Testing creates better memory than constructivism?

LabOutLoud, a podcast and blog for science teachers, posted a tweet with the intriguing text "To really learn, quit studying and take the test".  This linked to the blog Blaseria with a picture of a graph showing student's recall of a science text after four different study methods.  This blog linked to an article in the New York Times based on an article in the journal Science.  So I read the article (Epub ahead of print in the week's Science), and this is what I think.

We know that each time you access a memory, you change it.  Drs Karpicke and Blunt at Purdue wanted to know whether retrieving memories also helps you learn them, and if it helps you learn any differently than other popular study methods. 

Undergraduates studied a science text under one of four conditions:  study-once, where they studied the text in a single study period; repeated study, where the studied the text in four consecutive study periods; elaborative concept mapping, where they learned how to make concept maps and created concept maps while reading the text; and retrieval practice, where they studied the text then practiced retrieval by recalling as much of the information as they could on a free recall test.  The latter three strategies were controlled so students spent the same amount of time with the text.  One week later, students returned and took a short answer test including verbatim and inference questions, which assess different depths of conceptual knowledge. 

Students who used retrieval practice performed significantly better on the test than did students who used the repeated study or concept mapping methods.  Single study students did the worst. 

They repeated the experiments with 120 new students, all of which studied two texts and used cognitive mapping on one and retrieval practice on the other, and half took the original short-answer test and half took a cognitive mapping test.  In all cases, students who studied using retrieval practice still outperformed other students. 


(figure from the New York Times)

So what does this all mean for us as teachers?  I think the important thing to note is the method of the retrieval practice group. They studied the text for 5 minutes, then sat down and did a free write for 10 minutes about the text but without the text in front of them.  This could help in a few ways.  One, they were practicing the skill they would have to use on the test, of recalling the material without seeing the text.  Second, they could identify gaps in their knowledge and focus on those parts during later sessions. 

The surprising part of the study is how much better the retrieval practice group did compared to the concept mapping group.  Cognitive mapping is a constructivist, progressive task that is very popular.  Retrieval practice is just taking more tests.  I certainly don't think we should give up constructivist learning tasks in favor of more tests based on this article.  However, I do think we could use the results of this study to make learning more effective.  The best way to do that would be to have the students read the text, then make their cognitive map without the text in front of them.  This would force them to retrieve the text will making connections.  Once they have a draft of the map, they can go back and reread the text and adjust their maps accordingly.  The important point is that students remember better when they have to think back and recall a text, not just work with it.  It will be interesting to see what new studies come out based on this one, and where they take us!