I have been asked about the “argument” made by Rick Hess that perhaps none of Rhee’s students were actually tested in Harlem Park, and therefore we can’t tell anything about whether Rhee achieved the miracle that she claimed, so therefore the vitriolic blogger Brandenburg is full of it.

His argument is laughable.

The kids who had no test scores reported, according to the study itself, were those who were

(a) absent during the testing time period,

(b) enrolled at the school after February 1, or

(c) were in severe special education categories and thus were exempt, or

(d) students who put their heads down and didn’t answer anything at all, or scribbled all over the test, or some such thing, and thus were given a score of “1”. (I left this reasons b and d originally)

As far as I understand, there were no other reasons to exempt students. Anybody find any such reasons? In any case, NONE of those groups of students are the ones that Michelle Rhee could possibly be boasting about.

I agree that it is awfully suspicious that only sixty-four percent of the students at Harlem Park had reported test scores in 1994-5. No other school in the entire study had such low numbers (the next-lowest percentage is in the mid-70s; most are in the 80s to 90s). Of course, back then, it was rather common for principals and unethical teachers to subtly encourage their low-performing students to take testing week off, or to allow them to put their heads down or doodle during the test, or else to send them to a special room to watch films during testing period, and so on. That raises the scores for the remainder.

I can’t think of any good, legitimate reasons for the low test-reporting rate, and I have also publicly worried, in print, why the number of students tested AND the entire cohort that Rhee and her colleague taught at shrank so drastically from grade 2 to grade 3 at HPE. Such a high percentage of missing scores only occurred at that school, and at none of the other ones, either Tesseract or regular Baltimore Public schools. It makes me think that the principal at Harlem Park at that time was not doing her job.

Keep in mind that Rhee has said and written many times that she AND her team-teaching colleague raised the scores of the entire group of 70 students from the bottom to the top, i.e, from the 13th percentile to 90% of them above the 90th percentile. She uses the word “we” a lot in this regard, and of course, later, claimed that she learned about the scores from her principal, but, of course, conveniently has no records.

Let’s do a little math. There are reading scores reported for 43 third-graders at Harlem Park in 1994-5. The overall fraction of students at the school with scores reported is 64%. That would mean that there were probably about 67 third-graders overall. There is no evidence whatsoever in the report of reporting or testing irregularities at Harlem Park, except for that 64% figure. But 67 kids is more like three classes, not four classes, it seems to me.

But then again, it might just be, if you believe Rhee’s boasts, that she and her colleague taught every single one of them. (After all, 67 is less than 70.)

If you believe it is likely or possible that only 10 of Rhee & Colleague’s students got tested and that 9 out of those ten got scores at the 90th percentile — while all of the other students in that other, non-cooperating teacher got tested and they scored really low, then I would say this:

(1) I have a nice bridge in Brooklyn that I’d like to sell you.

(2) You have no understanding whatsoever about statistics or probability.

(3) Even, if by some miracle, this actually happened, then:

The only HONEST thing Rhee could have said was, “I lost all of my class data for all of my students, lost all of my roll books, and somehow almost every single one of my students was absent on the entire week that the CTBS was administered, and I forgot to call any of their parents to please, please bring their children to school that week. But, of the ten students in my and my colleague’s class that DID get tested, well, 9 of them did really, really great!”

Or else, perhaps Rhee simply found that at some unit of study, she and her colleague taught some unit of study, and over 90% of the students got scores of 90% or better on this teacher -made test (or perhaps it was one that was purchased from some commercial venture — and the company claimed that the test was ‘nationally normed’ or some such foolishness).

My response would be, “That’s nice. Good for you.”

But I would continue: “Michelle, don’t you know the difference between a teacher unit test and an actual, nationally-normed, standardized, comprehensive, end-of-the-year test like the CTBS? Didn’t you take any educational statistics courses at all? And how come you don’t have any records whatsoever to prove this, and yet you have boasted over and over and over again about this miracle that didn’t happen? Are you just full of chutzpah, or are you just stupid?”

So, I am going to revise my conclusions a little bit. It is now a tossup

(a) She’s a bald-faced liar

b) She’s stupid and clueless and can’t keep records, and doesn’t understand the first thing about educational statistics.

Either way, she has no business running any of the following:

(a) a business recruiting long-term subs (NTP)

(b) any school as a principal

(c) any school district as a superintendent or chancellor

(d) advising any governor or educational body whatsoever about anything. Unless that ‘something’ be how to lie and get away with it.

==================================

Aside: It’s no surprise that Rick Hess, who gets paid big bucks to prostitute his intellect for the right-wing millionaires and billionaires who fund the American Enterprise Institute, leapt to the defense of his personal friend, Michelle Rhee, since she is basically in the same boat. Me, I do this entire body of research and advocacy for precisely no financial remuneration whatsoever, and I am not going to go around asking billionaires or ordinary people for any funds for it. This, however, won’t prevent folks who are rabidly anti-public-employee-union for saying that I am receiving millions of dollars from a dastardly union plot to take over the nation, and that teachers are selfish and only in it for the money. If it weren’t such a serious situation, it would be laughable.

I see that you nixed my earlier comment on your blog showing why the numbers shown in the UMBC study are consistent with Rhee’s claim on her resume. That’s not going to make the truth go away. Curious readers can check out eduwonk.com, and search in thread “Rhee Invented!” for a breakdown of the UMBC numbers.

GFB – I’ve posted this response on a number of websites and they have all mysteriously disappeared. Now, they weren’t offensive, they didn’t attack anyone personally, so it is a mystery. I have no great deep understanding of statistics, but I know enough to know that Rhee’s claims of 90% going from the 13th to the 90th percentile is very likely pure bullshit. Let me just illustrate a hypothethical given to me by my wonderful statistics professor – I paraphrase, but the gist is the same.Imagine that you have 100 two-year-old children and you want to track how tall they become as adults. It’s a pretty-accepted anecdotal “fact”that adults are approximately twice the height they were when they were 2 years old.Your group of child subjects range in height from 30″ to 36″ (2’6″ to 3’0″). When they turn 18, you would expect them to range in height from 5’0″ to 6’0″, but you wouldn’t be terribly surprised if one or two made it to 6’6″ or perhaps one or two others ended up being 4’8″, right? However, wouldn’t you be utterly ASTOUNDED if 90 of them, 90 PERCENT, grew to be 6’6″? or if 90 of them grew to be 4’8″?

Rhee’s claims – improbable? They sound like the rantings of a sociopathic delusional liar. And these are the kinds of delusions the Ed Deformers want informing public education? God help us.

There was one other important reason that test scores were not included in the UMBC study: the student received a “1” score. If you look at the chart on page 142 of the report, you will see that Harlem Park had the highest number of “1” scores in 93-94 and among the highest in 94-95. This probably explains part of the reason Harlem Park’s percentage of not-included scores is high.

I also noticed a discrepancy in the data. Test scores are included for each grade in each year, starting on page 143. Then, on page 149, scores are provided for students enrolled for two years. The 1994-95 3rd grade students would include those taught by Rhee for two years. For some reason, the number of students in the two-year dataset is higher than the single year students. One would expect the reverse. I’m not sure what to make of that.

Saw that as well.

But remember, it the students scores on page 149 and 152 that are to be examined. These are the ones she had for 2 years and worked her magic on.

Sidenote, A former collegue and defender of Rhee has tried to argue that maybe only 5 or 10 of her students took the test. While that is certainly possible, 4 or 8 out the atleast 35 she taught is not what she is implying.

She’s as slippery as an eel.

Dear Coach Brandenburg,

Have you found any evidence that Michelle Rhee, the fearless leader actually ate a bee in front of her students? Does she suggest special techniques for taping student’s mouths shut so that there won’t be marks or tears since has discussed these strategies for garnering student attention and raising tests scores? Seriously, why have her comments about these techniques been ignored.

One possible reason why the number of students taking tests at HPE was so low is that the principal maybe kept those kids off the test knowing they would drive the scores down. Remember, Rhee’s class actually did gain – I think you said 13% to 45% – just not the huge gains she claimed. Leaving your lowest performing students off the test will cause your scores to go up. After all, that is what Rhee did in DC and we had a gain of around 4%, didn’t we? She maybe learned this trick from her time at HPE.

That is my suspicion as well. Scores of “1” are what a student gets if he or she answers nothing at all, or just scribbles and/or defaces the test and makes no effort. If you get your low-scoring students to do that (or merely don’t discourage them from doing that), it raises your class average by a lot!!

There were fewer instances of this in 1995 at Harlem Park than in past years.

No, Chris, as usual, you are wrong. The percentage of students whose scores do NOT appear was higher in 1994-5 at Harlem Park than at any other school in the study, in any year. Something very, very suspicious was going on. Michelle Rhee’s claims are, in fact, disproved by the evidence, despite your delusional dreams to the contrary.

Did you even read Hess’s post? You sure as hell didn’t offer any counterarguments.

And there’s a reason no one is going to remunerate you for this “research”, and that’s because your understanding of statistics is equal to that of maybe a 1st year undergrad. The numbers are consistent with Rhee’s claim bearing out as true even if half of the tested students came from Rhee’s class. There is no possible way that she taught a class of 60+ students; there were either 2 or more classrooms total, and so just like Hess and myself wrote, you need to know exactly where those scores came from to level the asinine accusations that you are making here.

Why read Hess?

He calls it this hatchet job.

No Fred, hatchet job is Miss Rhee telling FastCompany she riffed teachers who abused students and who had had sex with them.

More dissembling.

I wonder how Fred would feel if the same was said of him

There is no possible way that she taught a class of 60+ students;Michelle Said that Chris. She said she taught 70 kids.

November 2008 Rhee at the Aspen Institute:“I had a life-altering experience through that experience [teaching in Baltimore], I came to realize this is all about the teachers, because for those 70 kids nothing changed….”

“And so I became obsessed with this idea that if we were really going to change the quality of urban education in this county, it’s going to be about high quality teachers.”

http://mefeedia.com/entry/dc-schools-chancellor-michelle-rhee/15966031Are you calling her a liar?

I spoke recently with the Dr. Lois Wilson, the now-retired educational researcher at UMBC who was the principal investigator on the ERIC-UMBC-Tesseract evaluation report. She told me (and it’s also in the report) that one positive thing that Tesseract did was to try to keep the class sizes down and to have a qualified teacher’s aide in every classroom. Class sizes at HPE and the other Tesseract schools appear to have averaged in the mid-20’s, not in the mid-30’s, or even mid-60’s or mid-70’s.

It sounds to me like she is talking about the work that she claims she did with one of her second-grade (later third-grade) colleagues (Carole Jacobs?). She makes repeated claims about having worked with that other teacher to bring BOTH classes from the very bottom to the very top.

Now, of course, all that might be made up, too. I can’t tell.

I wish that some teachers and students from Harlem Park Elementary from that era would come forward.

page 126 of the pdf has Satffing and Class Size Information

For HPES, grades 1-5:

17 classroom

1 Self contained DEC

466 enrollment

26.7 estimated class size.

Yet on the next page, they have data on the size of classes they observed:

Gr 2 28, 26, 21

Gr 3 (Michelle’s) 19, 20

gr 4 26, 19, 27

Gr 5 24

Mean size 23.3

If the grade 3 classes are Michelle Rhee’s and Michele Jacobs, I think there are small because they would have had only students they had in grade 2. That is, any new 3rd graders were not assigned to either of their rooms, nor were they reassigned students if their numbers drop.

I think based on this, one could say Michelle Rhee taught 40 kids in 1994-1995- half hers and half Miss Jacobs.

So, 90% of 40 is 36.

36 of the students listed on pg 149 and 152 would have had to score in the 90th percentile (77 NCE).

Reading

(36*77+20*y)/56=45 NCE

In this formula, y (the NCE for the 10% and the non Rhee kids) would be less than 1.

So, her 90% is less than 36.

Let’s try 32 students:

(32*77+24*y)/56=45

Here we get a 2.5 NCE for the other untaught by Miss Rhee.

Doesn’t seem likely.

Try 27

(27*77+29z)56=45

15 NCE.

A little more probably, but for this to take hold, she co-taught 30 students. That doesn’t seem as impressive.. I’m sure many regular teachers in DCPS would have love to have had Michelle give them only 15 kids, or let them co teach 30 with another.

For the Chris Smyres or Frederick Hessis, 27 out of 40 is 67.5% with other kids doing only 15 NCE.

That’s not the impression Michelle has created.

Hey Guy.

You’re retired with all this free time on your hands. :)

Drive up to Harlem Park.

Maybe they have a booklet with class photos.

Maybe the secretary or building sup was there 16 years ago.

They always have the good stuff.

Ed. HOW does a 2nd grade teacher teach 70 kids? Give an example. You really think that’s what she meant?

Chris,

maybe she had 35 and Michele Jacobs had 35.

Remember, they co-taught the same students they had the year before.

One year, I had a class of 36 and my wife had a class of 38.

You posted a few times since.

What do you think she meant?

I’m just taken her on her word at this point. Why shouldn’t I?

Also, over at the eduwonk link, Chris displayed an interesting math concept.

He described Miss Rhee’s resume from 3 1/2 years ago about her work from 15 1/2 years as being “20 years” ago.

Maybe, he can run my mortgage company.

No, dumbass, my quote you are referencing was this:

“There’s nothing that suggests Rhee lied about her classroom success 20 years ago”

You want to discuss how much more important her results are given they are ~16 years prior instead of rounded up to 20?

Dumbass,Really?

And you get mad at PhillipMarlowe for calling you Mr. Peck, “dickless” from GhostBusters.

“Great minds discuss ideas. Average minds discuss events. Small minds discuss people.”Ideas like truthfulness, fidelity, honesty and humility.

Ed: I will call you out if you misquote me. Phillip was the moron who would toss out dick jokes for no reason at all (or no reason he could ever articulate, since he hardly ever counters my arguments).

If there were even just 2 classrooms in 3rd grade, the data is consistent with Rhee’s claim. Your number game above is really bad: think more about the assumptions you’re making with enrollment numbers.

Chris: you are absolutely delusional!

You give “smearing” a bad name!

(Sorry, bad joke.)

If there were even just 2 classrooms in 3rd grade, the data is consistent with Rhee’s claim. Your number game above is really bad: think more about the assumptions you’re making with enrollment numbers.No it isn’t.

And here’s Math (pg 152)

Gr 3 53 students NCE 51

(36*77+17t)/53=51

By this formula, she couldn’t have had 36 students.

try 32 kids (out of36)

(32*77+21t)/53=51

The others got 11 NCE.

try 27 students (30)

(27*77+26u)/53=51

The others NCE is 24. Again possible.

But you see the pattern here.

As we drop the number of kids getting into the 90th percentile, in order to raise the NCE of the other students to something a little more reasonable, the size of Miss Rhee’s student group begins to look ridiculous.

As I wrote,

I’m sure many regular teachers in DCPS would have love to have had Michelle give them only 15 kids, or let them co teach 30 with another.Over at eduwonk, you were using the same formula Chris, but giving her class size as 20.

MAybe Michelle should say this:

When I was a teacher at Harlem Park ES, I experienced how one teacher could make a difference in raising student achievement (as measured by CTBS test scores). I co-taught with another teacher and I got 20 of our students to move from the 13th percentile in their 2nd grade fall administration of the CTBS to the 90th percentile in the Spring 3rd grade test session.