Thanks to our illu­sory supe­ri­or­ity we con­sis­tently over­es­ti­mate our per­for­mance on tests, and, with­out qual­ity feed­back, rapidly become obliv­i­ous to the gaps in our knowl­edge. Fur­ther­more, many con­sider test­ing to be an inef­fec­tual tool for assess­ing per­for­mance and errors to be coun­ter­pro­duc­tive to learning.

Chal­leng­ing this pre­con­cep­tion is research sug­gest­ing that mak­ing mis­takes on tests–and being informed of them–is an inte­gral part of the learn­ing process.

We tend to assume that the best way to con­sume and remem­ber infor­ma­tion is through the appli­ca­tion of rig­or­ous, extended study. What we fail to see, however, is that the process of try­ing to work through a prob­lem to which we don’t know the answer focuses our atten­tion on it in a way that sim­ply study­ing it does not. The desire to get the answer right, and the frus­tra­tion of fail­ure, is partly to account.

But there’s another ele­ment as well. When we strug­gle to learn some­thing, and fail, the moment we finally get the answer it imprints itself more deeply on our mind than it would have had strug­gle and fail­ure not pre­ceded it. […]

If I had to iden­tify one over­ar­ch­ing les­son from  our study it would be this: When you make mis­takes, don’t just let them slip by — cor­rect them. Cre­ate chal­leng­ing learn­ing envi­ron­ments, make mis­takes and then learn from them.

There is much in com­mon here with the evidence-based approach to teach­ing.