Suppose that retrieval testing helps future retention more than concept diagrams or re-reading. I’ll go further and suppose that it’s the stress of trying to recall imperfectly remembered information (for grade, reward, competition, etc. - with some carrot-and-stick stuff going on) that really helps it take root. What conclusions might flow from that?
Coursera-style short quizzes on the 5 minutes of material just covered are useful to check understanding, but do next to nothing for retention.
Homework is useful, but the stress it creates may be only indirectly related to the material we want to retain: lots of homework is solved by meta-guessing, tinkering w/o understanding, etc. What kind of homework would be best to cause us to recall the material systematically under stress?
When watching a live or video lecture, it may be less useful to write detailed notes (in the hope that it’ll help retention), and more useful to wait until the end of the lecture (or even a few hours/days more?) and then write a detailed summary in your own words, trying to make sure all salient points are covered, and explicitly testing yourself on that somehow (e.g. against lecture slides if available).
“The best way to learn is to try to teach”. I thought that when I try to organize something I don’t know very well into a lesson to teach others, I end up knowing the material well because I just have to go over all of it, sift it, rearrange it, anticipate questions etc. But maybe the stressful situation of anticipating being inadequate, failing to transmit the ideas well etc. that is responsible.
Spaced repetition, Anki-style flash cards. Do they work because they present to you the information you’re just starting to forget in just the right moment? Or maybe their success is (also?) due to placing you in the stressful situation of trying to recall in a competition-like setting? Contrast spaced repetition flash cards that just repeat data to you vs. those that ask you to recall with a question/answer format. If the hypothesis is correct, the latter will be much more successful. (It may be said that asking is necessary to determine the time delay till the next questioning in the spaced repetition formula, so comparing the two is difficult. True, but there may be some approximation, e.g. in a deck of 20 cards on the same topic we could play the first 5 cards as questions/answers to approximate how well we remember the whole deck, and then the other 15 cards either as questions/answers or straight answers).
Spaced repetition, Anki-style flash cards. Do they work because they present to you the information you’re just starting to forget in just the right moment? Or maybe their success is (also?) due to placing you in the stressful situation of trying to recall in a competition-like setting? Contrast spaced repetition flash cards that just repeat data to you vs. those that ask you to recall with a question/answer format. If the hypothesis is correct, the latter will be much more successful.
To Really Learn, Quit Studying and Take a Test
Suppose that retrieval testing helps future retention more than concept diagrams or re-reading. I’ll go further and suppose that it’s the stress of trying to recall imperfectly remembered information (for grade, reward, competition, etc. - with some carrot-and-stick stuff going on) that really helps it take root. What conclusions might flow from that?
Coursera-style short quizzes on the 5 minutes of material just covered are useful to check understanding, but do next to nothing for retention.
Homework is useful, but the stress it creates may be only indirectly related to the material we want to retain: lots of homework is solved by meta-guessing, tinkering w/o understanding, etc. What kind of homework would be best to cause us to recall the material systematically under stress?
When watching a live or video lecture, it may be less useful to write detailed notes (in the hope that it’ll help retention), and more useful to wait until the end of the lecture (or even a few hours/days more?) and then write a detailed summary in your own words, trying to make sure all salient points are covered, and explicitly testing yourself on that somehow (e.g. against lecture slides if available).
“The best way to learn is to try to teach”. I thought that when I try to organize something I don’t know very well into a lesson to teach others, I end up knowing the material well because I just have to go over all of it, sift it, rearrange it, anticipate questions etc. But maybe the stressful situation of anticipating being inadequate, failing to transmit the ideas well etc. that is responsible.
Spaced repetition, Anki-style flash cards. Do they work because they present to you the information you’re just starting to forget in just the right moment? Or maybe their success is (also?) due to placing you in the stressful situation of trying to recall in a competition-like setting? Contrast spaced repetition flash cards that just repeat data to you vs. those that ask you to recall with a question/answer format. If the hypothesis is correct, the latter will be much more successful. (It may be said that asking is necessary to determine the time delay till the next questioning in the spaced repetition formula, so comparing the two is difficult. True, but there may be some approximation, e.g. in a deck of 20 cards on the same topic we could play the first 5 cards as questions/answers to approximate how well we remember the whole deck, and then the other 15 cards either as questions/answers or straight answers).
Most of these seem testable!
Thoughts?
Active elicitation and testing does work better than mere exposure; see http://www.gwern.net/Spaced%20repetition#background-testing-works and also search for ‘feedback’.