Nukemarine Wrote:I do like the idea of SRS being incorporated into individual learning plans or self paced classes.

Sort of on topic, but here's how I describe SRS to people:

Imagine you're in College and you walk into a new class. The professor says: "Good morning, this class you will be roughly seven days a week. Tests can be anywhere from 100 to 200 questions or more. If you miss a day, those questions will be added to the following day's test. The questions will cover any and all material previously studied". Now, hearing this, will you drop out of that class?

I really would like to make a SRS test making program. The school I work at, and most others, just don't have the resources to do SRS by computer in school, but there's no reason you couldn't do SRS by paper testing.

The problem is, of course, that there's no time to make an individual test for every student. It should be quite easy, however, to write a computer program that makes many tests and print them all out out in one swoop. My idea is to do it with something like vocabulary, and the computer could keep track of which ones the students have failed just like a computer SRS.

There are various problems which make it difficult. One is that it's good for the students to be given the right answer directly after they answer a question, and this seems to be impossible without some ingenuity. It would be possible, however, to do so directly after a test by having the computer print out the answer sheets as well. You could give this to the students, and then give them time to look it over alongside the questions (it would be difficult to actually get the students to pay attention to this stage). An alternative way would be to give them a list of what they failed that day, as well as recently failed answers, before the school day ends.

There are other problems: The amount of paper used, grading the tests (whether to do it, and how to do it), and the fact that the students don't come to school every day (you could always just have a bigger test on their first day back after the weekend, but holidays are even more difficult).

Probably, by the time I could work out all these and other problems, computers will be playing a far bigger roles in schools. Then again, maybe not--it's been quite slow so far.

Tzadeck Wrote:There are various problems which make it difficult. One is that it's good for the students to be given the right answer directly after they answer a question, and this seems to be impossible without some ingenuity. It would be possible, however, to do so directly after a test by having the computer print out the answer sheets as well. You could give this to the students, and then give them time to look it over alongside the questions (it would be difficult to actually get the students to pay attention to this stage). An alternative way would be to give them a list of what they failed that day, as well as recently failed answers, before the school day ends.

i find that knowing what i got right is just as important (if not more) than knowing wht i got wrong. if not for correcting myself for motivation alone. it would be quite desmotivating getting a list every day that reads "this is what you got wrong." and nothing saying "but this over here you got right."

perhaps if you take off the pressure, make it un-graded practice, and get the students to understand what they are doing, you could give them question/answer pairs (in columns, i guess), and tell them not to look at the answers untill they've answered the question. getting all of the pass/fail info into the computer every day is gonna be hell though.

I won't go onto the topic of extensive reading since I'm assuming we're all high school students here and we don't actually care about learning, instead we care about good marks.

The best way to memorise things in maths is to do more questions. For example a topic which required heavy memorisation in mathematics was circle geometry.

Reconstructing proofs in circle geometry was a pain in the neck. But, it is made much easier when you have a look at (or attempt to) derive the axioms yourself - which is what our teacher made us do. After I did the 30 exercise questions on circle geometry knowing the rules wasn't a problem, it was actually doing difficult questions.

I can say the same thing about any other mathematical axiom or theory. I really don't see the point in memorisation derivation tables when you should be able to construct them yourself from scratch. Not once did I "memorise" derivation or integration tables, even though they had them at the back of our sheet for reference, they came at the top of my finger tips because I had done questions which were the actual fun part for me.

Chemistry - This subject was slightly more text based than another country would have their chemistry courses but the same thing applies, all the facts learnt in chemistry made sense after you thoroughly looked through and understood them. All one really needed to do was memorise one fundamental reaciton pattern, say (off the top of my head, has been 5 years since I've done chemistry), - Chemical + 02 -> combustion + other shit.
Wala and after you do exercises, read more on the topic that equation gets buried in your mind. I clearly don't remember it now, but when I pick up a text book it will spring back to mind.

Physics - Same thing. Read up on Feynmann, learn how to derive Maxwells laws from newtons laws (lawl) and there you have it. Every physics professor I've seen pulls out equations from their finger tips, how do they do that? Not through SRS, but through understanding and doing many sorts of questions. The time wasted SRS'ing formulas/unit conversions or whatever in ones head could have been better spent working out questions, understanding physical processes getting an intuitive feel for the formula, and intuitive feel for the equation.

English - Ok skip.
History - I kinda understand why one would SRS here,
Languages, Sentences, words, whatever, understandable

I really love SRS's don't get me wrong, but if someone is going to spend their time memorising equations when they can do the real stuff which is what is actually being asked in an exam/real life situation then I think it's better to just do questions.

liosama Wrote:I won't go onto the topic of extensive reading since I'm assuming we're all high school students here and we don't actually care about learning, instead we care about good marks.

I agree most high school students (hell, even most college students) don't care about learning at all, can't assume everyone's like that though.

liosama Wrote:Reconstructing proofs in circle geometry was a pain in the neck. But, it is made much easier when you have a look at (or attempt to) derive the axioms yourself - which is what our teacher made us do.

uh.. i really do hope you mean theorems and not axioms... because... it makes no sense to "derive" axioms, given that axioms are the assumed base for deduction, and thus require no proof. and yes, you can usually deduce theorems, but i find, for myself, that memorizing certain things before starting to solve problems, makes the problem solving a faster and more efficient learning tool.

liosama Wrote:Physics - Same thing. Read up on Feynmann, learn how to derive Maxwells laws from newtons laws (lawl) and there you have it. Every physics professor I've seen pulls out equations from their finger tips, how do they do that? Not through SRS, but through understanding and doing many sorts of questions.

yes, those teachers have them formulae at their fingertips, then again, they've been teaching that same course, what, two times a year for a couple of years at least? i would think the reason for them remembering the formulae is because they repetitively recall it, all the time. Think of it as "spaced repetition" with a few extra reviews in the middle.

again, i doubt anyone here is advocating for mere SRSing without actual practice... srs is a tool to memorize, understanding comes only with analysis and practice. memorization just makes the understanding process go faster.

liosama Wrote:I won't go onto the topic of extensive reading since I'm assuming we're all high school students here and we don't actually care about learning, instead we care about good marks.

I would assume that anybody doing Heisig probably cares about learning more than good marks. Also, really? I thought most people here were in their 20s/30s...

dbh2ppa Wrote:

liosama Wrote:I won't go onto the topic of extensive reading since I'm assuming we're all high school students here and we don't actually care about learning, instead we care about good marks.

I agree most high school students (hell, even most college students) don't care about learning at all, can't assume everyone's like that though.

welldone101 Wrote:I would assume that anybody doing Heisig probably cares about learning more than good marks. wink Also, really? I thought most people here were in their 20s/30s...

Extensive reading kills all forms of anything to do with SRS anyway, since it involves reading widely and deeply into a topic in order to become a savant. I was merely responding to the ones saying SRS would have helped back in high school. SRS's and Extensive reading are not synonomous, in fact they're both the exact opposite of each other.

dbh2ppa Wrote:

liosama Wrote:Reconstructing proofs in circle geometry was a pain in the neck. But, it is made much easier when you have a look at (or attempt to) derive the axioms yourself - which is what our teacher made us do.

uh.. i really do hope you mean theorems and not axioms... because... it makes no sense to "derive" axioms, given that axioms are the assumed base for deduction, and thus require no proof. and yes, you can usually deduce theorems, but i find, for myself, that memorizing certain things before starting to solve problems, makes the problem solving a faster and more efficient learning tool.

Yes sorry I meant deduce theorems and memorise axioms,

dbh2ppa Wrote:

liosama Wrote:Physics - Same thing. Read up on Feynmann, learn how to derive Maxwells laws from newtons laws (lawl) and there you have it. Every physics professor I've seen pulls out equations from their finger tips, how do they do that? Not through SRS, but through understanding and doing many sorts of questions.

yes, those teachers have them formulae at their fingertips, then again, they've been teaching that same course, what, two times a year for a couple of years at least? i would think the reason for them remembering the formulae is because they repetitively recall it, all the time. Think of it as "spaced repetition" with a few extra reviews in the middle.

again, i doubt anyone here is advocating for mere SRSing without actual practice... srs is a tool to memorize, understanding comes only with analysis and practice. memorization just makes the understanding process go faster.

That's true, but how come when I took my electromagnetism class in 2nd year and in 3rd year I could write all of Maxwells and other messy electromangetic equations off the top of my finger? This was by using David J Griffiths Introduciton to electrodynamics (perhaps the most popular undergraduate electromagnetism book used across the world) With no SRS, but actually reading the chapters and going through their derivation. Same thing with quantum physics in 2nd year, the equations were very messy but the Shrodinger equations were much easier to understand.

Statistics popped up just know. Everyone here who has taken at least one course on stats would know that stats has more formulas than any other course. For most of the basic distributions I'd go through the derivations at least once. There's no way I could re-derive them but it gave me a better understanding of them. However there are still way too many equations that one has to use, at my university we were given formula sheets, I'm pretty sure other universities follow the same foot. Again here I think it is more important if one knows how to use the formulas instead of actually memorising them. Memorisation could help save time grabbing a formula sheet, I'm still kind of iffy on that one.

Someone could implement an SRS here, but I would probably kill myself out of boredom trying to run these through.

The thing is with mathematics everything has patterns, there are patterns in very complex equations, that's what I've spend almost 3 years in my Engineering degree, learning how to exploit these patterns and use as little memorisation as possible and more of understanding. I agree that memorisation makes things faster, much faster , but I believe it's much more fun to have this memorisation process coincide with the understanding one. I think SRS's remove that link.

Well SRS can probably help with highschool but its so pathetically easy that there's no need. That's why I keep trying to bring the conversation back to university, where a leg up in learning is really useful! I know for a fact that I passed several of my units through sheer guesswork and memory at the exams since the questions get repeated (lazy engineering lecturers). The SRS would've helped here of course, but that's not as important as knowing if using an SRS to learn the fundamentals would help solve problems later.

Could somebody explain to me how putting problems into Anki to help study math is fundamentally different from putting sentences into Anki to help study language? Why is the consensus that the memorization problem will work itself out in the latter case and not the former?

Mcjon01 Wrote:Could somebody explain to me how putting problems into Anki to help study math is fundamentally different from putting sentences into Anki to help study language? Why is the consensus that the memorization problem will work itself out in the latter case and not the former?

Fundamentally they are different, but I don't think that's why the consensus is occurring. I think it's occurring because we've had pioneers who have proved that they can learn languages via SRS, and we don't yet have a champion for the SRS math movement. One who took the plunge and went all the way through. I'm pretty sure SRS for languages seemed equally silly when the first person started out on it. Don't you still get guffawed at by anybody that doesn't know better?

Stop reading this post.
Fundamentally different (note: math and languages are fundamentally different, that doesn't mean that the SRS can't be applied to them in the same fundamental way, I'm just debating semantics here): I believe they are fundamentally different. This is an a philosophical debate that has been going on for.. a while. I believe one is created by humans (math) and one is born, not created, of humanity (language). Also the fundamental principles of math are derived from the universe and it's a system we are still inventing to describe it's actions.

No SRS were just an upgrade to flashcards. Something inevitable in this age of computers and portable devices. No one would ever use physical flashcards to learn mathematics, so I don't see why anyone would use SRS to learn mathematics.

I don't think SRS deserve their own category of learning paradigms as everyone makes SRS out to be. I see them as a convenient replacement for flashcards. If one was taught the exact algorithm that SRS programs use, you use that and chuck it on your physical flashcards, but no one is going to run around carrying 2000 cardboard papers, instead why not just make a little program and chuck it on portable devices which most people carry around these days.

Sure I would be impressed to see someone demonstrate their math skill through an SRS but as it stands now I highly doubt I'll ever see such a thing happening

Just out of curiosity how many here have actually studied further than first year mathematics/physics at university and actually agree with the OP et al?

Does Engineering count? =(

Nii87 Wrote:Does Engineering count? =(

Yes I'm a UNSW Engineer, nice to meet you Aussie

liosama Wrote:Just out of curiosity how many here have actually studied further than first year mathematics/physics at university and actually agree with the OP et al?

Do I count? I've got a degree in mathematics focusing on operations research and complex analysis. I know of course I agree with me, but I'm not exactly sure what I am agreeing with me about. I was just wondering who else saw the Star Trek movie and thought, "OMG the Vulcan's are using an SRS, hah!"

I think the reason that an SRS allows more language learning than a flashcard, is that since it's electronic and efficiently managed you're suddenly able to have 10,000 flashcards. You learn the language through exposure, just.. efficient, memorized exposure.

Is the same possible in math? I believe for starters you can cut out the memorizing part from the classroom and put it into an SRS and be faster about it. But there's really very little to memorize about math (say compared to chemistry or bio or languages) once you start really getting into it. However, for people who aren't interested in math, and just want to get through grades 1-12 and year one college, maybe it'd be awesome.

Maybe it's possible to do more and actually help teach math by using electronic flashcards to give lots of exposure to certain problem archetypes? Don't know, we don't have a http://www.allmathallthetime.com guinea pig.

Mcjon01 Wrote:Could somebody explain to me how putting problems into Anki to help study math is fundamentally different from putting sentences into Anki to help study language? Why is the consensus that the memorization problem will work itself out in the latter case and not the former?

Because solving a problem and remembering a fact are two different things.

Studying language requires the memorization of large amounts of knowledge. Most notably vocabulary. Anki and software like it help you to retain knowledge of vocabulary over the long term. In mathematics, knowledge is important and for that Anki could be helpful. However, most people have trouble with mathematics, not because they can’t remember something, but because they don’t understand. Anki won’t help you to understand things, it just helps you not to forget things.

The R in SRS is Repetition not Remembrance in my opinion. The spacing we use can be applied to more than just flashcards. One can use it to space out music, dialogue, hobbies, chords of music, whatever. Like I posted earlier, you can even apply it to a process such as mathematical equations or grammar conjugations.

What turns SRS into Spaced Remembrance System is when we add a "Reset" button. You don't have to reset anything, which may turn it into more of a Spaced Reminder System (ok, how many different R's can I put into that acronym?).

An SRS is great to ingrain fundamental concepts, facts or procedures into your head. The spacing allows new stuff to come into play without the expense of ignoring older material. Ask me, that makes it perfect for many, many aspects of scholarly work. You just have to decide what fundamentally needs to be added. Then you want to break into smaller chunks in some manner or fashion.

The best way to get information in the brain is to understand it in a meaningful way. That is the entire point of RTK. It's primarily the mnemonics and understanding of primitives that etches the kanji into our brains, followed by the continous use of said kanji(reading/writing) that keeps it there . The SRS is a just a mechanism to temporarily trick the forgetting curve (or whatever it's called). In the case of math it's less necessary because there aren't so many facts to be memorised nor is there as much benefit in memorizing them. I also agree with others who said that formulas are best remembered by doing the proofs. Similarly I think it's also not necessary for programming languages (another thread) because functions/etc can just be looked up as needed. I think there could be some use for srs in maths study though. In school, we learned by repeatedy doing the same kinds of questions over and over and over again. Obviously this isn't necessarily the most efficient way to go about it, and it slows down forward progression. An srs program could be used to test students on different types of questions (rather than teaching formulas). If the student can consistently answer questions of a certain form, the srs progresses to more difficult questions. The srs program would have to support variables in the question side, so the answer on the answer side is also variable. Otherwise the student could just memorise the answer. I don't know if any current SRS programs can do this or not, but I don't think it would be hard to implement. Questions using variables might be useful for language study too. For example, you could write sentences that draw from a random list of names/pronouns or nouns so that the question/answer pair can't be memorised but rather needs to be understood. eg:
question: [random name] は [random food] を食べた。
answer: [random name] ate [random food]

I think it's important to realise the SRS just does whatever you make it do. It can be used to remind you about things you already know. It can be used as a quick and dirty way to memorise new facts (less efficient than mnemonics imo). Or it can be used to test understanding. The latter obviously requires more thought in setting up, and I prefer to just use books/listening for this purpose.
The SRS is a powerful tool that can be used to help in the learning of probably just about anything, but the learner needs to be aware of exactly what they are using it for. I was slightly surprised at the rather anti SRS responses in some of the recent threads such as the learning music topic posted by me.
After all I don't think anyone here is solely using an SRS or even using it as their primary method.

wrightak Wrote:

Mcjon01 Wrote:Could somebody explain to me how putting problems into Anki to help study math is fundamentally different from putting sentences into Anki to help study language? Why is the consensus that the memorization problem will work itself out in the latter case and not the former?

Because solving a problem and remembering a fact are two different things.

Studying language requires the memorization of large amounts of knowledge. Most notably vocabulary. Anki and software like it help you to retain knowledge of vocabulary over the long term. In mathematics, knowledge is important and for that Anki could be helpful. However, most people have trouble with mathematics, not because they can’t remember something, but because they don’t understand. Anki won’t help you to understand things, it just helps you not to forget things.

I'm well aware of what an SRS is used for. It's why I chose to write "help study" rather than "learn". I do most of my actual learning by reading (and watching the occasional video tutor), but I still have the problem of forgetting things I don't review -- and the more I learn, the more there is to review. My math-learning project is on the back burners right now, since I'm getting ready to maybe-study in Japan this fall and need to crack down on my Japanese, but the way I envision it is having the "question" be a problem related to whatever concept I've just learned. I'm sure there's a better way to do it, but this seems simple and a good starting point for my grand math experiment. The "answer", then, wouldn't actually be the answer to the problem, but rather if I understand the problem, how to solve it, and all the "whys" that go into that. There wouldn't even really need to be anything actually on the answer side of the card, since the flashcard aspect of the SRS is incidental in this case. It's really just a glorified scheduling system saying "Hey, time to contemplate this old material so your brain knows it's still important." Though, I suppose having the answer there could help check for accuracy.

Of course, this is all just theory bouncing around in my head right now, so it may turn out that the SRS is useless for math study, or that even trying to use it will actively destroy your ability to count. Who knows? I still intend to try. Hopefully someday I have something interesting to report back about it. Don't expect an allmathallthetime, though. Blogging isn't really my thing.

I can't offer any practical experience in using an SRS to learn math, since I haven't tried it myself, but as long as we're talking theory, Dr Wozniak (the author of SuperMemo) says math is not an exception. To summarize: mathematics is a field where facts are less important than inference, thus SRS is not as useful as other fields (like language and medicine) that require memorizing more facts. However SRS still provides a benefit where:

a) Facts need to be memorized.
b) Memorization is more time-efficient than repeated derivation.
c) Memorized problems illustrate underlying rules and patterns.

Personally, I don't see a theoretical difference between SRSing math problems and SRSing example sentences. In both cases the goal is not rote memorization, but spaced repetition of problem solving. If you learn 10k sentences by route, you can only recognize and generate those 10k sentences. Does that mean that SRSing sentences is a waste of time? Not if you discover and practice applying the underlying patterns of a language in doing so.

It should be the same for math problems.

However, it's also worth noting that those who have learned languages using an SRS did not use an SRS only. Understanding a language and solving math problems are both learned by doing, with spaced review acting as a powerful supplement.