BONUS: References for some of the things that I mentioned in my comments during the "supercharging learning" workshop.

• "Too rapid feedback can be harmful":

"This research does not mean, however, that greater frequency of feedback is always better. Again, timeliness of the feedback is a significant factor. For example, consider a study in which college students were learning to write mathematical functions in a spreadsheet application (Mathan & Koedinger, 2005). The particular goal for students’ learning in this situation was not only that they be able to write these functions accurately but also that they be able to recognize and fix their own errors. Students who received feedback immediately after they made a mistake scored lower on final assessments compared to students who received “delayed” feedback. Although surprising at first, this result makes sense when one realizes that the immediate feedback group was missing the opportunity to practice recognizing and repairing their own errors. In contrast, the students receiving delayed feedback had a chance to fix their own errors so they had more practice at the corresponding skills. That is, when the delayed feedback group made errors, feedback was given only when they (a) showed sufficient signs of not having recognized their error or (b) made multiple failed attempts at fixing their error. In this way, one could argue that even though it was not immediate, their feedback was given in a more timely manner relative to the learning goals at hand."

(Susan Ambrose et al., How Learning Works: 7 Research-Based Principles for Smart Teaching, citing Mathan & Koedinger: Fostering the Intelligent Novice: Learning From Errors With Metacognitive Tutoring.

See the cited paper for a broader discussion. Note that based on the paper, Mathan & Koedinger would probably somewhat disagree on Ambrose et al's characterization of the worse-performing group as receiving "delayed" feedback: Mathan & Koedinger would argue that the feedback wasn't so much immediate, but rather reflected a different conceptual model of the best kind of feedback. From their description, though, it is true that the worse-performing group did get feedback at an earlier stage than the better-performing one.)

• "Shorter and easier math problems are better than longer and harder ones": don't have conclusive evidence since the following paper discusses a method that's broader than just changing exercises into shorter and more numerous ones, but see regardless pages 4-5 of http://icme12.org/upload/UpFile2/TSG/1801.pdf and the results of their method.

• Bonus bonus: another method for increasing intensity that I don't remember being mentioned is interleaved practice. E.g. if you're practicing movements associated with a physical skill, rather than repeating one move over and over and then doing the same with another move, it's better to increase the challenge by alternating the two, or even picking randomly between the two. May generalize to cognitive skills as well. See http://www.bulletproofmusician.com/why-the-progress-in-the-practice-room-seems-to-disappear-overnight/ .