It turned out that the students in the last group retained more information -- 50 % more, evidently -- than those in the second and third groups (from the Times' report it seems the students in the first group, who read the passage once for five minutes, must have served as something like a no-recall control group). A second experiment, comparing just concept-mapping and testing, also found that the tested students "did much better."
These are very interesting experiments, but for now I want to make just one observation: What these experiments appear to show is that the way to increase recall is to practice recalling. Put that way, these experiments really shouldn't be surprising. If they are -- and apparently they are -- that's a measure of how confused our thinking about learning has become.