Rhee Invented!

Here’s the site for Michelle Rhee’s new education advocacy venture.

By the way, say what you want about Rhee, she’s not wrong about this (and if you haven’t been close to it, it’s even worse than you probably think):

Policymakers, school-district administrators, and school boards who are beholden to special interests have created a bureaucracy that is focused on the adults instead of the students. Go to any public-school-board meeting in the country and you’ll rarely hear the words “children,” “students,” or “kids” uttered. Instead, the focus remains on what jobs, contracts, and departments are getting which cuts, additions, or changes. The rationale for the decisions mostly rests on which grown-ups will be affected, instead of what will benefit or harm children.

44 Replies to “Rhee Invented!”

  1. I see fracture lines in the future for the ed reform movement. Rhee is at her best when she plays the lone warrior in the trenches. I don’t think you want her as the face of your tea party.

  2. It’s rather ridiculous that anyone really needed Rhee to clarify to effective teachers that, “You don’t have anything to worry about. My job is to make your life better, offer you more support, and pay you more.” At no point did she ever say anything to the contrary. The charge that she hates all teachers (every last one!) was at its core a particularly slimy bad-faith argument adopted by a vociferous and vocal opposition. Lou Fleming must realize the irony his Tea Party reference elicits when opponents of Rhee heavily relied on the visceral rather than the actual in criticizing DC reform.

  3. I think Lou might have been comparing Michelle Rhee to Sarah Palin with that Tea Party reference.

    They do have a similar zeal for righteousness.

  4. Chris – key phrase in your comment is “effective teachers.” When you don’t believe teachers are important, it’s hard to make the distinction between effective teachers and everyone else.

  5. Steve:

    We can brainstorm a list of other notable key figures in and outside of education that also have a “similar zeal for righteousness”. That doesn’t make the comparison any more valid, nor does it avoid the noted irony inherent in such.


    True, and how funny it is that so few commenters here consider that sentiment much more “anti-teacher” than anything Rhee has ever done.

  6. I love that her site’s domain name is ‘children first’.
    Michele Rhee is a jokefest hall of famer. I will put tape on my mouth
    and falsify unverifiable “data points” on my resume in her
    honor. Plus, I promise to say “crappy” a lot when interviewed to
    prove I take a no nonsense approach to education reform.

  7. Jokefest,
    Her time in Baltimore should be labeled the “Baltimore Miracle.”
    A miracle that many believe in despite the lack of any evidence and in the face of evidence that it didn’t occur.

  8. I definitely think that the arguments proffered by my namesake are critical. Clearly the most important thing to discuss when talking about education reform is the question of whether Rhee’s classroom test scores rose 20 years ago. And the most important way to discuss that is to use a catchy phrase found on the posts of other conspiracy-oriented commenters.

    I’m a little disappointed though that we couldn’t have thrown in “Billionaire Boys Club” or “education deformers” or other really clever phrases to advance the discussion. Then we could have all done a virtual high-five, which would have been cool.

    I know when I hang out with other lowly taxpaying school parents, these are the main things of interest to them. We really like to focus our conversations on conspiracy theories about former superintendents in other cities, and whether their actions will fracture some other amorophous group of policy wonks. That’s way more important to us than whether our kids’ teachers are any good, and whether we can get more good teachers into our kids’ school and clear out the dead weight.

  9. I left out that Ms. Rhee was making these miraculous claims just three ago.
    She posted them on her bio at her NTP.
    So, three years ago, she was typing lies.
    A year and a half ago, she told some tall tales about the rising test scores at Shaw Middle School due to the work of one of her principals she hire away from Montgomery County. Jay Mathews and John Merrow fell for the ruse, though John Merrow did make a correction later.
    11 months ago, she told FastCompany magazine she got rid of teacherS who had sex with students. It turn out to be just one had sex.
    For all her derision towards collaboration and communicating with parents, she was quite willing to meet with parents who kids did not attend Hardy Middle School and then decide to remove the principal. She never did meet with parents who children attended Hardy.

    I know when I hang out with other proles, we like to be lied to.
    I know someone who will lie to me will hire the best teacher possible.
    They will be an excellent role model for the children.

  10. ^ Too stupid to think of your own handle, eh? Need to steal someone else’s? If you need help, I can think of a few new ones you can adopt.

  11. Chris is correct – oddly that wasn’t me above. Sort of funny in a post about lying.

    But hey, stealing handles and obsessing over weird things will definitely convince me of your arguments. That’s a smart strategy.

  12. Not only is it a weird fetish, it’s also false: there’s nothing that suggests Rhee lied about her classroom success 20 years ago, as the evidence that does exist (grade level data at her school showed significant gains, and other coworkers and her boss were willing to vouch for her) suggests she made gains. Of course, this issue is a nonstarter, but it’s funny to see it continually, and incorrectly, referenced.

    The irony is also lost on these anonymous crusaders: Rhee’s inability to offer the testing data that detailed her classroom success is not indicative of any wrongdoing on Rhee’s part, but directly implicates the failures of the accountability system at the time to make the testing data more transparent. Which is kind of something that Rhee has been adamant about fixing this whole time, if you can believe that.

  13. Eduwonk already posted on stealing handles which I hadn’t realized was an issue until someone stole mine above.


    There are some really bizarre people commenting on this site. It’s amazing how unhinged people get in the education community – I guess they don’t like seeing their monopoly challenged by new ideas. That said, the bizarros do help bring clarity to the debate.

  14. gee whizz,
    slippery fingers get a couple people into a tizzy, and provide the opportunity to avoid the facts.

    edconsumer (the first) is willing to accept someone who will lie to them.
    I know a few car mechanics, doctors, and teachers who like you.

    As for Chris who makes a pretense of having his fingers on data, is now turning 15 into 20.
    Chris, I’ve got a five and a ten. willing to trade for a twenty.

    Ms. Rhee was making false claims about herself 3 years ago, and she continues to fudge the facts on her time.
    Chris, data does exist on Miss Rhee’s time at Harlem Park ES. The GAO did an examination of the private company operating the school and HPES gave the standardized data to the GAO. No Miracle scores there.
    If you wish, I’ll send you a link, but you can google it.
    It seems you are reverting to that comparison made to you earlier this year, in that you think like a creationist.
    You were highly insulted.

    PS. When I write that college tuition check, I am a edconsumer

  15. Tedconsumer:

    Oh come off it. Not only is it immature to try and replicate someone else’s handle with a “clever” twist, but the post in question was absolutely intended to be read as if “Edconsumer” him/herself had written it. If you were the anonymous party guilty of this, an apology is in order; that kind of crap will get you rightfully banned from most internet forums.

    You are also sniveling about how we accept someone who will “lie to us”, but you are the bigger liar here. You are referring to this GAO study, yes?


    Suggesting that the GAO proved here that Rhee is a liar is as willfully ignorant and equally deceptive a claim as when you implied that a study on the Air Force Academy had implications on VAM in DCPS:


    There’s a reason we only see anonymous loonies referencing this paper as proof that “RHEE LIED!” I will indulge you as to why:

    * The study’s experimental group included 7 of the 9 privately managed public schools in Baltimore, one of those being Harlem Park Elementary. While this analysis assessed test scores of students who remained in the schools, the longitudinal data did not provide any specifics on schools, and certainly not on any of the teachers. Nothing can be said of Rhee with these data, as her unknown individual impact may have amounted to no more than ~5%.

    * Cross-sectional analysis compared test scores each year “without regard to changes in student population from year to year”. This analysis gave school-level data, which again doesn’t give specifics on teachers and their individual impacts. Also, since this analysis takes into account the high turnover of students at these school, it may also provide an incorrect picture of how a teacher might have done with the same class over time, as we don’t know what type of turnover Rhee saw in her classroom relative to classrooms outside of it. Nothing can be said of Rhee with these data, either, and her unknown individual impact on these data may have amounted to no more than ~10%.

    * Also, cross-sectional analysis relied on measuring differences between schools, but did not provide data on how well students did in relation to all schools nationally. This can make the comparison misleading, particularly if the control school Harlem Park was compared to was not representative of the national trend.

    * Cross-sectional analysis of the data, despite all of the flaws described above in relating it to Rhee, suggested that in the period of ’92-’95 Harlem Park Elementary underwent incredible gains relative to its control school, jumping from an effect size of -0.54 in 1992 to an insignificant difference in 1995. None of the other schools in the experimental group achieved any such growth. This was measured from reading scores; math scores showed similar impressive gains over Harlem Park’s control school which no other school matched.

    EVEN IF this data provides no outlook on how Rhee did individually with her students over time (which it DOES NOT), it DOES suggest that Harlem Park showed marked gains in the period Rhee taught there.

    And this is ALL BESIDE THE POINT that her test scores 20 years ago have NO BEARING WHATSOEVER on her effectiveness as DCPS chancellor.

    I eagerly await your rebuttal.


    Speaking of edlharris, he had this to say about me on another blog:

    “There’s is [sic] a TFA alum who comments over at Eduwonk where he has taken an abusive attitude with me for calling him out for believeing [sic] in Michelle’s Baltimore Miracle without evidence.”

    “Abusive”! “Calling [me] out”! HA!

    I am developing quite the fanbase in my time here on Eduwonk. Anyone else come across instances of pathetic whining about me elsewhere along the Internets? I know Phillip has done it at least 3 separate times so far. I do get a kick out of the publicity.

  16. Anyone, go to the GAO report and find the evidence that show where 90% of the 3rd graders scored at the 90th percentile. Michelle Rhee claimed she did that while at Harlem Park ES, where she left 15 years ago.
    (Not 20 years, as Chris repeated misstates.)
    Chris can’t find it, so he dances, using blue smoke and mirrors.

    Chris The Creationist
    Chris the Discovery Institute graduate

  17. Tedconsumer:

    You are frantically backpedaling. You asserted that there was a GAO report that specifically showed that Rhee was lying. Give a link to this report or refute the arguments I’ve addressed to you above.

  18. To Chris The Creationist (who taught high school science):
    Ms. Rhee was making false claims about herself 3 years ago, and she continues to fudge the facts on her time.
    Chris, data does exist on Miss Rhee’s time at Harlem Park ES. The GAO did an examination of the private company operating the school and HPES gave the standardized data to the GAO. No Miracle scores there.

    And, where is that “90 % of the students scoring at the 90th percentile”, as Miss Rhee put it.

  19. Tedconsumer:

    1) Can you be any more obtuse? You are asserting that the GAO did not find miracle scores for Rhee, but the GAO did *NOT* report or even analyze individual teacher effects! They didn’t even parse data by grades; it was all school-level data! And from that school-level data, what did they see? They found LARGE gains in Harlem Park relative to its control school, gains that were not seen in any of the other 6 schools in the experimental group.

    2) I’ll repost a key paragraph I wrote from before:

    “The irony is also lost on these anonymous crusaders: Rhee’s inability to offer the testing data that detailed her classroom success is not indicative of any wrongdoing on Rhee’s part, but directly implicates the failures of the accountability system at the time to make the testing data more transparent.”

    That the specific testing data is not readily available or not reported anywhere does NOT stand as evidence of Rhee lying. The only data that HAS been given suggests there WERE gains made. I don’t know why you think this “Guilty until proven innocent” bit is at all persuasive.

  20. So, no apologies for pretending to be someone else, and no retraction of your claim regarding the GAO report?

    Go on, I believe you were saying something funny before about honesty.

  21. Oh, poor Chris the Creationist can’t wait.
    I have nothing to apologize for, unlike you who can’t read the GAO report and see the breakdown which includes Miss Rhee’s 3rd graders in 1995.
    As you may recall, she taught second grade her first and second year (1992-93; 1993-94). Her third year (1994-95) she travelled with her second graders to 3rd grade. So the 1995 test scores given to the GAO by Harlem Park reflect her class and the other third grade class.

    I’ll be sitting at the airport tomorrow morning (just a few miles from Harlem Park.) Part of my time will be spend responding to you.
    Which will be difficult because the creationist in you will reject facts.

  22. Not sure why you are stalling (well, I sort of know why). What report are you referencing? There is no such grade-level analysis in the GAO report I cited above.

  23. The Fantasies of Michelle Rhee

    The GAO report shows no 90% at 90 percentile.
    The follow-up response in the report by EAI makes no mention of it.
    Gee, you’d think EAI would like to trumpet it to the heavens what a miracle Michelle wrought.

    The MSPAP scores show no evidence of students with 90 percentile. (Yes, I know we are dealing with different tests, but if you went back to the data in Maryland and compared MSPAP scores to CTBS scores, you will see that the higher percentiles are matched by a higher MSPAP scores. Not causation, but more than a coincidence.)

    The GAO report states:

    Although small, the cumulated effects for CTBS
    scores in both reading and math were statistically significant in favor of
    the comparison group (ie, not EAI schools).
    andThree of the seven privately managed schools (Dr. Rayner Browne, Harlem
    Park Elementary
    , and Sarah M. Roach), however, show a more definite
    pattern of underperformance compared with their matched schools. In
    these cases, effect sizes were great enough to warrant further attention.

    The GAO report does not provide breakdown of CTBS scores per grade level at each school, but it does mention the UMBC study on EAI. This study , commissioned by the mayor and city council, does provide a breakdown per grade level. And when you read the data, your heart will break Chris. You will see in black and white that your heroine, your idol, your heartthrob, your golden girl, your Braveheart, your Superwoman didn’t tell the truth.

    What happened with the CTBS scores?
    These are NCE scores for the 3rd graders in 1995 who had been at the school for 2 years-grade 2 and grade 3. Most would have been taught by Miss Rhee and Miss Jacobs:
    55 test takers NCE 45

    53 test takers NCE 51

    (Note to all-In this study, the NCE is a few digits above the percentile)

    The UMBC study has a line in it that you and Mr. Rotherham will love:
    pg 85

    A veteran teacher said “We are told everything done in the past has been wrong, and made to feel that we have been part of a conspiracy to keep children ignorant.”

    So readers to review:
    What Miss Rhee said on her resume of 3 1/2 years ago (not 20 years ago as Chris loves to dissemble) is false.
    Her “success” in the classroom did not result in acclaim from the Hartford Courant, the Wall Street Journal, the Home Show and Good Morning America
    The Good Morning America show of Oct 24, 1993 was about EAI in Baltimore and taped at Harlem Park. Notice, it aired at the START of Miss Rhee’s second year, ie. BEFORE her miracle.
    She did not take her 2nd graders from the 13th percentile to 90% at the 90th percentile near the end of third grade.

  24. Tedconsumer:

    Well that only took over a month. Were you busy researching? The extent of your response doesn’t suggest it:

    1) Reread my last few comments: the GAO study did not analyze data at teacher-level, so there is no possible way it could have reported on the scores Rhee is referencing.

    2) EAI did not “trumpet [Rhee’s successes] to the heavens” in their reply because the report wasn’t about individual teachers and their successes! They replied with corrections they felt needed to be noted, not with new and unrelated material.

    3) EAI in its response DID, however, rightfully disagree with the conclusion that students at all of their schools did not perform better academically.

    4) Rhee’s claims concerned her students’ CTBS scores, not their performance on the MSPAP. They are very different tests, as you should already know.

    CTBS: assesses basic skills in reading, language, spelling, math, study skills, science, and social studies

    MSPAP: assesses critical thinking, high-level problem solving, and integration of knowledge in several subjects.

    5) Even assuming you can posit a correlation between achievement on the two tests, the MSPAP data was NOT reported for 1994-1995 in the GAO study, which was the year that CTBS scores dramatically increased for Harlem Park Elementary. Thus, WE DON’T KNOW if there were similar gains made on this test as were found with the CTBS data, and so you again have no leg to stand on with that claim.

    6) Give a link to the UMBC study you are referencing. Given your penchant for throwing out claims supported by an incorrect reading of the literature, I’d like to double-check your “facts” here, as well.

  25. #6 You’re a smart boy, Christopher.
    I’ll give you a week to find it.
    Google it.

    #4 I said they’re different. What’s wrong, can’t read?

    #5 The GAO report has a misprint where they identify 1993-1994 MSPAP data as 1994-1995 data.
    You can find the 1995 MSPAP results in this file.

    I noticed you backhandedly acknowledge the media reports don’t match what she claimed they said. The Good Morning America one is the best.

  26. Ted:

    re: to 6) YOU’RE the one citing a report that you’ve yet to share with readers here. It’s not my job to do your legwork. Google/ERIC queries return only a small summary of these studies, not the data sets you are referencing.

    You already flunked Research Basics 101 because I have already had to show why each of your claims thus far has erroneously cited such studies for support. Recall that you started this debate referencing the GAO report and how it brought ruin upon Rhee’s credibility. The only thing acknowledging it ruined is the trust readers should place in your ability to engage in good-faith and well-cited arguments.

    Give the source for your data.

    re: to 4) You ought to explicitly acknowledge that Rhee is referencing a different test than you are making claims about. This all ties in well to #5.

    re: to 5) The MSPAP scores, as already shown, are moot to the claims that Rhee made about her students. It’s also not true that every year where gains are seen on the CTBS they need to be seen on the MSPAP, so having the MSPAP scores readily available doesn’t bear greatly on Rhee’s claims, and it certainly does not offset the CTBS data already compiled that show significant gains at her school in 1995.

    Nevertheless, if we wanted to look at these MSPAP scores for 1995, we can’t look to the GAO report:

    “In addition, for students in third and fifth grades, we used reading and math outcome scale scores from the Maryland State Performance Assessment Program (MSPAP) for school years 1992-93 and 1993-94.”

    I’m now going to go out on a limb here and assert that you yourself haven’t bothered to check the MSPAP data for 1995 on the link you provided. The reason?

    “The files are large because they reflect information for 1600+ schools in Maryland. To ease the use of these large files for “serious number crunchers,” the data are stored in the SDF (standard Data file) format and contain no decimals. The user must place decimals in the numbers. The SDF format is fixed length with no separators or delimiters, relying on the accompanying design document.”

    You want to cite a raw data set the GAO paper didn’t analyze, yet you can’t even correctly read the data after the experts have already poured over it and compiled it into a paper. I thus find any further analyses that you will proffer on this matter rightfully questionable.

  27. Ted:

    I even perused the MSPAP raw data set you linked to above, and what did I find?

    1) Only criterion-referenced levels of performance are given, NOT the outcome scale scores given in the GAO paper. This shows without a doubt that you didn’t even bother to look at the 1995 data you linked to, since you can’t simply compare the two; we can’t look at your cited 1995 performance levels to simply check how they compare to the ’93 and ’94 reported outcome scale scores. This is more evidence of your dishonesty.

    2) Again, MSPAP scores are moot for judging Rhee’s claims since the MSPAP isn’t the test she was referencing when claiming her students scored impressively on standardized tests. Whether or not there were gains made in 1995 on the MSPAP has little bearing on the direction scores took on the CTBS, nor on the exact sizes of those gains/losses. In addition, there is already evidence of stellar improvement seen in Harlem Park on the CTBS in 1995 as shown in the GAO report.

    Yet even given all of that, what happens if we were to look through the data sets you blindly referenced? We’d see that her MSPAP gains in 1995 were impressive compared to past years in 3rd grade, and that the gains made by her 1995 3rd graders over past years were generally missing for the following year’s 3rd graders:


    MSPAP 3rd Grade READING

    1992) %EXCELLENT: 000.0
    1992) %SATISFACTORY: 005.3

    1994) %EXCELLENT: 000.0
    1994) %SATISFACTORY: 002.5

    *1995) %EXCELLENT: 000.0
    *1995) %SATISFACTORY: 012.1

    1996) %EXCELLENT: 000.0
    1996) %SATISFACTORY: 009.1


    1992) %EXCELLENT: 000.0
    1992) %SATISFACTORY: 000.0

    1993) %EXCELLENT: 000.0
    1993) %SATISFACTORY: 000.0

    1994) %EXCELLENT: 000.0
    1994) %SATISFACTORY: 003.8

    *1995) %EXCELLENT: 001.1
    *1995) %SATISFACTORY: 014.3

    1996) %EXCELLENT: 000.0
    1996) %SATISFACTORY: 001.3


    1993) %EXCELLENT: 000.0
    1993) %SATISFACTORY: 004.8

    1994) %EXCELLENT: 000.0
    1994) %SATISFACTORY: 002.5

    *1995) %EXCELLENT: 003.3
    *1995) %SATISFACTORY: 008.8

    1996) %EXCELLENT: 002.6
    1996) %SATISFACTORY: 010.4


    1993) %EXCELLENT: 000.0
    1993) %SATISFACTORY: 001.0

    1994) %EXCELLENT: 000.0
    1994) %SATISFACTORY: 006.3

    *1995) %EXCELLENT: 001.1
    *1995) %SATISFACTORY: 011.0

    1996) %EXCELLENT: 000.0
    1996) %SATISFACTORY: 013.0


    1992) %EXCELLENT: 000.0
    1992) %SATISFACTORY: 004.0

    1993) %EXCELLENT: 000.0
    1993) %SATISFACTORY: 000.0

    1994) %EXCELLENT: 000.0
    1994) %SATISFACTORY: 001.3

    *1995) %EXCELLENT: 000.0
    *1995) %SATISFACTORY: 009.9

    1996) %EXCELLENT: 000.0
    1996) %SATISFACTORY: 001.3


    1992) %EXCELLENT: 000.0
    1992) %SATISFACTORY: 002.7

    1993) %EXCELLENT: 000.0
    1993) %SATISFACTORY: 001.0

    1994) %EXCELLENT: 000.0
    1994) %SATISFACTORY: 002.5

    *1995) %EXCELLENT: 002.2
    *1995) %SATISFACTORY: 015.4

    1996) %EXCELLENT: 000.0
    1996) %SATISFACTORY: 003.9

  28. Did you find the UMBC report?
    While you search, you’ll want to compare Harlem Park to Pimlico ES. They were the comparison school for the UMBC study. You also need to pull out what was going on in the city and in the state.

    For the first 5 years of MSPAP, we were changing what we were doing each year as we saw more and more of the test and as the state modified. At the time, we noticed that the school systems that were doing the best on MSPAP were the system’s whose employees were writing the test.

    To find the report, think of who was running those schools.

  29. 1) So no reply to the facts that the MSPAP scores you linked to were not what you implied them to be, and that they showed big gains in 3rd grade in 1995 when Rhee said her kids scored well on the CTBS, and that your previous claims regarding the GAO report were bogus? This type of dishonesty is the reason no one tends to argue with you and any phonies like you when you all post these inane screeds.

    2) YOU need to link to the report, Ted. Not me. You ham it up about awful it is that Rhee didn’t provide her individual class scores despite their being poorly documented and reported, yet here you claim to HAVE the evidence but DON’T WANT to share it! You’re disingenuous and a hypocrite.

    3) The GAO report shows large gains at Harlem Park relative to its control school Pimlico. I already notified you to this. The MSPAP scores I tabulated above do not require a comparison school to allow you to see the gains that were made in 1995 which were nonexistent before and diminished the year after.

  30. It does, Mr. Peck.
    Quite applicable to the shit those of us who want to engage in an honest discussion have to put up with on this site.

  31. If you want an honest discussion I would advise you to start offering counterarguments instead of either nothing at all or what you think are witty comebacks. I can provide links if you cannot find (the many) instances of what I’m referring to.

    Ted has his work cut out for him here, so why don’t you carry on with the “honesty” you embody in the other threads where I have already responded to you?

  32. “what you think are witty comebacks.”

    “The witty repartee’ of Joey ‘Eat Me’ Donner.”

  33. Hess beat me to it by a matter of hours, but do go ahead and read his take as well:


    Here is a link to the report that Ted cited but obnoxiously refused to link to himself:


    And here are the reasons why Ted is wrong again, and why he again fails at reading research literature correctly:

    1) See pg. 113 regarding the conversion between NCE scores and percentile scores. You must consider this conversion if you want to relate the NCE scores reported to the percentile gains that Rhee claimed, as jumps in NCE scores may translate into larger jumps in percentile scores, as the case is here. Here’s a PDF conversion table:


    Rhee’s 90th percentile claim translates into an NCE of 77. Keep that in mind.

    2) See pg. 109 regarding the CTBS scores reported. The scores tabulated in this report do not include ~20-25% of the students actually enrolled at each school, and for 1995 at Harlem Park only 64% of enrollment is accounted for. More specifically, scores are not reported for students testing with a “1” score. This needs to also be considered if you are comparing the growth in NCE scores reported, as it will likely not be the exact measure that Rhee would have seen on her own classroom score reports in 1995.

    For more on this, see pg. 142 which lists the # of students who were removed from the score reports for “1” scores (in parenthesis are the # of students enrolled at the end of that year)

    Harlem Park 1992-1993: 31 (523)
    Harlem Park 1993-1994: 39 (493)
    Harlem Park 1994-1995: 17 (440)

    While there is no grade level breakdown, this suggests that fewer 1995 test scores were dropped for “1” scores, even after normalizing to enrollment, than in 1994 and 1993. Again, there is likely going to be a difference between the NCE scores reported here and the data that each teacher received for their respective classrooms, and the reported scores in years prior to 1995 may also be biased upward.

    3) 3rd grade data provided in the MSPAP scores I listed above and in this paper comes from at least 2 classrooms, and thus is NOT a direct assessment of how Rhee’s students performed in 1995. This must be considered, and I will explicitly show why momentarily.

    4) Before we get to the actual CTBS data, here again is Rhee’s claim:

    “Over a two-year period, moved students scoring on average at the 13th percentile […] to 90% of students scoring at the 90th percentile or higher”

    Let’s consider the claim as if the 90% of her students she references makes up the entirety of her share of the reported scores in this paper, which of course is an assumption that we can’t confirm. As I said before, scoring at the 90th percentile would mean scoring at an NCE of 77.

    Scores from 43 students were reported in 1995 for reading, and 44 students for math. How many of these scores were from students in Rhee’s class versus the other classroom? We again do not know, and this fact is very important for determining what the data suggests. Given that there were only 64% of total enrollment at Harlem Park that year with reported scores, this is a major problem for any analysis one wants to do with these numbers.

    For now, assume it was split down the middle and 22 of those 44 scores were from students in Rhee’s classroom.

    The reported NCE scores for 3rd grade in 1995 were 48 in reading and 52 in math, so an average NCE of 50.

    If half of the scores were from Rhee’s students, then the other half is from students in the other teacher’s class. What might the other teacher’s scores have been? Again: don’t know! If we want to get a very rough estimate of what this teacher’s scores might look like had they been similar to past years, we can look to previous testing years in 3rd grade and average the scores:

    3rd Grade Reading NCE (’92, ’93, ’94): 39, 35, 34
    3rd Grade Math NCE (’92, ’93, ’94): 37, 34, 42
    Average NCE for these years: 38 + 34.5 + 38 = 110.5 / 3 = 37

    Of course, we could also look to past data for that cohort of students, but we can’t look at 2nd grade 1994 scores since Rhee taught some of them as well, and there’s obviously differences in the samples of students’ scores reported at each year, so we might as well stick with the average scores that the school tends to see in 3rd grade as shown above.

    There are lots of assumptions being made so far, ones that Ted either didn’t know he had to make or that he dishonestly pretends that are magically confirmed when they have not been. These assumptions all may be close to the truth, so we can consider several possibilities to see how the numbers may work out.

    If the other teachers, who impacted half of the combined 3rd grade NCE score in 1995, had gotten their students to 37 NCE, and Rhee had indeed brought her students to 77 NCE as she claims, what combined score would we expect?

    (37+77) / 2 = 57 NCE

    This is reasonably close to the 50 NCE that was reported in the UMBC study. Another example: if of the 44 scores that were tabulated, what if 20 of them were Rhee’s and 24 were the other teacher’s?

    [37*(24/44)] + [77*(20/44)] = 55 NCE

    Even closer. What if the other teachers had brought their students to 34 NCE, instead?

    [34*(24/44)] + [77*(20/44)] = 53.5 NCE

    You get the point.

    Given the uncertainty in the assumptions made above, these numbers are entirely consistent with Rhee’s claim bearing out as truthful. This data obviously does not conclusively prove that her students hit the 90th percentile, but ALL OF THE EVIDENCE cited so far from these given studies and from personal accounts from Rhee’s coworkers and principal suggests either nothing, or that she was telling the truth. Claims to the contrary, as I have painstaking shown here and elsewhere, have ZERO evidence to support them. Yet that doesn’t stop the misguided character assassination from folks who have no idea what they’re talking about.

    As an important aside to the general thrust of this comment: the gains reported in 3rd grade in 1995 over previous years (that even tinfoil-hat-Brandenburg acknowledges in his clunky Excel graphs) were tremendous. That’s a gain in reading from the 14th to the 47th percentile in just 1 year. Fantastic. Now that we’ve established the approximate gains made of an anonymous group of kids over 15 years ago (and that these scores may actually not correctly gauge the actual gains made of this cohort per the extensive reasoning above), can Rhee’s detractors *please* find something more worthwhile to be petty about?

  34. Chris’s argument is that she can’t be lying because we can’t show she isn’t, nor can she show she isn’t, so we should believe her.

    A lack of data does not prove anything.

    And, Chris, seeing as you are an experienced teacher (I know. Funny!), doesn’t it sound a bit fantastic to raise kids from 13th to 90th in a year or two?

  35. No. My tl;dr argument is that the lot of you need to read more carefully.

    A lack of data does suggest this, however:

    “The irony is also lost on these anonymous crusaders: Rhee’s inability to offer the testing data that detailed her classroom success is not indicative of any wrongdoing on Rhee’s part, but directly implicates the failures of the accountability system at the time to make the testing data more transparent. Which is kind of something that Rhee has been adamant about fixing this whole time, if you can believe that.”

  36. 1) See pg. 113 regarding the conversion between NCE scores and percentile scores. You must consider this conversion if you want to relate the NCE scores reported to the percentile gains that Rhee claimed, as jumps in NCE scores may translate into larger jumps in percentile scores, as the case is here. Here’s a PDF conversion table:


    Rhee’s 90th percentile claim translates into an NCE of 77. Keep that in mind.

    When you compare the Texas conversion chart to the numbers given on page 113, they do not match.

    Pg 113
    NCE 45 Percentile 42
    NCE 45.8 Percentile 42

    pg 113
    NCE 38 Percentile 33
    Texas chart
    NCE 40.7 Percentile 33

    Reformulation perhaps?

  37. @Chris:

    The problem remains that, in the interviews with Rhee and her colleague, there was never any sense that Rhee just blew her colleague out of the water. They spoke about how “they” raised those scores in their two team-taught clases. In order to substantiate your numbers, you’d really have to throw her colleague under the bus — literally scores that are half of that of Rhee’s.

  38. Ted:

    Percentile to NCE conversion is not something that ought to undergo a “reformulation”. The Texas chart I gave was not just used in Texas; look up the charts anywhere else and you’d see the same distribution and conversion factors. Your response here proves my point further, in that not only is there a laundry list of assumptions we’d have to make in order to tie this data directly to Rhee’s students, but the numbers themselves may not accurately correspond with a normal distribution.

    Also, do I have to assume that everything you have ignored so far you agree with, since you still haven’t responded to it?


    In the interviews, did they explicitly say that *ALL* of the teachers in 3rd grade brought their students to tremendous gains? Could there have been other classroom(s) that just trudged along at or near the average seen in 3rd grade for that school? And then, again, how do we know what scores from students in each of these classes were pulled for this study’s analysis?

Leave a Reply

Your email address will not be published.