Gatto’s thesis is that forcing children to write about Cortez when they’d rather write about something else has an effect that is the opposite of education. I don’t want to get too lengthy here. I’d rather let you enjoy the book for yourself–it’s very well written and full of facts that require one to ruminate deeply. In the end, Gatto will blow your mind, even if he doesn’t change it.

My point is not that learning, or even schooling, of any kind is bad. I was saying that the formula by which schools achieve their version of education is fatally flawed, and more often results in lobotomization than in education. Schools hijack curiosity and enslave it to the clock, the bell, and to the mindless formulaic repetition of figures that are unimportant to the free mind. They force children to seek the approval of a teacher (in most cases, a complete stranger) in the form of grades or verbal praise, rather than to be proud of their own work for their own reasons.

For a less destructive alternative to the current model, take a look at the Sudbury Valley schools (http://www.sudval.org/)

Something’s amiss.

Students hate school until it functionally lobotomizes them, and then they graduate. They survive thereafter as barely-literate automatons who do as they’re told by the folks whom they mistakenly attempt to please. Then they experience an externally-imprinted version of success, have their own children, and then nostalgically subject those children to the same machine that ground up their brains for 12-16 years. Rinse, repeat, at infinitum, since 1895.

See John Taylor Gatto’s work for more on this (www.johntaylorgatto.com).

There may also another reason. When you apply phrases such as uniforms, no touching, no talking, no mandatory attendance, lockdowns, zero tolerance, police monitors, security cameras, barbwire, body searches, property searches and forced “community service” surrounding what you do in a “room” all day; where is the difference between school and the state penitentiary.

Margo may be correct in noting that WWC does not use state assessment data–but all of WWC’s vetted program assessments are linked to standardized testing data, either the NAEP or another nationally normed test. And standardized tests thus measure how well the curriculum is aligned to test items. Tests should be aligned to curricular goals if they’re measuring instructional effectiveness (or, especially, the value of a particular curriculum).

Lots of people who don’t understand that–or have an axe to grind– believe that WWC has cut through the fog and can reliably tell us which programs a school “should” adopt. The premise behind WWC is elevating randomized experimental trials as the best way to do social science research, and using standardized test data as the one best measure of student learning. And that–is a bias.

The recent IES study on the four math programs, with its “puzzling” results of two very diverse math programs leveraging the highest scores was not so puzzling when the items being tested were compared with curricular goals embedded in each program. And when the researchers casually noted that the Saxon program students received approximately an extra hour of instruction per week, per program recommendations– well, so much for pure, bias-free science.

Exclusively using standardized tests to measure learning is a bias against human judgment (which is used in the workplace and the real world as often or more often than measured outputs). Anyone who’s ever sat around a table trying to understand why a child with off-the-charts IQ numbers did poorly on a standardized test, while a kid whose daily work is mediocre scores at the top of the scale will tell you that standardized tests tell us some things, but not everything. And those huge data sets, while seductive to psychometricians and economists, are not impacting learning as much as one-to-one relationships and other soft, “anecdotal” practices. You can measure learning through student work products, too. It just takes a great deal more time and skill.

I agree that the econometric model of measuring student learning is ascendant, and controlling the discourse at the moment, with the attendant narrative of equating test scores to accountability. But more rigorously controlled research models haven’t gotten us where we want to go in improving student learning, have they? They’ve just given us more data. And more data doesn’t always lead to better choices.

I don’t pretend to be a research expert. I’m not. I’m a lifelong teacher, who had the opportunity use a very engaging curriculum model. The only scientific evidence I have was my observation that the content was challenging, and often hit the spot between what kids had mastered, and problems that pushed their mathematical thinking. And I repeat–I would never choose a curricular program based on standardized test data–I would choose one based on the learning goals embedded in the program, not the items selected by the test developer.

What I find intriguing is the WWC’s lack of ideological bias. Both Every Day Mathematics and Connected Math are constructivist in outlook, but the WWC found one to be effective while the other was ineffective.

It may be that CMP is effective in other ways such as improving problem solving abilities or increasing algebra readiness, but the burden of proof lies with CMP. Surely, someone at the University of Michigan could create a randomized experimental or quasi-experimental test that would demonstrate whether or not CMP is actually effective in these ways.

Thanks for the note. I am a huge Seinfeld fan and watch the reruns religiously – not that there’s anything wrong with that. Tonight it’s Darren the intern working at Kramerica Industry. A scream.

That must have been a challenging experience jumping into a math class like you described.

I think that you are misinformed with regard to What Works. I have been following their publications for a while, and while it is true that they are looking for quantitative randomized experimental or quasi-experimental studies, the data is based on sources other than state tests. Certainly one of the overall impressions from what works is that there is not so much available that meets their criteria. Certainly the alignment of a curriculum to state standards would be a worthy consideration when selecting a curriculum, particularly as so many teachers are heavily reliant on what is included in the text. Certainly there would be other criteria as well.

I know that there are lots of lofty discussion regarding how research is most appropriately done and the absolute value of knowing and all that, and whether anything can truly be randomized (or known). But, I remain skeptical of studies that do not have some basis (such as randomization or a quasi-experimental design) for claiming an effect. There is certainly also room for study to demonstrate an effect for a particular group (low SES, girls, hearing impaired, whatever) that may not generalize, but may still be useful.

I haven’t seen too many WWC reports that found more than a few studies focused on what they were looking at. Says a whole lot more to me about the need to support educational research than about any WWC bias or about the fallibility of things being done in the classroom.

What the data might actually be measuring is alignment of the CMP curriculum to individual state tests. If a state test is mainly composed of arithmetic calculation items, some other math series may well do a better job of preparing kids. If the test includes constructed items (which are more expensive to score, but generate different kinds of data), or problem-solving, a program like CMP is likely to do a better job. The “evidence” that WWC has provided on any number of strategies or materials is skewed toward a particular conception of what kinds of research can be trusted, but in the social sciences there is never true randomization.

I would never adopt a curriculum solely on the basis of standardized state test data–and if you’ve been following the National Standards discussion, you’ll know that any number of state tests have recently been skewered for being constructed around low-level skills.

I also think that Rich brings up another factor: socio-economic factors, including earlier education experiences in a subject, strongly impact the value of a particular kind of teaching. Although I teach in an area considered rural, we are on the outskirts of the wealthy suburbs. About 90% of the kids in my district go on to college. Last year, about 85% of MS parents in my district showed up for PT conferences. And they all care about rigorous math instruction. In fact, on Parent Open House night, I shared the CMP books with parents. Lots of automotive engineers sitting in the room–and they liked it. A few confessed that helping with homework was a new challenge–they actually had to think about how to set up problems.