Comments of the week: Teacher's-eye views of Regents scoring

It’s that time again — the time for us to highlight some of the week’s informed, thought-provoking comments from our readers.

This week, we especially appreciated the comments posted on our story about a new grading system for high school exams that the Department of Education quietly piloted this year. The system, called distributed scoring, requires for the first time that teachers not grade their own students’ Regents exams, which are required to graduate.

We reported that some teachers worried that distributed scoring would depress scores, sometimes unfairly. Several readers backed up that hypothesis. For example:

…My school’s passing percentage went from 70+% to 20% (!) when they were graded by another school. No doubt there is biased grading going on there.

Some commenters said they thought students from struggling schools were likely to feel the effect of the new grading system most:

While it clear that my whole grading team were really rooting for the kids whose papers’ we scored (even without knowing them), I worry about how my students would fare if they’re papers were graded by teachers who were used to only top tier students. I suspect that a Brooklyn Tech or Stuyvesant teacher (or a teacher on the north shore of Long Island if the pilot expands state-wide) would expect much more from a response for it to earn credit then would most city teachers. I saw two teachers from a selective high school grade papers from a struggling school and give a whole school 0 and 0 on questions 26 and 27 of the ELA. Makes it almost impossible to pass. The rubric would have dictated at least one point for most of the responses. A colleague and I were grading the essays of the same envelope and gave 3’s, 4’s, and even one or two 5’s to the same students. Their incompetence will likely lead to a school’s closure.

But other readers said their experience suggested that higher-performing students would be penalized:

Our school was chosen.  The passing % didn’t change but mastery fell, a lot. Shockingly low. It wasn’t the students work, it was the graders. A lot of schools gave low scores to ‘better’ schools. It was known and talked about. I heard that it was planned out in some schools. Some schools want their tests regraded and some kids are demanding it. I have a feeling that some of the “good’ schools are going to be screwed with this system. I’m in a better school and I was rooting for the weaker kids from the other schools. I was sad when I got blank papers because I couldn’t even look for points to give. Pretty much all the teachers at my school (and, in all subjects) wanted the kids we graded to do well. Sometimes a weaker school student didn’t express themselves as well, but they knew what they were talking about. We gave as many points as we could. The problem was that some of the other schools resented grading the papers of stronger students. They complained about reading longer essays, having to work through math problems, etc. Our students often paid for it.

A commenter succinctly offered one reason why distributed scoring might not solve the problem of scores not always accurately reflecting student performance:

I think there is an interesting point about incentives here. Originally there were concerns about teachers inflating their own student’s grades, but given the competition inherent in progress reports, there is an incentive to deflate the grades of students from other schools.

And several comments included suggestions for the city and state as distributed scoring is required for all schools next year:

It does not seem like it would be that difficult for the state to maintain a website with a list of all questions asked by teachers and the responses given. This would be tremendously helpful to teachers across the state and help ensure that grading is consistent. As a math teacher who participated in the electronic grading pilot this June, I have one major complaint. Open-ended math questions were scored by only one rater. There was no check process in place at all. I strongly suggest adopting the model used for grading essays in Global History and English: two raters independently score each essay with a third rater used to settle discrepancies.