Tuesday, February 23, 2010

Multiple Measures of the Same Data

RIDE:

The Rhode Island Department of Elementary and Secondary Education’s method for identifying persistently lowest-achieving schools as defined in this Protocol includes analysis of the following factors:

  1. School-wide student performance in mathematics and reading against the state- wide average performance in these subject areas;
  2. No Child Left Behind Classification with respect to number of years in need of improvement;
  3. Subgroup performance in reading and mathematics against the state-wide average performance;
  4. Student growth percentile at elementary and middle school levels in reading and mathematics and graduation rates at high school levels against the state-wide average growth; and,
  5. School-wide improvement in reading and mathematics between 2005-2006 and the 2008-2009 school years against the state-wide average improvement.

Since AYP is based on math and reading scores, and graduation rates, broken down by subgroup, taking into account annual growth, these five factors are just chewing over the exact same data five different ways.

I still have trouble believing that I live in a world where decisions to close schools would be made without even bothering to move your finger over one more column on the NECAP report to look at the writing scores. Hey! You already paid for them! They're right there. See?

Never mind all the data RIDE has collected every year for the past decade in surveys of parents, students and teachers. Never mind RIDE's formerly strong and unique (in the US) school inspection system. And I have a strong suspicion that they couldn't be bothered to look at five year graduation rates either.

Never mind college admissions data. Never mind college retention data.

And, for the record, the test data for the "2008 - 2009" school year is the data that came out after the "lowest-achieving" schools were announced.

I'm about ready to go back to concentrating on internet spaceships.

No comments: