Winnipeg’s Charter School Confusion

Publication, Education, Frontier Centre

Frontier responds to Letters to the Editor in the Winnipeg Free Press criticizing a recent piece on charter schools.

Letter 1: Reviews are mixed on charter schools

This is in response to Peter Holle’s article Charter schools receive high grade, which appeared on Aug. 16 on your editorial page.

Mr. Holle’s proof that charter schools provided better services vis-a-vis conventional public schools came from a select number of charter schools across Canada. He compared them individually to other schools in their districts.

The same day, however, the New York Times revealed that a national assessment done by U.S. President George W. Bush’s education department clearly showed, on average, charter schools in the U.S. lagging far behind their public counterparts. This assessment showed that Grade 4 students in charter schools were half a year behind in math and reading.

Perhaps Mr. Holle should draw his conclusions after a careful examination of available information, rather than beforehand, as ideologues are so often prone to do.

BEN GRANT, Powerview

Frontier responds – Sample of U.S. Study Needs Careful Examination

Granted our comparison was drawn from a couple of charter public schools; however, research conducted by three university professors at the University of Alberta and University of Calgary support the statement charter schools receive high grade. Long waiting lists, high parent, teacher, and student satisfaction are strong evidence that charter schools are meeting the needs of many families and educators.

It’s important to review the demographics of students in the USA who attend charter schools. The majority are students at risk, stem from low socio-economic backgrounds, and are from minority groups—all factors that, according to research, have significantly contributed in low student achievement. (See response to third letter below for a more “careful examination” of the New York Times report)

Letter 2: Parents: Don’t leave public school system

In the article Charter schools receive a high grade (Aug. 16), the author asserts “that the slide in public school quality and results continues apace, especially in the challenged inner-city core.”

He concedes that the problem of providing a quality education for every boy and girl is a complex one. It is a complex problem in that Canadians who believe in the common good are committed to providing every child, regardless of ethnic background, sex, religion and economic status, with an equal opportunity to succeed. The means to this end has been our public education system.

How should the problems in that system be addressed?

The writer concludes that, “More school choice means more learning.” In his view, the best means for improving learning are charter schools, that is, schools that a group of individuals can organize and manage with public funding and with some government oversight. Implicit in such schools is that they can determine the curriculum, determine who can attend and charge fees, and that parents and others may contribute funds or gifts in kind.

This approach to improving learning can best be characterized by Tommy Douglas’s remark, “Every man for himself, said the elephant dancing among the chickens.” Consequently, those parents who are motivated by what they perceive as best for their children and who have the money will, given the opportunity, leave the public system for charter schools.

Would not their staying in the public system to help strengthen it and to solve its complex problems better serve the public good?

VIC PRUDEN, Winnipeg

Frontier responds – Charter schools are public schools:

Vic obviously lacks understanding and knowledge about charter public schools. The schools are an option under the public education system.

Charter public schools:

  • Are attended by choice
  • Charge no tuition
  • Have no religious affiliation
  • Employ certified teachers
  • Are locally governed
  • Encourage parental involvement
  • Have a unique focus
  • May not discriminate in admissions policy
  • Must teach the mandated provincial curriculum
  • Are accountable to the public
  • Parental affluence is not a factor in choosing a charter public school as there are NO tuition fees. The provincial curriculum must be followed. Vic, you are right that as part of the public system, the ripple effect from charter schools positively influences neighbouring public schools—all for the public good.

    Interestingly, any public school can accept donations or gifts in kind and provide tax deductible receipts as they are registered nonprofit organizations. Check with your local school district/municipality.

    Letter 3: Writer ‘spoke too soon’ on charter schools

    Peter Holle spoke too soon when he proclaimed that Charter schools receive high grade (Aug. 16).

    The same day, the New York Times reported: Nation’s charter schools lagging behind, U.S. test scores reveal. The article adds: “The first national comparison of test scores among children in charter schools and regular public schools shows charter school students often doing worse than comparable students in regular public schools.” U.S. supporters of charter schools are scrambling to find plausible explanations that will help them turn teaching over to church groups and private companies.

    DANIEL STONE, Winnipeg

    Frontier responds – Test Score Reliability a Problem:

    According to Education Analyst, Richard Innes, the national comparison of test scores are unreliable (see complete response that follows). Putting the comparison aside, Daniel appears to be misinformed about charter public schools. In Alberta, for example, all public charter schools are led by teachers holding Alberta teaching certification, have several years as teachers under the conventional public school system, have held positions as principals, central school office administrators, or with Alberta Learning.

    The following detailed response comes from Richard G. Innes, Education Analyst, Bluegrass Institute for Public Policy Solutions ( an independent think tank based in Bowling Green, Kentucky.

    There has been a considerable amount of discussion on the American Federation of Teachers’ new report:


    The AFT’s basic conclusion, based on an analysis of NAEP data, is that charters are not performing as well as regular public schools.

    I dug into the NAEP Data Tool (which the AFT used to get their data) and tried to see where AFT actually got their numbers. So far, I have discovered several interesting things.

    First, there is no breakout of the exclusion rates in the different types of schools. As long-time members of the list know, I discovered exclusion can be a major problem in 1999 when Kentucky posted a big increase in 4th grade NAEP reading scores compared to their 1994 performance. It turned out that the entire 6 point rise was actually questionable once one considered that the Kentucky exclusion rate for students with learning disabilities had exploded from 4 to 10 percent of the raw sample NAEP selected for testing. If you cut off the bottom 10 percent of the bell curve, your average scores will certainly rise, but it won’t mean anything.

    Sadly, without this data, all the NAEP data for charter schools is of dubious value.

    But, I discovered something else even more interesting. In the same section of the Data Tool where I found the question with the charter school scores, I also found some other questions regarding school type.

    The lead question for these parts of the Data Tool asks: “What type of school is this?”

    The charter school 4th grade math scores are found by selecting the answer in Table 159: “Charter school.”

    For the District of Columbia, this kicked out the information that a total of 2,614 students were tested in DC and 12% were in charters while 88% were in schools that showed “no response” to this answer. The charters scored 203 with a standard error of 1.6 and the “no response” schools scored 205 with a standard error of 0.7.

    As a note, it is 95% probable that the real score for a school will be within 2 standard errors of the published score. So, for the charters, the real score could be anywhere from 199.8 to 206.2. That means the charters could have outscored the other schools, but the NAEP is too imprecise to tell. This is a problem with most of the other charter data, by the way. So, NAEP measurement error seems to large to support the AFT’s conclusions. But, that isn’t the really interesting finding.

    I also checked out the other possible answers to “what type of school is this?” Table 165 provided an interesting result. This table’s answer is “Regular elementary school.” For this response, the “Yes” category represented 87% of the sample and scored 204 with a standard error of 0.7. The “No Response” category amounted to 13% of the sample and scored 209 with a 1.7 point standard error.

    Now, the interesting question is whether the charter schools reported themselves as regular elementary schools because their more logical answer would have been in the earlier table 159 as “charter school.”

    Note that the percentages for these two questions are such that the charters could have all been grouped with the “no response” category and still fit the Table 165 answer nicely. However, the “No Response” schools in answer 165 scored well above the other schools, which is a sharp conflict with Table 159. Thus, there could be a serious problem with the data in the Data Tool.

    This isn’t conclusive, of course. Perhaps the charters answered “yes” to both questions. But, this is something that needs to be clearly sorted out before much faith can be placed in the charter school data in the Data Tool. And, this might explain why the charter school report is taking so long to get released.

    One other interesting problem showed up. Beginning in Table 168, the question changed to “Is this a school with a special focus on any of the following?” Table 168’s answer is “arts.” However, in this table, the total sample size suddenly dropped from 2,614 students to only 2,312 students. Where did the other kids go? Also, the scores are *dramatically* lower for the “yes” and the “no response” answers to this and the following questions. In Table 168, the “yes” results show a score of 194 with 7% of the students included. The “no response” results are a score of only 187 and this represents 93 % of the sample.

    Clearly, the data in Table 168 does not agree at all with the earlier data.

    So, at this point, there seems to be some very interesting anomalies in the NAEP data that AFT relied upon in its report. Absent more information about the anomalies, I would treat this NAEP data as unreliable. So much for NAEP being the “Gold Standard,” as AFT suggested.

    More on this in September when I have some time to get some answers.

    Richard Innes