Random problems from my problem sets: a suggestion
You must be logged in to post a reply. If you don't have an account, you can register.
|Plumflower||I find it difficult to access older problems within my saved problem sets. It would be nice to have an option for random problems. I seem to continually only be able to access a cascade of about 30-40 problems this is a nice feature for learning new content. But it seems difficult to access old problems outside of going in and manually selecting a certain set.
I would love the option of including more randomization in my problem sets. As I never see some of the earlier problems which I have saved. Thanks for the work you guys do!
|Darrell Malick||Hi Plumflower,
I've considered two extremes as an answer to your suggestion: 1) good idea. Or 2) a very long explanation of the reasons for why the scheduling works the way it does. I'm going to attempt an in-between solution here, but will expand if you want to hear more.
The scheduling in the Training System is designed to give you the *smallest* amount of repetition and study time necessary for you to remember what you've learned. How you rate problems affects this very profoundly. If you are seeing the same problems too often, then you should be clicking "Good" or "Easy" more often. (You could also adjust your settings to be more aggressive - meaning less frequent repetition)
If you want to review some problems *before* they come up in the scheduling rotation, then go to the Lesson Page for that lecture and click the "Quick Look" button and just look at those problems.
Looking at random problems from your problem sets is not likely to be particularly efficient in either learning or refreshing memory, but it might be entertaining.
|Darrell Malick||Hi again Plumflower,
I just took a look at your account and see you have a large backlog of problems that are due for review. This is why you are only seeing your most recent studies. The ones you are wishing to see more of are deeper in the queue.
There is a reason for this: A problem that you studied recently will have a short review interval. Let's say for example, 10 days. If you are 2 days backlogged, then you could say you are "20% past due" on that problem.
Now take a problem that you studied long ago and the review interval is 200 days. Imagine you are 20 days past due on that problem. Well that would only be "10% past due."
The system will give you the "20% past due" before the "10% past due" because you are in more danger of forgetting the 20% than the 10%.
This is based on the theory of a Spaced Repetition System: Each time you re-study something - especially near the time you would have forgotten it - you forget it more slowly next time around.
The solution to your problem is - if possible - catch up on your past-due reviews. If it's not too boring, I'd suggest not adding new problems until you make some headway on the overdue ones.
Also, if there are problems you just don't care to study enough remember, turn those problem sets off so that you get more time on the ones you DO care about. For me, these are problem sets that were too easy or too hard. Turn them back off. Don't leave them there to waste your time or to discourage you.
Hope this helps.
|Plumflower||Right. Thanks. I do enjoy the Spaced Repetition System and i agree that its a great way to learn new content, yet i am unsure of how to adjust the settings correctly so that i see more old content occasionally. I set the starting easiness factor to 1.7 in an attempt to use the 'easy' selection to not show certain problems for longer periods.
I suppose it may just be a personal preference for me to want to see random problems occasionally. My workaround i suppose is to simply go back and do an individual problem set. Its just unfortunate that i am not 'surprised' by problems i may have forgotten about. I enjoy being suprised because the problems have left short and even long-term memory which forces me to utilize principles to solve, again, the problems. This is a different form of learning that utilizes different algorithm-like processes to solve problems instead of memorization.
The thing is, Spaced Repetition is still a precursor to utilizing principles to solve problems instead of forced memorization of those forgotten problems. I prefer not to memorize 'harder', but to instead learn how to 'solve' problems by forming (mostly subconscious) principles.
(Neuroscience is my major)
Post a reply