we can't selectively invoke parental satisfaction

I think it's important for those of us who are critical of the school reform movement to be consistent in our application of evaluative criteria. Part of that lies in taking empirical measures that demonstrate benefits or strengths of reform efforts we don't like as seriously as those that demonstrate deficiencies or weaknesses. There are of course complications. One of the difficulties for the reform movement is that it typically weds a strong faith in standardized testing as effective assessment to a belief in particular mechanisms for raising them. That means that when structures like private school vouchers are advocated as a solution for raising standardized tests scores, and then fail to, those advocating them have to deal with their preferred method of improvement failing on their preferred method of assessment. Those critical of much of the school reform movement often criticize both the method of improvement and the method of assessment-- leading us, at times, to embrace data that discourages particular reforms even though we might reject the tests that were the mechanism through which the data was gathered.

(The key, I think, is to be careful in saying "by your own preferred method of assessment, your program is failing to achieve the gains you have predicted"-- and incidentally, I'm actually fairly amenable to certain kinds of standardized testing in certain contexts, at least compared to other school reform critics.)

Additionally, I will continue to say that the mere efficacy of anti-union school reforms is not enough to compel their adoption. Union rights are rights. They are not legitimately curtailed simply because it becomes convenient to society to do so. School reformers need to do more than demonstrate the effectiveness of their proposed reforms; they need to demonstrate that implementing them will not illegitimately trod on union rights of teachers. (Of course, until the school reform demonstrates consistent, valid, and reliable gains that are not later revealed to be the product of fraud, the question is somewhat academic.)

Anyway, to be more specific and useful, I want to say that we need to take care not to trumpet parent satisfaction data as proof of success in public schools when we wouldn't do so when it comes to charter schools, private school vouchers, and the like. Many have pointed to encouraging statistics about parental satisfaction with their local public schools, and also to the disconnect between a parent's perception of his or her own child's school and American public schools in general. USA Today summarizes a study by PDK:

Nearly eight in 10 Americans — 79% — give an "A or B" grade to the school their oldest child attends, according to findings released today by Phi Delta Kappa (PDK) International, an educators association. That's up from 68% in 2001, and the highest percentage of favorable ratings since PDK began asking the question in 1985. That year, 71% of parents gave their kids' school top grades.

It is indeed encouraging and important that Americans feel their local public schools are succeeding, and there are some complex epistemelogical questions about how to evaluate and incorporate this kind of data. But let's be clear: if we consider this evidence for the success of public education, we must also consider similar data evidence of success in charter schools.

Susan Phillips's School Choice: Policies and Effects: An International Literature Review (2004), while not a text that I would endorse without qualification, has a good rundown of extant evidence. See also Buckley and Schneider (2006), available here (PDF), and this Pew Research Institute report (2010, PDF) which includes satisfaction data for parents of students in charter schools, local Catholic schools, and public schools in the Philadelphia area. Generally speaking, a broad collection of data suggests that parents rate themselves as highly satisfied with their child's charter schools.

As I've said many times, there is a fierce debate about how we develop knowledge and what constitutes appropriate epistemology hiding in the school reform debate. It's perhaps unsurprising that parents of both public school students and private school students profess high satisfaction with their child's school. Parents are the definition of invested respondents; to rate low satisfaction for your school is essentially to say that you're failing your child. I don't take parent's self-reported satisfaction too seriously for school quality for the specific reason that it's quite hard for anyone to effectively rate a school's quality and the general reason that self-reported data has to be taken with a handful of salt.

And, indeed, the self-same Pew report that shows high satisfaction levels with Philadelphia charter schools admits that this satisfaction comes despite "widely publicized reports of financial mismanagement at several schools and test results indicating that students in some charters are not performing as well as those in district-run schools." So we have again this basic dynamic in school reform: empirical evidence contradicts conventional wisdom, deductive thinking, and the opinion of interested parties. I will continue to say that we have to privilege that empirical evidence over the alternatives, which for now redounds to the benefit of critics of the reform movement like me. But I also think that we have to be consistent in how we evaluate our evidence.