"A child's learning is the function more of the characteristics of his classmates than those of the teacher." James Coleman, 1972

Thursday, October 27, 2011

Tennessee Charter School Data Gives Good Reasons to Go Slow or Begin Backing Up

On Monday the Commercial Appeal in Memphis carried a story on efforts by local officials to take a deep breath and assess the charter school situation as these same officials work to create a Metro school system out of separate city and county systems.  This seems to be a reasonable proposal, but, of course, the hedge funders, vulture philanthropists, Koch Brothers puppet, Gov. Haslam, and other corporate welfare bottom feeders are in an instant lather to block any effort to call a timeout on the spread of the intensely segregated (by race and class) exclusionary charter schools in Memphis, with their punishing protocols that no middle class parent would consider for their own children.

Greg Thompson, former investment banking analyst and new chief of the Tennessee Charter School Incubator, is leading the charge for the uninterrupted flow of public money into these corporate charters.  He and the front office of KIPP, Inc. and TFA, Inc. argue that the spread of these urban chain gangs should remain unhindered by planning, evaluation, or just a little common sense. To bolster their insistent whine, they refer to "a report issued this month on behalf of the Center for Research on Education Outcomes (CREDO) at Stanford University," even though the State has yet to produce such a report despite repeated requests to various offices within Commissioner Kevin "TFA" Huffman's Department of Education.

So far, all that we have is a joint press release from TN's Achievement School District/New Schools for New Orleans, both of whom are supposedly contracting with CREDO for some kind of services to help determine which schools get the federal "innovation" grants.  The Press Release has never even been posted to the Web, even though it is dated September 30, 2011 (email me if you want a copy of it).  From the Press Release:


   So if a school does not have three years of data, then its "effect size" cannot be determined, right?   Again, from the Press Release:

Now here is the first oddity in this.  Of the 15 Memphis charter schools on the list of high flyers from 2008-2011, there are 5 schools with only one year of data, having opened only in the Fall of 2010:

Memphis School of Excellence
New Consortium of Law and Business
Circles of Success Learning Academy
Omni Prep Academy

Two others, Freedom Preparatory Academy and Promise Academy, have only been in operation for two years and, thus, have only 2 years of data.  Promise Academy is listed as the school with the highest math growth, and Veritas, just opened last year, is listed the school with highest reading growth.

Second oddity: Three of the schools on the "CREDO" list have simply awful results on the state TCAP tests, which are supposedly used by "CREDO" to arrive at the list of schools that are outperforming public schools:

Memphis Academy of Health Sciences in Memphis
Soulsville Charter School
Memphis Business Academy

The TCAP Composite Scores for these three schools are below, as they appear on the State's public site:

2011 School Value Added

Memphis Academy Of Health Sciences in Memphis

TCAP Composite

In 2010, the Tennessee Department of Education implemented significant changes in state testing. For the purposes of these analyses, the new testing has been equated so it can be reported on the same scale. The Growth Standard reflects these changes and is based on statewide student achievement in 2009.


Estimated School Mean NCE Gain
Grade678Mean NCE Gain over Grades Relative to
Growth Standard0.00.00.0
State 3-Yr-Avg0.2-0.4-0.6Growth StandardState
2009 Mean NCE Gain12.1 G*1.6 G*-4.4 R*3.13.4
Std Error1.10.80.80.50.5
2010 Mean NCE Gain8.8 G*-5.5 R*-7.3 R*-1.3-1.1
Std Error1.00.80.70.50.5
2011 Mean NCE Gain11.1 G*-0.9 R1.6 G*3.94.2
Std Error1.20.80.80.50.5
3-Yr-Avg NCE Gain10.6 G*-1.6 R*-3.4 R*1.92.2
Std Error0.60.50.40.30.3
Estimated School Mean NCE Scores
Grade678
State Base Year (2009)50.050.050.0
State 3-Yr-Avg47.147.046.7
2008 Mean


2009 Mean48.648.441.6
2010 Mean44.143.141.1
2011 Mean46.043.244.7
G*Estimated mean NCE gain is above the growth standard by at least 1 standard error.
GEstimated mean NCE gain is equal to or greater than growth standard but by less than 1 standard error.
YEstimated mean NCE gain is below the growth standard by 1 standard error or less.
REstimated mean NCE gain is more than 1 standard error below the growth standard but by 2 standard errors or less.
R*Estimated mean NCE gain is below the growth standard by more than 2 standard errors.


Soulsville Charter School in Memphis

TCAP Composite

In 2010, the Tennessee Department of Education implemented significant changes in state testing. For the purposes of these analyses, the new testing has been equated so it can be reported on the same scale. The Growth Standard reflects these changes and is based on statewide student achievement in 2009.

Estimated School Mean NCE Gain
Grade678Mean NCE Gain over Grades Relative to
Growth Standard0.00.00.0
State 3-Yr-Avg0.2-0.4-0.6Growth StandardState
2009 Mean NCE Gain2.6 G*-0.5 Y-0.4 Y0.50.8
Std Error1.30.90.90.60.6
2010 Mean NCE Gain-8.4 R*1.1 G*-3.5 R*-3.6-3.3
Std Error1.91.00.90.80.8
2011 Mean NCE Gain5.7 G*-2.9 R*-2.3 R*0.10.4
Std Error1.71.00.90.70.7
3-Yr-Avg NCE Gain-0.1 Y-0.8 R-2.1 R*-1.0-0.7
Std Error0.90.60.50.40.4
Estimated School Mean NCE Scores
Grade678
State Base Year (2009)50.050.050.0
State 3-Yr-Avg47.147.046.7
2008 Mean


2009 Mean35.640.544.7
2010 Mean31.336.737.0
2011 Mean44.728.434.4
G*Estimated mean NCE gain is above the growth standard by at least 1 standard error.
GEstimated mean NCE gain is equal to or greater than growth standard but by less than 1 standard error.
YEstimated mean NCE gain is below the growth standard by 1 standard error or less.
REstimated mean NCE gain is more than 1 standard error below the growth standard but by 2 standard errors or less.
R*Estimated mean NCE gain is below the growth standard by more than 2 standard errors.

2011 School Value Added

Memphis Business Academy in Memphis

TCAP Composite


Estimated School Mean NCE Gain
Grade678Mean NCE Gain over Grades Relative to
Growth Standard0.00.00.0
State 3-Yr-Avg0.2-0.4-0.6Growth StandardState
2009 Mean NCE Gain6.6 G*-2.9 R*-2.1 R0.50.8
Std Error0.90.91.20.60.6
2010 Mean NCE Gain7.0 G*-5.4 R*-10.0 R*-2.8-2.6
Std Error1.10.80.80.50.5
2011 Mean NCE Gain8.9 G*-3.4 R*-2.5 R*1.01.2
Std Error1.20.80.70.50.5
3-Yr-Avg NCE Gain7.5 G*-3.9 R*-4.9 R*-0.4-0.2
Std Error0.60.50.50.30.3
Estimated School Mean NCE Scores
Grade678
State Base Year (2009)50.050.050.0
State 3-Yr-Avg47.147.046.7
2008 Mean


2009 Mean45.242.637.2
2010 Mean43.036.631.7
2011 Mean42.132.431.4
G*Estimated mean NCE gain is above the growth standard by at least 1 standard error.
GEstimated mean NCE gain is equal to or greater than growth standard but by less than 1 standard error.
YEstimated mean NCE gain is below the growth standard by 1 standard error or less.
REstimated mean NCE gain is more than 1 standard error below the growth standard but by 2 standard errors or less.
R*Estimated mean NCE gain is below the growth standard by more than 2 standard errors.


Finally, consider the case of KIPP, which is listed as #2 on the list of "CREDO" high flyers.  It is hard to say which public schools they were matched with, but when matched with the state average, they did not do so hot, or at least there were no detectable differences (NDD) between KIPP's Explore Math scores or their End of Course Algebra I scores when compared with the State Averages.  Again, from the State's public site:


















































2 comments:

  1. Anonymous11:23 PM

    This is very interesting, and disturbing. Amazing how data can be misleading.

    ReplyDelete
  2. Anonymous8:26 AM

    This comment has been removed by a blog administrator.

    ReplyDelete