the multiple choice.
Forming Next Steps
The only option available now is to find out from our current students. We would have to investigate what our students are able to do on specific problems when they come up in the curriculum and then address their needs with an assessment we’ve made ourselves.
NY State is essence requires schools assign their own assessments to figure out their performance on the state tests.
Let’s recap all the work involved with this test. Students have a raw test score converted into a scale score and then a proficiency level. This confuses everyone about the actual performance of all students. Schools then receive an instructional report, they also get one that looks the same but contains individual students, that they need to analyze all standards against multiple years of performance and against local district wide results as well as class performance within the school. After analyzing the test results they need to then review all relevant test questions. Finally, they have to guess why students did poorly.
By the time this is done schools are well into the fall and have their own baseline assessments and class assessments to also analyze.
Schools need assistance in making assessments relevant.
After my attempt at understanding report grades I did come to a conclusion that allowed me to move on. I determined that a grade only measures the ability of a person to give another person what they ask for. In this way, this might be a valuable measure. Students who can understand what each of his/her teachers want and then produce what each of his/her teachers want to see have a valuable skill set.
I think students taking standardized tests have the same skill. They have to understand what is being asked of them on a test they have never seen that may contains material they can’t predict.
We ask teachers to prepare them for these exams. I help schools break through the numbers to get to the actual questions and the real assessments, but it takes a lot of effort to figure it out. If the purpose of the assessment is to help students and schools then it shouldn’t be so difficult to accomplish.
Subscribe to get our latest content by email.