From the archives: Rhee's Test Score Claims in Doubt

[jsteele]

For most of 2007, I published a blog focused on local District of Columbia issues. The blog is now offline, but I've decided to republish some of the school related posts here. Those interested in background to many of today's school issues may find them interesting.


Originally published June 28, 2007.

An earlier article described how acting DC schools Chancellor Michelle Rhee's much-heralded teaching experience in Baltimore was part of an aborted experiment in for-profit management of public schools. That article pointed out that while Rhee claimed fantastic test score gains by her students, test scores for the school as a whole had been disappointing. This suggested that Rhee's students and done inordinately well. Today, Gary Emerling in The Washington Times, reports that the test score improvements cannot be substantiated and he casts doubt on whether such gains were possible.

Rhee's resume says the following:

Taught in Harlem Park Community School, one of the lowest-performing elementary schools in Baltimore City, effecting significant measurable gains in student achievement.  Over a two-year period, moved students scoring on average at the 13th percentile on national standardized tests to 90% of students scoring at the 90th percentile or higher.

According to Emerling, Rhee said the test in question was the Comprehensive Test of Basic Skills (CTBS). While it appears impossible to recover test scores from that period, quite a bit of data is contained in a 1995 report compiled by the University of Maryland, Baltimore City. The report includes CTBS scores of Harlem Park 2nd and 3rd graders for the years Rhee was a teacher. That data suggests that Rhee could not have made the gains her resume claims.

According to the Washington Post, Rhee taught 2nd and 3rd graders during her first year of teaching. It says that for the next two years she "and another teacher co-taught a group of 70 students, of which only 13 percent were reading on grade level when they entered the class." The focus on grade level reading is inconsistent with focus on test score percentiles on her resume (and in the Times article), though in both cases the alleged improvement is from 13 percent/percentile to 90 percent/percentile. According to the Times, Rhee couldn't remember her class size which is another inconsistency between the Post and Times stories.

At any rate, according to the UMBC report, during the 1993-94 school year, 79 second graders took the CTBS scoring a NCE (similar to a percentile) of 27 on the reading portion of the exam. The Times reports that Rhee taught second-graders that year, so the 79 students taking the exam is reasonably consistent with the 70 students the Post says Rhee taught along with a co-teacher. However, while the NCE score and percentile don't always match perfectly, there is a big discrepancy between the 27 NCE and 13th percentile claimed by Rhee.

During the 1994-95 school year, the Times says Rhee taught the same students as third-graders. The UMBC report records 56 third-graders who had also attended Harlem Park the previous year scored a 45 NCE on the reading portion of the CTBS. Again, there is a big discrepancy between the 45 NCE and 90 percentile claimed on Rhee's resume.

Based on the UMBC data, which was obtained from McGraw-Hill data tapes, Rhee's students' reading scores would have improved from a 27 NCE to a 45 NCE over the two-year period. This is not a bad achievement, but far from the 13-90 percentile increase her resume suggests.

Moreover, the UMBC report also offers an explanation for at least part of that increase. According to the report, schools managed by Education Alternative, Inc., such as Harlem Park, had a "preoccupation with test scores" and "concomitant use of instructional time for testing and test preparation". In addition, the EAI schools had a fall testing program in which alternative forms of the CTBS exams were given which created a "small 'practice effect' that increases scores."

Finally, it is noticeable that the number of test takers decreased from 79 second-graders to 56 third-graders. That could also have a significant impact on the scores.

An administration that produces plagiarized education plans and Potemkin review panels would really have to stretch to reach new levels of perfidy. But, if the acting schools Chancellor has a resume containing inaccurate information, Fenty and his cohorts may have done just that.

Update: The Washington Times is running an article today that says Council Chairman Vincent C. Gray is investigating Rhee's test score achievement claims. The article quotes several Councilmembers expressing dismay at the possibility that the claim may not be true. However, one Councilmember, Tommy Wells (Ward 6) seemed to indicate indifference at prospect, saying "We're not asking her to be a teacher".

[Anonymous]

Thanks for pointing this out again. I thought there were questions surrounding test results in DC from last year, but I can't find anything online about it.