Do you get the feeling that legislators really don’t know what they’re doing anymore?I’m not talking about the smorgasbord of sex scandals we’ve been treated to, or the logjams in Albany or Washington that indicate our elected officials don’t spend much time staff developing themselves in the basic skills of governance.
I’m talking more about how good these people are at judging what’s best for the profession of teaching: where are they getting their facts, and do they have the time, interest, or skill to really assess what their advisors are telling them.
At Lobby Day up in Albany a couple of weeks ago, a whole bunch of us went in to speak to our local assemblyperson, who, like many others, was not even in the building. Instead, we got to talk to — for want of a better word — a kid, an aide in his early 20s, whose face showed he hadn’t a clue what we were talking about. He took no notes, remaining alert, tense and frozen. He was probably praying he’d be able to respond intelligently to a couple of our points and remember a handful of things he could report back to his boss on. It was decidedly uncomfortable, for him and for us, who had given up a day to make the trip.
I don’t trust that elected officials know enough about education to legislate it, and I don’t know how we can get them to absorb what they need to know about teaching in underfunded schools and hostile environments.
Jennifer Medina’s reported in the NY Times last Tuesday that the Assembly passed a bill the week before that would keep school districts from linking teacher tenure to the test scores of students. Not bad for starters, but she went on to mention that the School Board Association director thinks legislation to toughen tenure standards keeps getting “watered down.” The rules on tenure passed last year, applying to teachers hired after this coming July 1st, say that tenure decisions must include, among other things, “an examination of how well teachers use data.”I can’t imagine any legislator knows how to use data in the classroom. What I can imagine is how people whose business it is to use data try to persuade them that data is essential to good teaching.
Here’s an example of data in the classroom from when I was in high school some 45 years ago, before computers were invented and way before BloomKlein and their ilk could squander $80 million on computer projects like ARIS.
A teacher gave a test. No one got 100, even the best and most reliable students. He took the highest score and raised it to 100, then put all the others up the same number of points. Looking us square in the eye, he then told us that (a) he had misjudged our capabilities, and (b) he was disappointed in us. We got the point that it was not so much the mark that counted, but that our standards for ourselves were too low. Now, that's what I call a good use of data!
We don’t need test data to tell us how to teach. We need the time to assess our students’ work and speak to them individually about it. We need the time to look over their shoulders, offer comments in the margins of their papers, read over the details of what they write and the computations they make, we need time to correct their spelling and their grammar.
This does not come from legislation on the use of data. This comes from legislation on class size and making sure that when teachers are in school, they’re helping kids acquire skills instead of doing data entry and custodial duties or attending staff development sessions of negligible importance.
I’ve pretty much been staying out of the school scores discussion, because as an arts teacher I don’t have hands-on experience in the subjects that do get a lot of test prep (and a lot of tests!) and because there are others, like Eduwonkette, who are really good at peeling back the layers of KleinLieb with precision.
But, I have joined in on a couple of occasions, because this whole “progress" issue has such a stink to it that even an electives teacher like myself is under a moral obligation to look where it’s coming from.
What I can do, and have done already, is ask some questions, and I can also be grateful for the responses and happy that I have a site where I can post them.
The first question I asked was on a couple of listservs was this:
I've been hearing a lot of parents and teachers saying their school got an A or a B or a D, and I'm not sure they've understood that these letter grades are assigned to percentiles, not scores in the way we use them every day to grade papers.
In the DOE school grading system, an "A" does not mean 90 or above, but rather that a school is in the top 15%, and so on thus:
85 or above = A
85 - 45 = B
45 - 15 = C
5 - 5 = D
Below 5 = F
This table is from a printed report called "Educator Guide: the NYC Progress Report", and it was given to me by a teacher trainer who takes administrative classes with superintendents. She got it there. When we checked the online version today, this table does not appear. Another table is used, on p.29. It gives rather:
67.6 - 100 = A
48.8 - 67.5 = B
35.1 - 48.8 = C
28.9 - 35.1 = D
0 - 28.9 = F
Apart from the fact that neither of us [editorial comment: nor a reporter who subsequently contacted me about this] can understand why one table appears in the printed version and another online (maybe one version is for HS and another for another level, I'm not sure), I am very troubled by the ambiguity the DOE has created by using LETTER grades in the first place. The general public looking at a letter grade does not think of statistics and percentiles. They see what they are used to, that A is usually in the 90s, B in the 80s, C in the 70s, D in the upper 60s, and F below 65, or failing. I am sure a less ambiguous system could have been used so that no one could possibly confuse the percentile groups they've labeled A - F with the normal and widespread meaning of letter grades.
More obfuscation and confusion on top of the contortions of the computations themselves.
The teacher trainer also told me this also: "The progress report is an attempt by the DOE to show NYS that NYC is meeting the mandates of NCLB. Thus the huge number of grades that seem above average (A's and B's) but are actually pretty darn low."
I do admit to not knowing a whole lot about this whole grading project/fiasco, but my reaction is visceral, that I believe there is a purposeful effort on the part of the DOE to misinform the general public. Can someone tell me if I am at least getting the facts right, or is there something I am not understanding?
One response came from Gary, who not only linked to a parody he had just written, but drew my attention to Leonie Haimson’s 2004 testimony that included a section on “Creative Confusion.” Very helpful in understanding the general ideology, but not specifically dealing with the weird use of letter grades that represent sets of percentages far different from what the hoi polloi are used to, like A=90s, B=80s, etc.
I asked another question in November, this time on comparisons of test grades from a long time ago to the present. I was basically interested in finding out if there’s been a measurable “dumbing down” of the population over time.
Has anyone ever administered a test, let's say for 8th grade math, written in the 1960s to a current 8th-grade class to see what kids now and what they knew 40 years ago?
This led to some back and forth on how some tests are somewhat comparable over time (e.g., the Long-Term Trend NAEP, explained by Ravitch) while others (like the city and state tests) are not. I looked at the NAEP site as she suggested and found you still cannot get very accurate long-term comparisons, since accommodations were allowed in some years but not in all.
Yesterday I read a post on NYCEducator that brought to mind some other questions about testing and the obfuscation thereof:
Kids taking a certain exam this year, say, will be different from the kids taking the same exam next year, right? So what is the point of comparing this year's scores against next year's scores in the first place? Different set of kids means different set of test-takers. What good does it do comparing one set of test-takers against a later set?
Shouldn't progress be measured by comparing how well the same set of kids do from one year to the next? I know that's not really possible, so what's the point of these comparisons anyway?
And here are some of the answers to those 5 questions, first from xkaydet65:
If you deal with the Elem and Int schools you can follow a kid's progress. Does he go from a 2 to a 3 or vice versa (BTW a change of 2 mult. choice answers can change a high 2 to a 3 and a low 3 to a 2)
For you HS folks there is no way to compare results. The tests are different and the kids are different. Yet some people try to do that and they don't work for Bloomklein. Colleges trumpet increases in the SAT scores of their frosh classes from year to year even though the exams change. This is probably the paradigm that the DoE is working with.
Then from Schoolgal:
Elementary and Jr. HS students can be tracked. And it is true how one wrong answer can be the difference between a 2 and a 3.
However our principal informed us that the report card measurement will no longer look at the grade as much as the raw score. So if a child got a high 3 or high 4 two years in a row, they will see it as zero improvement rather than meeting or exceeding the standards. That is why schools with very high scores are getting Bs or Cs.
What is a teacher or principal supposed to do with this crazy method?
Schools in the high 80s are pressuring their teachers to bring scores up will only be hurting themselves in the future once they hit the ceiling. Yet the most violent schools got As because their low scores showed some improvement. If I knew how to spell in Yiddish, I would say this was one cockamayme plan. (my apologies to the Yiddish-speaking readers)
[editorial note: the alternative spelling is kakameyme, which brings up an unfortunate comparison between cock and kaka.....]
And from 15 more years:
This business of kids not improving from year to year is a big crock of hooey. If they were taking the same exact test every year, then yes, as they go from 5th grade, to 6th, and onward, then you naturally would expect a child to show growth. BUT, plenty of kids top out in 5th or 6th grade. The ELA in 7th grade is much easier than the ELA in 8th grade –– so it is almost expected for a child to remain static, if not drop a bit if he is a struggling student to begin with. And what if a child is sick the day of the test? There are so many variables to consider, the concept is ridiculous and ill-conceived.
The mantra coming out of Tweed these days is “accountability,” and Bloom/Klein have made much of testing and test score comparisons to convince people that they know how to make schools accountable.
But it is clear that from almost every possible angle – different tests, different test-takers and what side of the bed they woke up on, different amounts of accommodation, different degree and quality of prep, different testing conditions, and even different morality in marking – comparisons of test scores do not stand up to scrutiny or further the cause of accountability.
Test score data is useful for one thing only: Spin. And that’s the farthest away from accountability you can get.
What are they really trying to do with our schools and with our kids?
New link!! Thanks to Ednotesonline, here's a Test you can take, put out by Prof. Celia Oyler of Columbia TC, on the KleinLieb school grading system. Find out if you already know all there is to know about this colossal boondoggle and if it will ever in a hundred million years be an accurate, fair, or relevant tool for judging NYC schools.
Here’s an example of data in the classroom from when I was in high school some 45 years ago, before computers were invented and way before BloomKlein and their ilk could squander $80 million on computer projects like ARIS.
