What's Wrong with Educational Sciences?

One of my favorite websites to visit is called Eurekalert at http://eurekalert.org.  It is an aggregate of press releases from universities, colleges, and research facilities.  And, of course, as an educator, I find myself scanning through the education section looking for interesting articles to tell my students about and post to this blog.  The problem is that the educational sciences are not that rigorous.  Studies rarely meet the standards set by the medical community, which require the randomized double-blind method.  

The randomized double-blind study is the gold standard because it minimizes confounding variables.  However, this standard is rarely met in the field of education.  For example, students are often selected randomly, but rarely from the general population.  Now, that is not to say that these studies aren't worthwhile, for they can be used for meta-analytical studies later.  Taken on their own, however, they should not be extrapolated out to the general public.  Or, as in the case of charter schools, are not randomly selected from the general population but from a self-selected group of students who are biased towards charter schools.

The other problem with methodology is the placebo effect.  Students know they are going to a different school or can figure out when something dramatic has changed in the curriculum.  This creates the idea of a treatment to the student.  We know from research that the placebo effect counts for up to 30% of observable improvement even if the person knows they are receiving a placebo. 

In conclusion, the science problem eventually leads us to the present policy problem: wanting evaluations for teachers that aren't scientific.  Any evaluation of a teacher that isn't scientifically based is a fraud and a waste of tax payer dollars.  The question is this.  Do we really want to solve the problem? Or are we simply dumping blame at a group of people because we can and it is easy?