At a policy level, decisions are made based on this kind of blind testing. In order to have your research accepted to the policy discussion it MUST adhere to this kind of testing… this is the problem. David didn’t add this in his post, but I’m pretty sure that’s what he’s talking about.

Mean deviations, whatever they may reflect, are considered to acquire truths based on how they remove the human factors from the equation. They are successfully constructed the more they remove context from the equation… the more they are ‘standardizable’. This implies that there is A correct or A SERIES OF correct ways of teaching and that they are observable. It also subsumes the discussion of what we are teaching in the first place. If we are teaching ‘content’ or ‘literacies’ or ‘models’ the results will differ.

No one, I think, is trying to imply that this results in teachers wandering around like single-minded robots teaching each student as if they are the same as another. It means that the underlying research methodology that is acceptable for gov. funding and application to policy ascribes to this worldview so that it can be ‘verifiable’. It does not address wether this verifiability is desirable or even possible.

It posits a world where there IS verifiability, whether that world reflects the real classroom or not.

However, when colleges put a few hundred students in the same lecture hall they are treating them as if they were all the same.

But even if we could afford to give every student his/her own private tutor (which some AI buffs think might happen someday), would we really want students to learn in isolation from other students.

But if we put students in group learning environments, do we want those groups to be homogeneous or heterogeneous. One could easily argue both ways.

So, to return to the original question/issue, if we put students in heterogeneous groups, that implies that we’re teaching students with dis-similar learning styles/abilities in the same learning environment (i.e. we’re treating them in the same “way” even though we know they’re different).

Alternatively, if we group students by learning styles/abilities then we open ourselves up to the charge that we aren’t providing a “diverse” learning environment.

Finally, if one tries to strike a compromise by providing some one-on-one tutoring, some homogeneous group learning, and some heterogeneous group learning then critics can always complain that we haven’t got the mix “just right.”

For those who like clear cut solutions, the problem is that education is a “divergent” problem situation, unlike a “convergent” problem such as putting eight queens on a chess board so that none of them can attack one another. There is a solution to the eight queens problem because the rules of chess are clear cut and the objective is well defined and agreed upon.

In education, the rules are not clear cut and the objective is not well defined. So, issues in education policy cannot be solved like eight queen problems.

Those who look for “eight queen” like solutions try to impose well defined rules and objectives. But, those who object to the outcome of such an analysis can always complain about the initial definition of the rules and objectives. So, we continue to spin our wheels.

The above blog sounds like college classrooms instead of elementary and middle school ones to me.

It’s the same discussion as the “evidence based medicine” debate. Off course randomized trials, double blind, etc. etc. is are indispensable tools, but of course they are not applicable in every situation.

The irony is that people tend to be hired based on how they make the hirer feel. Perhaps it’s appropriate, then, that what most people learn from their school years is how to fit in and do as they’re told.

Of course, when they’re fired, it’s because they’re just cogs in the system. And one cog’s as good as another when it comes time to “shed excess weight.”

If we couldn’t generalise across a body of people, nothing would work – ever. If everyone was completely different, life would be a complete shambles. The fact that we all have so much in common in terms of cognition is what makes us such a powerful species. What works in this situation with these people may very well work with those people in that situation, all things being generally equal. Of course we must adjust for the individual as required, but I guess what well-designed studies do is try to give a foundation from which to adjust.

I think what’s most interesting about the What Works Clearinghouse is seeing that so many studies that are used to “scientifically prove” the effectiveness of a program, actually fail to meet a rigorous standard of research. Kind of devalues any claims to being “research-based”.

My point is, I don’t have a problem with randomized control-group studies, and research clearinghouses because they tell us what works MOST of the time. But I DO have a problem when policymakers or others put too much faith in these studies. The best approach, I feel, is to know the scientific research, and then make the best judgment based on the students involved.