Slashdot Mirror
UCLA Professor Says Conventional Wisdom on Study Habits Is All Washed Up
An anonymous reader writes "Taking notes during class? Topic-focused study? A consistent learning environment? According to Robert Bjork, director of the UCLA Learning and Forgetting Lab, distinguished professor of psychology, and massively renowned expert on packing things in your brain in a way that keeps them from leaking out, all are three are exactly opposite the best strategies for learning."
For a more general set of suggestions on study skills based on cognitive science, see "How to Get the Most Out of Studying Video Series". This is by Steve Chew, who was recently named a "U.S. Professor of the Year" for his teaching ability. For something printed, but not as detailed, see his "Improving Classroom Performance by Challenging Student Misconceptions About Learning". I recommend the video to all my students (I'm a college economics professor).
When I teach, I prepare a set of lecture notes, mostly an outline with key details. I leave room for notes. Then I give the students a copy of those at the start of the class. That way, they can listen and focus on understanding the information during the lecture. If they need to make some additional notes, they can add to the pre-printed lecture notes I handed out, but since the key points and details are already there, they don't need to add many notes. My experience is that students who spend too much time taking notes don't understand the material and don't remember it, so I make is easy for them to not spend time taking notes.
My classroom time is spent expanding upon the material, having discussions with the students, making sure the students understand it and how to apply it, doing hands on or thought experiments as appropriate, and refining my notes for the next class.
make imaginary.friends COUNT=100 VISIBLE=false
The article says to avoid learning disparate skills like dancing and tennis at the same time.
it says something like "learn several moves of a dance style interleaved... that way you will integrate the different moves and learn the dance faster overall".
She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
TFA is actually a little light on details, so I'll add a few more. But I am not a psychology researcher, so take my explanation with a sack of salt.
There are basically 3 separate issues that they are looking at with respect to learning.
One is spaced repetition. Spaced repetition is the idea that you forget things over time. The longer you go without recalling something, the higher the odds are that you will not be able to remember it when you try. However, if you remember something, the association will be stronger and the speed at which you forget it is lower (i.e., the drop from 90% recall to 80% recall will take more time). That speed is called the "forgetting curve". The shallower the curve, the less often you have to review something to rememer it. Software like Mnemosyne, Anki and SuperMemo try to time when you are likely to forget something so that you don't waste your time reviewing something that you aren't going to forget.
Another concept is spaced learning. This is different from spaced repetition. It turns out that the space between reviews is actually necessary for long term memory. If you memorize something and then wait for a long time, even if you forget it, you will make a stronger connection the *second* time you learn it. Spaced learning intentionally puts spaces between reviews with the intent that it creates a stronger memory (makes the forgetting curve shallower) the *next* time you learn it. In other words, you intentionally make it difficult to remember the second time around (as opposed to choosing a time when you are likely to remember it). The distinction between spaced repetition and spaced learning is fine but important, I think.
Finally there is interleaved learning. It turns out that time is not the only thing that causes you to forget. As you learn new things, the ability to recall old things gets worse. So if you learn A and that's all, you will forget it slower than if you learn A and then B. Learning B makes it difficult to recall A. You can use this to your advantage. Remember that with spaced learning, if you forget something, it is retained *better* the second time around. So if you learn A and B, and then return to A, you will remember A better than if you spent twice as long on A and then did B.
Basically spaced repetition programs that use SM2 algorithm are implementing spaced repetion. I will argue that they aren't making use of spaced learning, at least intentionally. When you initially try to remember something, you should space repetitions so that it is difficult to remember the item. One of the weaknesses of SM2 is that it doesn't really have any strategy for first learning the item (on the other hand, you are free to adopt your own strategies within the framework of the software). Specifically, there's no concept of getting an item correct and then waiting a short time and reviewing it again. It goes ahead and schedules it for a day or so later. Also, when you get an item wrong, you are back to square one, with the "difficulty" set at the same level it was at before you got it wrong. Spaced learning would suggest that at least the item will get less difficult every time you forget it. So I think there is considerable room for improvement.
SM2 also specifically does not implement interleaving. When learning new material (or even items that you forgot in the review) it would be rather interesting to have it introduce one new fact from 4 or 5 different quizes at a time. It would accellerate the speed at which you forget the item and provide opportunities for spaced learning faster (presuming there was support for spaced learning).
I'm actually the author of another spaced repetition program for studying Japanese, called JLDrill. I use a different algorithm, which I describe here: http://jldrill.rubyforge.org/Strategy.html I'm going to try to implement some of these other ideas in the near future.