View in your browser.

Welcome

Today, the Center for Research on Education Outcomes (CREDO) at Stanford University released an updated study of charter school performance in selected states, including North Carolina.  The bottom line is that charter schools, both nationwide and in North Carolina, continue to improve.

Bulletin Board

  • Attend. A list of upcoming events sponsored by the John Locke Foundation can be found at the bottom of this newsletter, as well as here.  We look forward to seeing you!

  • Donate. If you find this newsletter mildly informative or entertaining, please consider making a tax-deductible contribution to the John Locke Foundation.  The John Locke Foundation is a 501(c)3 not-for-profit organization that does not accept nor seek government funding. It relies solely on the generous support of individuals, corporations, small businesses, and foundations.

  • Share. The North Carolina History Project seeks contributors to the North Carolina History Project Encyclopedia. Please contact Dr. Troy Kickler for additional information.
  • Discuss. I would like to invite all readers to submit brief announcements, personal insights, anecdotes, concerns, and observations about the state of education in North Carolina.  I will publish selected submissions in future editions of the newsletter. Requests for anonymity will be honored. For additional information or to send a submission, email Terry at [email protected].
  • Revisit. We have archived all research newsletters on our website.  Access the archive here.

CommenTerry

Before we get to the 2013 Center for Research on Education Outcomes (CREDO) charter school report, let’s go back in time.  In 2009, CREDO published a report that was somewhat damning for charter schools, particularly those in North Carolina.  The report showed that North Carolina charter school students performed better than their traditional public school peers in reading and worse in math.  Overall, the results showed that North Carolina charter school students had results that were no different than the gains for traditional school peers.  One likely explanation for this was the presence of the 100-school cap on charters.  Researchers found that "the presence of caps puts significant downward pressure on student results." (p. 40)

Since 2009, lefties, particularly those who seek to discredit the charter school movement in North Carolina, have often mentioned the performance results without a word about the effects of the charter cap.  Few in the mainstream media have challenged their selective reading of the 2009 CREDO study and most have joined their associates on the far left in the chorus declaring charter schools to be dismal failures.  We found out today that there was a serious error associated with charter school data from North Carolina.  I’ll discuss that a bit later.

The 2013 CREDO study shows that North Carolina’s charter schools improved in both reading (from 0.00 to +0.03) and math (from -0.03 to -0.02) since the 2009 study.  Nevertheless, charters maintain a slight disadvantage in math performance compared to their district school counterparts.  According to the study, the state’s charter school students received the equivalent of seven fewer days of instruction than comparable students in district schools.  That is no surprise.  In recent years, North Carolina’s math performance has been superb.  District school students in fourth- and eighth-grade consistently score above the national average on the rigorous National Assessment of Educational Progress (NAEP).  In other words, the fact that charter schools are moving closer to district school performance is extremely good news.

On the other hand, our district schools have struggled to match the national average on the NAEP’s fourth- and eighth-grade reading test.  If charter schools performed no better than district schools in reading, then we should be alarmed.  Fortunately, North Carolina’s charter school students easily eclipsed the district student performance, providing charter students the equivalent of one school month of additional instruction.

But the above findings do not tell the entire story.  As a group, North Carolina’s charter school performance surpasses the statewide average academic performance of all tested students.  In reading, North Carolina is 0.23 standard deviations above the state average in English and 0.13 standard deviations above the average in math. In fact, North Carolina charter schools have the highest mean in both reading and math of any state in the study.

Nationally, charter schools continue to improve, although it is critical to note that performance varies from state to state and subject to subject (See Facts and Stats below).  In reading, charter school students in the study had larger learning gains in reading and similar gains in math relative to comparable district school students.  Researchers concluded that three groups — students in poverty, African-Americans, and English language learners — gained "significantly more days of learning each year in both reading and math compared to their traditional public school peers."  Moreover, charter school enrollment has become more racially and socioeconomically diverse.

What accounts for these gains?  Researchers attribute them to the fact that weak charter schools close and better ones take their places.  In other words, academic accountability, paired with the charter closure mechanism, appears to be working just as charter school advocates claimed it would.

Interestingly, CREDO researchers identified significant errors in two sets of data, from North Carolina and the District of Columbia, that alter the 2009 findings.  Researchers report, "In North Carolina, 9,449 student records were incorrectly reported to us as charter records in the 2006-2007 school year dataset." (p. 26) The fact that CREDO received thousands of records that were improperly coded, presumably by the NC Department of Public Instruction, raises serious questions about the competence of those who manage data collection and analysis.

Not everybody agrees with the methodological approach used by CREDO researchers.  In 2009, Caroline Hoxby of Harvard University published a researcher memo titled, "A Serious Statistical Mistake In The CREDO Study Of Charter Schools (PDF)."  Hoxby argued that the 2009 CREDO study, "causes a negative bias in its estimate of how charter schools affect achievement."  As a result, she contends that the report findings are not valid.  Some, including the Center for Education Reform, have criticized the 2013 report and I suspect that several more groups will follow suit.

We should be mindful of arguments from those who object to CREDO’s charter school studies.  Indeed, participants in the charter school debate should recognize that all research studies have methodological limitations and other challenges that have a direct bearing on the results.  More importantly, it is unwise to derive broad generalizations about charter schools from a single report, regardless of the size and scope of the study.

Facts and Stats

State Charter School Impacts, 2013

State

Reading Charter Impact

Reading Days of Learning

Math Charter Impact

Math Days of Learning

Arizona

-0.03**

-22

-0.04**

-29

Arkansas

-0.03**

-22

-0.03**

-22

California

0.03**

22

-0.01**

-7

Colorado

0.01**

7

-0.01**

-7

District of Columbia

0.10**

72

0.14**

101

Florida

-0.01**

-7

0.00

0

Georgia

0.02**

14

-0.02**

-14

Illinois

0.02**

14

0.03**

22

Indiana

0.05**

36

0.02**

14

Louisiana

0.07**

50

0.09**

65

Massachusetts

0.05**

36

0.09**

65

Michigan

0.06**

43

0.06**

43

Minnesota

0.02**

14

-0.01

-7

Missouri

0.02**

14

0.03**

22

Nevada

-0.16**

-115

-0.19**

-137

New Jersey

0.06**

43

0.08**

58

New Mexico

0.00

0

-0.04**

-29

New York

0.05**

36

0.11**

79

New York City

0.00

0

0.13**

94

North Carolina

0.03**

22

-0.01**

-7

Ohio

-0.02**

-14

-0.04**

-29

Oregon

-0.03**

-22

-0.07**

-50

Pennsylvania

-0.04**

-29

-0.07**

-50

Rhode Island

0.12**

86

0.15**

108

Tennessee

0.12**

86

0.10**

72

Texas

-0.03**

-22

-0.04**

-29

Utah

-0.01

-7

-0.06**

-43

** Significant at p ≤ 0.01

Source: CREDO, "National Charter School Study 2013," June 25, 2013, p. 51-52.

Education Acronym of the Week

CREDO — Center for Research on Education Outcomes

Quote of the Week

"Students in poverty, black students, and those who are English language learners (ELL) gain significantly more days of learning each year in both reading and math compared to their traditional public school peers. Performance differences between charter school students and their traditional public school peers were especially strong among black and Hispanic students in poverty and Hispanic students who are ELL in both reading and math."

–  CREDO, "Press Release: Charter Schools Make Gains, According to 26-State Study," June 25, 2013, http://credo.stanford.edu

Click here for the Education Update archive.