Were large numbers of students not tested in Baltimore?

Here is Mafara Hobson’s response to my criticism of Rhee’s claims of achieving a miracle in 1994-1995:


Our public schools are in crisis. Instead of talking about how to fix them, we’re getting unfounded attacks on Michelle. To get back to the debate about public schools, we want to address this misinformation head-on.

A blogger has posted some error-laden numbers, based on a 1995 study, claiming that Michelle was not an effective teacher. A couple of mainstream journalists have picked up and re-broadcast this storyline without reviewing the underlying analysis.

Here are the underlying facts about the 1995 study:

  • This was not a study of Michelle’s students. It was a study of the school’s entire grade level, which had four teachers.
  • There is no way to know if any of Michelle’s students were even included in this study. The study included only certain students at the school, and excluded large numbers from their sample.

Some have expressed surprise that credible journalists would swallow a blogger’s analysis without looking at the facts for themselves. We were quite frankly surprised ourselves. To our members, this episode is further proof of what we’re up against and why we need your support to get the message out.


A few points:

I’m not the one making up stories about “Michelle Rhee’s miracles”. Unlike Hobson, I try to point to actual data so you, the reader, can check it for yourself.

The study states that basically all of the elementary students were administered the CTBS, but that the scores of  somewhere between 20 to 25% of the students were excluded because they enrolled after February 1, were absent on testing days, or were in certain levels of special education (see pages 25 and 109).  The average exclusion rate at the Tesseract schools in 1994-1995, the study claims, was 20%, up from 16% in 1991-2 (see pages 109 and 33), although the percentage of special education students had declined from about 8.6% to about 3%.

According to the tables later on (page 143), there were reading scores reported at Harlem Park in the 3rd grade in SY 1994-5 for 43 students. If this only represents 75 to 80% of the grade-level cohort for that year, then there were between 54 and 57 students in the third grade. That is not nearly enough students for four classes (14 students per class?!?!).

Also, on page 127, they report the number of students that they observed in the various classes at the various schools at various years. For grade 3 at Harlem Park in 1994-5, they report these numbers: 19 and 21, which I interpret to mean that in one classroom they saw 19 kids that day, and in the other class, they saw 21 kids. I presume some of the children were absent on that day. Which of the two classes was Rhee’s? I have no idea, nor does it matter.

Let us now look only at scores for “Two-Year” students who remained at Harlem Park from second grade in 1993-4 through the third grade in 1994-5, which you can find on pages 152 for math and 149 for reading. These are precisely the students on which Rhee has repeatedly claimed that she brought to the 90th percentile, because she had them for two years. In math, the 53 students (out of a total of 66 to 71, total) achieved an NCE [Normal Curve Equivalent] score average of 51, which is equivalent to either the 51st or 52nd percentile. Hmm. In reading, the 56 students for whom scores are reported (out of somewhere between 70 to 75 students, if we assume the same exclusion rates) reached an NCE score of 45, which is equivalent to the 40th percentile. Hmm again.

Not a miracle in sight.

It is deceptive for Hobson to claim that there is even a remote possibility that not a single one of Rhee’s students was tested. What — every single one of them was absent, or came in after February 1 of 1995, or was in a severe special education status? That defies belief.

Michelle Rhee’s own repeated interview statements speak of 70 students that were team-taught by her and her team-mate; she speaks of team-teaching with this other teacher and taking the same cohort of students from the second grade in 1993-4 through the 3rd grade in 1994-5. Well, it is certainly possible that there really were 70 students in the cohort in their two combined classes in the second grade, given the table on page 143, when the scores for 79 students were reported. With the exclusions added back in, that would be somewhere between 99 and 105 students, which is certainly enough for four classes. Perhaps that is what Hobson is referring to, in a statistical bait-and-switch?

Rhee never claimed, anywhere, that she did a super-fantastic job  while her team teacher’s students all scored in the cellar.  That’s the only way that Michelle Rhee or her spokesperson, Mafara Hobson, can have it that Rhee performed a miracle, given the aggregate scores for the whole cohort.

Are you ready to throw your fellow team-teacher-and-miracle-worker under the bus, Ms. Rhee?

Or is it more likely that you and she were responsible for the entire 3rd grade class, and that by winnowing out the lowest-scoring students, you were able to bring the test scores of the remaining ones up to about the 50th percentile in reading and math on a highly suspect standardized test of so-called basic skills?


Oh, and who exactly is going around attacking public schools?

A lot of people think that it’s YOU, Michelle Rhee, who are leading the charge.

The URI to TrackBack this entry is: https://gfbrandenburg.wordpress.com/2011/02/09/were-large-numbers-of-students-not-tested-in-baltimore/trackback/

RSS feed for comments on this post.

5 CommentsLeave a comment

  1. What did Mary McCarthy say about Lillian Hellman?

    Also, unmentioned by Miss Rhee and her minions, she had an graduate or undergraduate student working with her in her class.

    Great job, Guy!


  2. great job


  3. I was the principal investigator for the study cited by Guy Brandenburg and referenced in the Washington Post, and, yes, he has correctly presented the information therein (converting NCE scores to percentiles) and come to the appropriate conclusion that one teacher’s “90+ percentile scores” of the estimated four third-grade classrooms would have led to a higher grade average. We were evaluating the Tesseract program as a whole rather than school results by classroom, so CTBS scores available to us were not disaggregated by teacher.


  4. It is very important for educators to be good role models. If Rhee still believes that she “Over a two-year period, moved students scoring on average at the 13th percentile on national standardized tests to 90 percent of students scoring at the 90th percentile or higher.” then she should say that. Why? Because that is a tremendous accomplishment! That would show students it is important to be proud of their success.

    If she doesn’t believe it, then she should admit she was wrong. Showing that kind of maturity would really help students by presenting them with a good role model.

    Until she does one of these two things, people should stop giving her donations and supporting her organization.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: