School data quality: Still a big problem
Honest principal in Nevada points to gross error in data for his school
US News and World Report's (USN&WR) annual ranking of high schools came out only a few hours ago.
Already, controversy is brewing about this dubious ranking program.
This time, part of the problem is that some of the data stored at the US Department of Education is clearly just plain wrong.
Principal Jeff Horn at the Green Valley High School in Nevada says there is a serious error in the enrollment data USN&WR used to show his school ranking 13th best in the nation.
That federal database shows Green Valley High has 477 students and 111 teachers, for a pupil teacher ratio of just 4 to 1. Apparently, that incredibly low ratio didn’t set off any alarms at USN&WR. It should have.
The actual enrollment at the school is 2,850 students, which (IF the teacher count is accurate) works out to a 26 to 1 ratio. Also, the incredibly low enrollment figure apparently messed up the USN&WR calculation of ‘pass rates’ on Advanced Placement tests. Per the magazine, the school’s ‘pass rate’ was 100 percent (another questionable number). The real pass rate is 64 percent.
The Kentucky story
USN&WR also ranks what are supposed to be the top 10 high schools in each state, including Kentucky.
I have problems with USN&WR’s #6 choice of Ballard High School and the #8 choice, the Eastern High School in Louisville.
In the 2010 Kentucky Core Content Test in mathematics, both of those schools had big black versus white proficiency rate gaps. Eastern’s was more than 30 points. Ballard’s was MUCH WORSE – an enormous 50 point difference! Check for yourself in the Kentucky Department of Education’s Gap to Goal Excel Spreadsheet for “Statewide,” “All Districts,” available here.
Blacks are getting left behind in these schools, but USN&WR's superficial analysis apparently won't show you that.
Unfortunately, the Kentucky Department of Education does not report disaggregated 11th grade ACT college entrance test results by race, but Eastern doesn’t make the top 10 for its overall average ACT composite score in 2010, and Ballard didn’t do as well as Eastern.
Stay tuned for more about the issue of racial performance gaps in Louisville’s schools. We are getting ready to release a paper on that shortly. Busing fanatics in that city and elsewhere won’t be pleased. Neither will anyone who really wants to see the achievement gaps reduced.
By the way, education database messes are not limited to the federal government.
Every year the Kentucky Department of Education publishes a school report card for each school. Included in those report cards are the per-pupil spending amounts for that school. The problem is that for the latest, 2010-11 school data, several schools have obviously wrong figures listed in the database at the department. Nine schools had ridiculously low funding figures of only a few hundred dollars per student. At the other end of the spectrum, one school had an extraordinarily high funding figure of more than $99,000 per student. No one else caught that error, either.
Thanks to those data errors, I have to manually track down the correct numbers before I can complete a new report. That is a pain, but at least I know enough to be able to tell when a number is grossly out of line.
Not so, apparently, at USN&WR.(9 May 12: Wording changed to clarify that KDE does not report ACT scores disaggregated by race)