Are “Best” Practices Really the Best? What is the evidence?

edited 12/20/20

Perfectionism is the voice of the oppressor. -- Anne Lamott

Imagine walking into a Cadillac dealer and saying, “Don’t waste my time with long-winded explanations, just tell me: what is the best car?” If the dealer responded, “The best car is a Cadillac,” would you be surprised?

Educators are more subtle. Ask an eighth grade science teacher what the best teaching method is and she will likely reply, “For what kind of student.?” You pull out your list of pigeonholes and say something like, “A middle class, eighth-grade, male student not in special ed.” If the teacher responded, “The best method is the one I use,” would you be surprised?

Interestingly enough, you will occasionally get the reply, “The best methods are ones we are not permitted to use in this school.” This happens especially in foreign language or math where parents or school board members insist that the kids be taught as they, twenty to fifty years earlier, were taught; educational research, or whatever, be damned.

Why do we lust after “best” practices? Why not be satisfied with ones that are “generally good,” or “adequate” and the like. After all, to be literally a “best practice” something has to be proven to be better than all alternatives for a particular kind of student, with a particular kind of learning history, in a particular schooling environment studying a particular subject.

Consider this: if there were only 5 possible kinds of students, with 5 possible kinds of learning history, with 5 possible kinds of learning environment, studying one of 5 possible subjects, then for a single subject we would have to test and compare 5x5x5x5 = 625 possible situations to determine the best of the 625. Being too costly and time consuming, such research is not being done. Given parental concern that their own kids always get what they consider "the best" the school has to offer (Test new methods on someone else’s kids!), it is unlikely it will ever be done.

If you are interested in pursuing this further, see

Are "Best Practices" Good Enough?

-- EGR