FAQ: Formulating knowledge in learning |
|
Do you have a question about writing simple and effective collections for SuperMemo? Write to Dr Wozniak. If you would like to have your items evaluated, please send 3-5 for review. Do not send whole collections in an e-mail attachment |
You can learn spelling with SuperMemo
You can learn to recognize diseases by their symptoms
You can use occlusion test to learn anatomy
Keep your items as simple as possible
The best way for providing context is to use labels inside your questions
Developer's Guide for authors of SuperMemo collections
You can memorize long sequences (e.g. Tai Chi)
Keep only one deletion in cloze deletions
Strategic recall is a matter of knowledge structure
You can learn programming with SuperMemo
It is recommended that you keep all your knowledge in one collection
Only items are used in learning (topics are only for material review)
Copying material from a dictionary
Combining multiple pieces of information in one item is not
recommended
Does minimum information principle hold in all cases?
You do not need to interrupt repetitions to edit items
Why do we remember weird words?
You can learn to retrieve words faster from memory
Do not create follow-up questions in your items. Use cloze deletion
Language ambiguities can be tackled with SuperMemo
If sequence is important, it must be included in a single item
Math proofs in SuperMemo
Use mnemonic techniques to memorize difficult material
(e.g. numbers, lists, sets, etc.)
Question-answer swapping will usually involve some reediting
Incremental reading requires some experience
Syllables may be easier to remember than numbers
Some narratives may be thin on memorizable material
Sequences are best learned through logical consequence
"Successful" leeches should not be reset
Inventiveness is the key to mnemonic techniques
You can learn Arabic with SuperMemo
How can SuperMemo help an actor?
You can learn programming language syntax
See also:
(Mike
Condron, USA, Dec 10, 2000)
Question:
I am finding that anatomy is a subject that consists largely of lists. The
branches of the axillary artery are... the layers of the abdominal fascia are
... the contents of the posterior triangle of the neck are ... and so on. What
is the best way to learn such lists?
Answer:
Enumerations should be discouraged. These always result in recall
problems. You should rather use multiple graphic deletion tests. You can, for
example, use a picture of the branches of the axillary artery, occlude or point to a
single branch in the picture and ask: Which branch is pointed to in the
picture? You will have as many items as there are branches of the artery.
See also:
(John
Meritt, UK, Nov 26, 1999)
Question:
How can I best learn programming with SuperMemo?
Answer:
(John
Meritt, UK, Nov 26, 1999)
Question:
How can I best learn spelling with SuperMemo?
Answer:
You should create a template
in which the answer will be a Spell-Pad (i.e. text input component). In the
question part you should ask about the word that is to be spelled. Because many
spelling problems come from the use of double letters (e.g. traveling vs.
travelling), you do not need to define the word. It is enough you ask to
choose a correct variant. For example, your question might look like this: traveling/travelling
or better yet trave(l/ll)ing.
It is very important to focus on one problem at a time. This is why instead of o(c/cc)a(s/ss)ion,
you should create two items:
Occasionally, you can make an exception to this rule. For example, you might ask Mi(s/ss)i(s/ss)i(p/pp)i as a request to spell Mississippi. In this case, it is easy to remember that all questionable consonants in this word must appear in double. Once you realize that, you may never experience problems with recalling how to spell Mississippi
(Luis
Gustavo da Silva, Brazil, Nov 9, 1999)
Question:
What is your view on the formulation of items in which the
question lists the symptoms of a disease and the answer provides the name of the disease?
Answer:
This sound like knowledge that should not cause much trouble in
learning:
(Jim
Ivy, USA, June 4, 1997)
Question:
What is the difference between a topic and an item?
Answer:
Topics are used to store articles or other content that presents knowledge, while items are used to
test knowledge by means of repetitions (usually they have the
question-and-answer structure or are cloze
deletions). See: Topics
vs. items
You can learn programming language syntax Q: mysql: To rename a table named table_name to
new_table_name, use: Q: mysql: To rename a table named table_name to
new_table_name, use: Q: mysql: To rename a table named table_name to
new_table_name, use: Q: mysql: To rename a table named table_name to
new_table_name, use: Q: mysql: To rename a table named table_name to
new_table_name, use:
(Noah Chanin, Friday, March 11, 2005 1:33 AM)
Question:
I want to memorize all of the valid forms of SQL. If you have some ideas about effective techniques to employ for learning syntax, it would be appreciated
Answer:
Before you begin, you should be aware of the cost-benefit balance. In most cases, syntax requires no SuperMemo. Simply put, your daily or regular use of the language will ensure this knowledge is easily retained without repetition.
However, if you are only a beginner, or you plan to know the language without using it much on a regular basis,
incremental reading is always the best option. You can import your entire programming language manual, and process it in proportion to priority of individual subjects.
In the end, you will arrive at simple cloze deletions as quoted below. It is worth remembering though, that there is not much benefit in memorizing details of syntax until you realize at some point that not remembering a particular piece of information is a stumbling block in your further progress. Knowledge of a single programming language is usually vast enough to take years to master to the last detail, even with the employment of the best techniques of incremental reading.
This is why a good principle is: "memorize only in need" and process the rest passively to reduce time costs.
Here are some examples of clozes to which you arrive in the end:
Q: mysql> [...] TABLE
new_table_name TO table_name;
A: RENAME
Q: mysql> RENAME [...]
new_table_name TO table_name;
A: TABLE
Q: mysql> RENAME TABLE [...]
TO table_name;
A: new_table_name
Q: mysql> RENAME TABLE new_table_name [...]
table_name;
A: TO
Q: mysql> RENAME TABLE new_table_name TO [...];
A: table_name
(Noel Clary, USA, Sep 6, 1998)
Item reviewed:
Q: a rod-and-tube element temperature sensor consists of:
A: a high expansion metal tube containing a low expansion rod. The rod & tube are attached on one end. The tube changes length with changes in temperature, causing the free end of the rod to move
Suggestions:
This is a typical case of combining a number of items in one with detriment to the ability to recall the combined item. The suggestion here is to split the item into a number of simpler items that reproduce the same information in student's memory:
Q: What are the two parts of a rod-and-tube temperature sensor?
A: rod and tubeQ: What is the expandability of the tube in rod-and-tube sensor?
A: highQ: What is the expandability of the rod in rod-and-tube sensor?
A: lowQ: How is temperature indicated in the rod-and-tube sensor?
A: tube moves relative to the rodQ: Where are rod and tube connected?
etc. etc.
(Noel Clary, USA, Sep 6, 1998)
Item reviewed:
Q: Step 5 HVAC duct design: Size ducts by the selected design method. Calculate system ..., then select ...
Suggestions:
Cloze deletions (i.e. questions with blanks) are generally a good learning tool; however, in most cases it is better to keep a single deletion per question
(Deron
Isaac, USA, May 21, 1997)
Question:
How can I edit texts of items during repetitions without
backing out of the test mode?
Answer:
Choose Q to edit the question, A to edit the answer, or E to
edit all text components. You can also edit all properties
of all components by using the component menu available with
the right
button click on a component in question
(John
Gibney, Australia, Sep 16, 1998)
Question:
In your materials you write that users should avoid memorizing sets (e.g. countries of
Europe) or long sequences (e.g. the alphabet). What if I want to remember the sequence of
a form of Tai Chi?
Answer:
Let us consider an example in which you want to memorize the entire sequence of letters in
the alphabet. It won't be very effective if you use the following item:
Q: What is the sequence of letters in the alphabet?
A: A,B,C,D,E,F,G,H,I,J,K,L,M,N,O,P,Q,R,S,T,U,V,W,X,Y,Z
You will notice that you frequently stumble on parts of the sequence and need to stop repetitions just to exercise the entire sequence in the traditional way (like we all learn poems by rote).
However, you can approach this in a way that guarantees quick effects:
Q: What is the sequence of letters in the alphabet
between A and E?
A: A,B,C,D,E
Q: What is the sequence of letters in the alphabet
between D and H?
A: D,E,F,G,H
Q: What is the sequence of letters in the alphabet
between G and K?
A: G,H,I,J,K
etc. etc.
After 2-3 weeks of repetitions, you may take on an extra task of recalling the whole sequence after each repetition of these simplified items. This will make sure you can recite the entire alphabet quickly. You will also frequently rehearse that parts of the sequence that are harder for your memory (e.g. V,W,X,Y,Z) as opposed to those that are much simpler (e.g. A,B,C,D,E)
Question:
What is the purpose of the option Search : Categories?
Answer:
SuperMemo introduces the concept of a category to help you
keep items related to different subjects in different branches of the knowledge tree (contents window).
It also makes it possible to give different item groups different
appearance (e.g. size of text components, screen layout,
font, color, and many more). Read more: Using categories
(Anatolyi
Lipatov, Ukraine, Jul 12, 1998)
Question:
I am using Advanced English to enhance my English and
business English. Now I am registering
for CFA examination (that is Chartered Financial Analyst program of Association of
Investment Management and Research). There are several organizations developing and
distributing methodological stuff for preparing to the exam. A lot of things should be
memorized for passing the exam. What do you think the best way to fit SuperMemo for
memorization and what approach should I use to prepare my own knowledge base for
memorizing the material. Is special programming knowledge needed for it?
Answer:
No specialist knowledge is needed to prepare simple knowledge systems in SuperMemo.
With Alt+A (i.e. Add new) you get
90% of functionality! All advanced editing options can be worked around by an appropriate
questions-and-answer approach. Perhaps it would be useful yet to learn how to add images
to your items (see help for details). To learn more about effectively structuring
knowledge in SuperMemo you might want to read 20 rules of formulating
knowledge in learning and Knowledge
Structuring and Representation; however, nothing works better as learning on one's own
mistakes in formulating knowledge for learning with SuperMemo
(David
Mckenzie, New Zealand, Apr 8, 1998)
Question:
Is there any point in keeping collections
separate?
Answer:
No. Once you master contents categories
and templates, there is no point. You gain global search,
global registries, global repetitions, global optimization, etc.
This would not be advisable back in SuperMemo 7 as item difficulty measure (E-factor) was
dependent on the average difficulty of items in the collection.
Presently, the item difficulty measure (A-Factor, or absolute
difficulty factor) is absolute and does not depend on the context in which an item is
placed (see: SuperMemo Algorithm). Only the length of the
first interval will significantly be affected by the average difficulty of items in the
collection. However, this shall not bear dramatically on the speed of learning. Especially that variable forgetting index for individual items makes it possible
to set different first intervals for whole contents categories
or branches of the knowledge
tree
Question:
What should I use Duplicate for? What
for do I need the same item in the same collection?
Answer:
You can duplicate an item, if you want to add to the another item
which is only slightly different. This way you can spare some time by reediting the old
item instead of typing in the new one
(Marc
de Ruiter, China, Apr 5,
2000)
Question:
It is possible to have SuperMemo do follow-up questions like this:
I am learning Chinese Characters, so first I have a
character which is the Question, then I have to give the answer which is the
pronunciation of that character, but then I have to give the meaning in English.
So basically there are three things: character - pinyin - translation
Answer:
This is a case where you would like to learn three associated things
A-B-C, where A is your question, B is the follow up and C is the follow up to B.
It is never a good idea to learn more than one thing in a single item! SuperMemo
needs to separately understand your difficulties with linking A and B, B and C,
as well as A and C. Building a test according to your suggestions is possible,
yet it would be better to use cloze deletions here. For example:
Parent template item:
A B C
Items generated with cloze deletion (e.g. using Reading : Remember cloze):
Item 1:
Q: ... B C
A: AItem 2:
Q: A ... C
A: BItem 3:
Q: A B ...
A: C
It is important to know that Item 1 above may make you fail to answer with A to the question C if you only learn to answer Item 1 by understanding the association of B with A. In such cases, you will need even more work by formulating items: A-B (where A is the question and B is the answer), A-C, B-A, B-C, C-A, and C-B. Although you will get six items instead of one, you knowledge is likely to be more solid and you may actually spend less time on repetitions of those multiple items than on repetitions of the conglomerate A-B-C item
(Patrick
Mon, Jan
14, 2001)
Question:
Can SuperMemo be used to memorize a hundred digits of PI? What about memorizing
very complex formulas or foreign language characters? The 20
rules page did not include those things
Answer:
Memorizing PI falls into the domain of mnemonic techniques. You could try
to use SuperMemo to memorize PI as is but it would be tremendously
inefficient. On the other hand, if you use mnemonic techniques to memorize
things, you should use SuperMemo to handle the periodic review. Despite the
claims of many mnemonists, mnemonic techniques will not you remember for ever.
In other words, to memorize PI, use both mnemonic techniques and SuperMemo. For
memorizing numbers, the best technique is called the peg-list method. A peg-list
is a list of 10 or 100 pictures associated with numbers 0..9 or 0..99
respectively. Memorizing PI is then equivalent to building a list of pictures
that correspond with the number. The rule of the thumb in reference to mnemonic
techniques is: use them on material that is particularly difficult to remember
(e.g. numbers, sets, lists, etc.); do not use them on material that stick well
to your memory in SuperMemo. To learn more about mnemonics, see some of the
links below:
Question-answer swapping will usually involve some reediting
(Patrik Nilsson, Thursday, July 26, 2001 5:41 AM)
Question:
SuperMemo website claims that swapping questions with answers is rarely applicable beyond word-pair learning. I disagree. If you show a picture of a car for a little child and ask for what's at the
picture, next time you ask the child what a car looks like and want him to draw it or explain it by words
Answer:
You are right about swapping the roles of stimulus and response. However, in
practice, optimum knowledge representation will enforce a degree of editing. In your presented example, we would rarely use name-picture pairs (unless the whole collection is uniform and leaves no doubt as to the question format). Optimally you will need a verbal clue:
What object is presented in the picture? and then How does a car look like? Those minor editing enhancement substantially speed up the learning process
The best way for providing context is to use labels inside your questions
(Steve Brown, Tuesday, August 21, 2001 4:01 PM)
Question:
Is there a way to tell what category a Q/A pair belongs to. The same question may have a different answer. For example,
"what is a strike" could have 2 different answers depending whether I was in the
"bowling" category or the "union" category
Answer:
Instead of using categories for such a purpose, you could accomplish the same with domain labels inserted before the question. This is the recommended method which immediately evokes the appropriate context. For example:
sport: What is a strike? or econ: What is a strike?, etc. This is how most collections in
SuperMemo Library are formulated
Combining multiple pieces of information in one item is not
recommended (#6101)
(Telepolis,
Spain, Thursday, December 20, 2001 1:57 PM)
Question:
What would happen if you presented the pronunciation sound at the same time
as its correspondent word? If you are not looking at the screen you are improving the
comprehension, and if you are looking, you can
check the spelling too. The associative learning will increase
Answer:
Ideally, you should adjust the mode of repetition to your goal. For that purpose it is better to separate learning the pronunciation, from learning the spelling, and from learning the semantics or synonyms. This means that you can create several items for the same word. Experience shows that this is the most efficient method. This is how
Advanced English was designed. However, it is rather
impractical to produce all combinations of items for all words and synonyms. For that reason, Advanced English includes the
pronunciation branch that includes only the words that are hardest to pronounce. Similarly, the spelling branch lists only words that cause most problems with spelling. Bombarding the brain with many stimuli at the same time may produce ambiguous
stimuli and you will not always learn that what you really want to learn. Some important aspects of information can be lost
Incremental reading requires some
experience
Question:
I do not know how to tackle this text in incremental reading. Any hints?
After the discovery of Pluto, it was quickly determined that Pluto was too small to account for the discrepancies in the orbits of the other planets. The search for Planet X continued but nothing was found. Nor is it likely that it ever will be: the discrepancies vanish if the mass of Neptune determined from the Voyager 2 encounter with Neptune is used. There is no tenth planet
Answer:
Here are some exemplary processing stages. Yours might be different. In the
end, you can convert the cloze deletions into more direct and well-formulated
questions-and-answers:
Extract 1: Pluto is too small to account for the discrepancies in the orbits of the other planets
Extract 2: Pluto was too small to account for the discrepancies in the orbits of the other planets. The search for Planet X continued but nothing was found
Extract 3: Pluto was too small to account for the discrepancies in the orbits of the other planets. The discrepancies vanish if the mass of Neptune determined from the Voyager 2 encounter with Neptune is used
Extract 4: There is no tenth planet
Developer's Guide for authors of SuperMemo collections
(anguskwong, Wednesday, August 28, 2002 8:01 PM)
Question:
I would like to develop a knowledge collection similar to 100 English Spellings for Kids with sound, spell pads, etc.
It seems I can not do all these as a normal user. Do I need Developer's Guide to do that?
Answer:
You
can develop such collections with SuperMemo. It is fully functional if you turn
on Full access and Professional
level. Please have a look at these:
Does minimum information principle hold in all cases?
(Simon J., Apr 08, 2004, 08:59:51)
Question:
My wife is using Advanced English and has now gone through about 9000 words. However, each time there is a big gap (say 1 month),
she will always forgets some items (e.g. amenable=responsive/susceptible). Her forgetting index
is set at 7%. It seems the word is too abstract and floating in space. Especially when both of the words were previously unknown. I wonder if the 'minimum information principle' holds true for everyone
Answer:
Not all items in Advanced English are structured ideally. We try to locate such cases and improve them in future upgrades. This is what differentiates Advanced English from a simple dictionary. However, it is not always possible to adapt this material for everyone's needs and some individual intervention is always necessary.
You are right that it is unacceptable to learn a word pair if both words are not understood. Adding translations or one's own examples in such cases is very important.
You could for example start with finding exemplary sentences where "amenable" is used and add them to your Advanced English. You might add then two additional items like these:
Q: Middle East is not [...](responsive) to improvements along American lines
A: Some things are not [...](open) to the approaches of science
Naturally, for passive recognition you need an item too. This could then be:
Q: amenable (adj)(e.g. some things are not amenable to science)
A: open/susceptible
As for the minimum information principle, you should remember that it does not refer to the number of words or characters in the question. It refers to the information that needs to be stored in the brain. For that reasons, cloze deletions built of exemplary sentences (as indicated above), are FAR easier to remember. Even though they are wordy, they produce a very simple association link in memory. They are highly recommended.
As for the forgetting index, you should rather keep a higher forgetting index for the entire collection and lower it only for the most important items. Do not use the forgetting index as a remedy against imperfect structure of knowledge! Modify your items first and lower the forgetting index only if you still have problems with recall and the materials is vitally important.
Why do we remember weird words?
(MRW, Mar 04, 2002)
Question:
Have you noticed that in learning English, the words that are easy to remember are often the "weird" ones (e.g. tintinnabulation)
Answer:
It all depends on the associations formed in your brain. Tintinnabulation may have stuck easily with you while it could be somebody else's pet peeve. However, weirdness often implies uniqueness which helps you avoid memory interference. Then, one day, you learn the word tinnitus and, all of a sudden, tintinnabulation may start causing serious trouble
"Successful" leeches should not be reset
(christian.roessel, Germany, Wednesday, January 10, 2001 11:26 AM)
Question:
What to with a leech that has been well-known for the last
n repetitions? Is it more time-efficient to forget and reintroduce this element or shall I keep it as is?
Answer:
If the leech has been remembered for a number of repetitions, its inherent problem might have already been resolved. In general, you should take postpone, reformulate or delete actions only at the moment of forgetting the leech. As
long as it progresses towards longer intervals, your best action might be to do nothing or to add minor edits that you
believe could improve retention
You can learn Arabic with SuperMemo
Question:
Can I learn Arabic with SuperMemo?
Answer:
Yes. See: Arabic
Verbs collection
Copying material from a dictionary (#17003)
(Rune, Norway, Monday, April 28, 2003 1:38 AM)
Question:
I copy word descriptions from the Collins Cobuilder dictionary and paste them into the answer field. It would be nice, if SuperMemo could create a new learning item and paste the description into the answer field. Now I first have to copy from Collins, create an new element, and paste into the answer field
Answer:
The best way to handle dictionary items is to paste the entire item to SuperMemo with Ctrl+Alt+N. Then extract individual definitions along with the headword with Alt+X. Finally, while learning individual definitions, create individual passive, active or detail items with Alt+Z
Here is an example of learning the meaning of the word trachea. Although there are 19 items on the output, not all these items are necessary to extract the basic meaning of the word. For that reason, the process can be executed incrementally. More specialized meaning can be refined in more advanced stages of learning.
traˇcheˇa
( P ) Pronunciation Key (tr
k
-
)
n. pl. traˇcheˇae (-k) or traˇcheˇas
- Anatomy. A thin-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs. Also called windpipe.
- Zoology. One of the internal respiratory tubes of insects and some other terrestrial arthropods.
- Botany. One of the tubular conductive vessels in the xylem of vascular plants.
trachea
- Anatomy. A thin-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs. Also called windpipe.
- Zoology. One of the internal respiratory tubes of insects and some other terrestrial arthropods.
- Botany. One of the tubular conductive vessels in the xylem of vascular plants.
trachea Anatomy. A thin-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs. Also called windpipetrachea Zoology. One of the internal respiratory tubes of insects and some other terrestrial arthropods
trachea Botany. One of the tubular conductive vessels in the xylem of vascular plants
trachea Anatomy. A thin-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs. Also called windpipe
a cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs trachea
trachea: A [thick/thin]-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs thin (thickness is a relative concept and you may want to skip that property)
trachea: a [bony/cartilaginous/muscle/membranous] tube descending from the larynx to the bronchi cartilaginous
trachea: A cartilaginous tube [descending/ascending] from the larynx descending
trachea: A tube descending from the [...] to the bronchi larynx
trachea: A tube descending from the larynx to the [...] bronchi/lungs
trachea: A tube descending from the larynx to the bronchi and carrying [...] to the lungs air
trachea: a tube carrying air to [...] (the) lungs/bronchi
trachea: A tube carrying air to the lungs. Also called [...] windpipe
Q: a cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs
A: trachea
Q: trachea: A [thick/thin]-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs
A: thin
Q: trachea: a [bony/cartilaginous/muscle/membranous] tube descending from the larynx to the bronchi
A: cartilaginous
Q: trachea: A cartilaginous tube [descending/ascending] from the larynx
A: descending
Q: trachea: A tube descending from the[...] to the bronchi
A: larynx
Q: trachea: A tube descending from the larynx to the[...]
A: bronchi/lungs
Q: trachea: A tube descending from the larynx to the bronchi and carrying[...] to the lungs
A: air
Q: trachea: a tube carrying air to [...]
A: (the) lungs/bronchi
Q: trachea: A tube carrying air to the lungs. Also called [...]
A: windpipe
Q: zool: [...]: one of the internal respiratory tubes of insects and some other terrestrial arthropods
A: trachea
Q: zool: trachea: one of the internal[...](function) tubes of insects and some other terrestrial arthropods
A: respiratory
Q: zool: trachea: one of the internal respiratory tubes of[...](main animal group) and some other terrestrial arthropods
A: insects
Q: zool: trachea: one of the internal respiratory tubes of insects and some other [aquatic/terrestrial] arthropods
A: terrestrial
Q: zool: trachea: one of the internal respiratory tubes of insects and some other terrestrial [...](phylum)
A: arthropods
Q: bot: trachea: one of the[...] in the xylem of vascular plants
A: (tubular conductive) vessels
Q: bot: trachea: one of the tubular conductive vessels in the[...](tissue) of vascular plants
A: xylem
Q: bot: trachea: one of the tubular conductive vessels in the xylem of[...](division) plants
A: vascular
Q: trachea: one of the tubular conductive vessels in vascular [plants/animals]
A: plants
Q: bot: [...]: one of the tubular conductive vessels in the xylem of vascular plants
A: trachea
How can SuperMemo help an actor?
(David Baldwin, Saturday, August 17, 2002 4:45 AM)
Question:
As an actor specializing in one-man plays, I'm required to memorize and regurgitate hours of material. How can I best use SuperMemo to quickly and permanently memorize?
Answer:
SuperMemo is primarily useful in learning material that is to be retained in memory for months and years. It might be very useful in lifelong retention of quotes, citations or proverbs. However, if you need to memorize long passages, and you need to be able to recall them for weeks rather than years, SuperMemo's applicability will be limited. You could use it, for example, to cram pieces on which you
stumble or get stuck. You could paste the problematic piece of text and generate lots of cloze deletions (i.e. questions with blanks to fill). With regular repetitions, the recall of such pieces of text will become quite automatic.
You cannot expect much qualitative change in your performance as an actor by employing SuperMemo today. However, it is still recommended that you get to know it. Gradually, you will get a better grip on the inner workings of your memory. You will also find optimum ways of employing the program. With time, you might discover that SuperMemo is more useful that suggested in this answer. However, that will depend on your creative approach to selecting the right portions of knowledge to learn, and representing them optimally for your purposes
Strategic recall is a matter of knowledge structure
(Noel L, Sweden, Tue, 19 Mar 2002 13:11:25 -0000)
Question:
How can I bring knowledge learned with SuperMemo into my mind when I need it
Answer:
Ideally, the knowledge is already stored in your memory and it comes up
naturally when you need it. However, the picture is more complex if you consider
that the way you ask the question in SuperMemo may differ from the way your life
asks you the same question. In other words, you may store some material in
SuperMemo, but a real-life situation will trick you into being unable to recall
it. In other words, you need to properly formulate the material to maximize its
recall in all potential contexts (see: "20
rules" article). There are two basic tools that will help you keep
knowledge at hand in need. The first is simplicity. Simple things are easier to
remember, but they are also easier to apply in varying context. The second is
universality of rules. For example, it is better to learn a universal
mathematical formula than just the examples of its use. Examples can be used to
emphasize applicability in various context. To sum it up: to make knowledge
available at hand, you need lots of practical experience in using SuperMemo. See
for yourself what works and what does not
Sequences are best learned through logical consequence DNA Replication:
(Gary B., Canada, Mar 25, 2005, 19:37:03)
Question:
I am memorizing DNA synthesis. After formulating various extracts and cloze questions, how can I ensure that I am able to recall the events in the proper
sequence? Here is the material I am learning:
Answer:
All you need to do is to ensure that the sequence is a logical consequence of
extracts you are processing.
For example, the following sentence creates a link between events: Once DNA is unwound, a replication fork forms. The preceding event is unwinding DNA, and the succeeding event is the forming of the replication fork (naturally, you must know what a replication fork is before you go into details). If this exemplary sentence is well consolidated in your memory, you will have little problems with figuring out the correct logical sequence of events.
In other words, there is no need to memorize the sequence in a separate set of questions. If you understand the sequence, see the big picture, and consolidate individual steps, you shall have little problems with reproducing the events of DNA replication (today or 30 years from now). This guarantee requires primarily that: (1) you understand the text and (2) you generate some material redundancy to counteract the forgetting index. You generate redundancy with prolific cloze deletions. Here is how your material might look once incremental reading process is completed:
Q: Once parental DNA is [...], replication fork forms at fixed origin of replication
A: unwound
Q: Once parental [...] is unwound, replication fork forms at fixed origin of replication
A: DNA
Q: Once parental DNA is unwound, [...] fork forms at fixed origin of replication
A: replication
Q: Once parental DNA is unwound, replication [...] forms at fixed origin of replication
A: fork
Q: Once parental DNA is unwound, replication fork forms at fixed origin of [...]
A: replication
Q: Once parental DNA is unwound, replication fork forms at fixed [...] of replication
A: origin
Q: One new DNA strand called [...] is synthesized continuously as DNA polymerase moves toward replication fork
A: leading strand
Q: One new DNA strand called leading strand is synthesized [continuously / discontinuously] as DNA polymerase moves toward replication fork
A: continuously
Q: One new DNA strand called leading strand is synthesized continuously as DNA polymerase moves [towards / away from] replication fork
A: toward
Q: One new DNA strand called leading strand is synthesized continuously as DNA polymerase moves toward [...]
A: replication fork
Q: One new DNA strand called leading strand is synthesized continuously as [...](enzyme) moves toward replication fork
A: DNA polymerase
Q: Because DNA polymerase only adds new nucleotide to [3/5]', RNA primer (synthesized by RNA Polymerase) is needed to start synthesis
A: 3
Q: DNA is synthesized in [5'->3' / 3'->5'] direction
A: 5'->3'
Q: Because DNA polymerase only adds new nucleotide to 3',[...] primer is needed to start synthesis
A: RNA
Q: Because DNA polymerase only adds new nucleotide to 3', RNA [...] (synthesized by RNA Polymerase) is needed to start synthesis
A: primer
Q: Because DNA polymerase only adds new nucleotide to 3', RNA primer (synthesized by [...](enzyme)) is needed to start synthesis
A: RNA Polymerase
Q: Because DNA polymerase only adds new nucleotide to 3', RNA primer (synthesized by RNA Polymerase) is needed to [...]
A: start synthesis
Q: [Lagging / Leading] strand of new DNA is synthesized in approx 1000 piece nucleotides called Okazaki fragments as DNA polymerase moves away from replication fork
A: Lagging
Q: Lagging strand of DNA is synthesized in approx [...] piece nucleotides called Okazaki fragments as DNA polymerase moves away from replication fork
A: 1000
Q: Lagging strand of DNA is synthesized in approx 1000 piece nucleotides called [...] as DNA polymerase moves away from replication fork
A: Okazaki fragments
Q: Lagging strand of DNA is synthesized in approx 1000 piece nucleotides called Okazaki fragments as DNA polymerase moves [towards / away from] from replication fork
A: away
Q: Lagging strand of DNA is synthesized in approx 1000 piece nucleotides called Okazaki fragments as DNA polymerase moves away [...]
A: replication fork
Q: DNA polymerase digests RNA primer and replaces it with [...]
A: DNA
Q: [...](enzyme) joins new discontinuous fragments of lagging strand
A: DNA ligase
Q: DNA ligase joins new discontinuous fragments of[...] strand
A: lagging
Before you start learning, you would best ensure you understand the big picture of the entire replication process as your short sequence does not fully illustrate it. You can also build the big picture incrementally, but that may take substantial time and is not recommended when you learn for an exam. If you have the big picture in your mind before you begin to learn incrementally, the savings in time may be substantial. However, if you are learning from lecture notes, this rarely is a case (a testimony to the dismal efficiency of taking notes during lectures in the "old world" of learning). Luckily, you can easily substitute with plethora of well-written materials imported from the web.
For the HTML source of the incremental reading process that led to the above questions see: DNA Replication example
You can learn to retrieve words faster from memory
(Samson Chen, Aug 10, 2005, 01:17:00)
Question:
I learn English. I use items like this one:
Questions: to move in a direction suddenly and quickly
Answer: shoot
Verb prep/adv
eg. A car shot by.
At the moment, I have no problem recalling the definition of a term. And if I see the word "shoot", I know one of its meanings is "move in a direction suddenly and quickly". However, I have trouble using these words in daily life. Although my oral English is fluent, I am unsatisfied with limited vocabulary in my speech
Answer:
The main problem with your item is that the question does not bring the word "shoot" instantly to your mind. A fluent speaker might answer your question differently (e.g. "rush", "dart", "hurry", "blast", etc. ). In other words, your question part is not specific enough.
Your optimum learning strategy should include (1) a passive recognition of the word "shoot" (which you have already accomplished), (2), the ability to use any term for "moving quickly" (e.g. the answer might then be "rush/dart/shoot" and any term would qualify for a good grade), and, optionally, (3) some enriching items that would help you use various synonyms and enliven your language.
For the latter, you could use cloze deletion examples that would anchor a specific synonym in a specific context. For example: "The police car ... by in a chase", or "Johnny English ... out from a hot tub".
Passive items will help you understand other people. Active synonyms will help you express yourself. Anchored synonyms will help you enrich your language.
You may memorize a couple of cloze examples for the word "shoot". It is important though, that you do not rigidly cram them if you fail to anchor "shoot" in a given context. If you fail a few times, change your example. Probably it is not specific enough.
Some narratives may be thin on memorizable material
(Tanya, Sunday, November 24, 2002 11:46 PM)
Question:
I am
trying to learn incremental reading. For starters, I decided to learn the incremental
reading FAQ using incremental reading. However, I have no idea how I might
cope with FAQs such as "
High priority of material or long review intervals will prompt you to run an
article preview".
The answer is too long and I don't know how to grab it.
Answer:
Not all materials are suitable for generating extracts and clozes. Some
narratives should just best be read passively. Like the quoted FAQ, they may be
a compilation of facts that are generally obvious. In such cases you can just
read and dismiss. Or you can read and schedule another review in a month or in a
year (if you worry you miss something important). Or you can try to write, in
your own words, a sentence or two on what new things you have learned from the
narrative. Your sentence would shortly extract the quintessence from an
otherwise lengthy passage. If it is meaningful and quintessential, you shall
find little trouble with locating keywords suitable for clozing.
Nevertheless, if you really wanted to process that particular FAQ with incremental reading, you should eliminate as much verbiage as possible, avoid memorizing it as an enumeration, and focus on its components that are the least obvious. In such a case you might arrive at the following set of questions and answers:
Q: The most important incentives for whole-article preview: [low/high] priority of the material
A: high
Q: The most important incentives for whole-article preview: [short/long] inter-review interval
A: long
Q: The most important incentives for whole-article preview: [presence/absence] of higher-priority fragments buried in a lower-priority text
A: presence
Q: Extract-preview will [increase / decrease / not change] your exposure to previewed material
A: increase
Q: Line-at-a-time reading will be equivalent to assigning a [lower/higher/same] priority to an article (assuming you do not interfere with intervals, A-Factors, priority settings, etc.)
A: lower
Q: If your reviews occur in very long intervals as a result of slow reading, you may opt for [...](passive method) or running a preview of the most important sections instead
A: shortening the interval
Q: If your reviews occur in very long intervals as a result of slow reading, you may opt for shortening the interval or [...](active method)
A: running a preview (of the most important sections)
Q: if you are reading texts from your e-mail tasklist, preview is highly recommended because [...]
A: not everyone starts their message with the most important points
Inventiveness is the key to mnemonic techniques
(Daria, Jan 12, 2006, 19:13:30)
Question:
I could never remember the name of tryptophan. I remember its name just more or less. Do you think it is a good idea to add an item like this:
Q: [Hit the fan] is an amino acid that is essential in human nutrition
A: tryptophan
Answer:
Yes. This item will work, esp. if "hit the fan" is what comes to your mind when speaking of tryptophan. However, your items does not say anything characteristic of tryptophan itself (many other amino acids are essential as well). In other words, you are only learning to associate the name and "hit the fan" sound. You could fare much better in the long run if you used the formula of tryptophan as the question (e.g.
"What amino acid is presented in the picture?"), or better yet ask a question based on something unique or memorable about tryptophan. For example, you could ask
"What amino acid used as a supplement caused an autoimmune disease outbreak due to supplement contamination (US,
1989)?"
Math proofs in SuperMemo
(JOSEPH PRIMO BELARMINO, Feb 06, 2006, 03:18:38)
Question:
I'm trying to format math proofs into SuperMemo, but I'm not sure exactly how to
make it so that it's always shown sequentially.
Hence if I have a proof that goes like:
often I'd get the step 2 window such that I'm unintentionally getting information for the step 1. How do I fix this?
Answer:
You can decompose the proof into individual cloze deletions. For example:
Q: Proving X:
- step 1: let a=b
- step 2: [...]
- step 3: using theorem d, c is the final answer
The sequence of review is not important. Repetitions are not supposed to help you build the big picture of the proof. They are just supposed to refresh memories on the assumption you have already built the coherent picture of the proof. The more clozes you create, the less likely you are to lose the coherence of the proof. Even for simple proofs as above, it may often be necessary to generate 6-8 separate cloze deletions. Moreover, individual steps are usually a bit more complex and need to be clozed too.
The difficulty is only in making sure that your clozes are well-structured and do not provide hints that make answers obvious. For example, you do not need to worry about the hint given by Step 1. After all, in real life, you will always worry about Step 2 only once you solved the Step 1. However, it may be necessary to delete Step 3 from the question as it may hint on the action taken in Step 2. You can, for example, move Step 3 from the question to the answer in the form of the comment:
Q: Proving X:
- step 1: let a=b
- step 2: [...]
A: b=c (step 3: using theorem d, c is the final answer)
It will also be helpful if you try the entire proof in a real-life situation to make sure your clozes indeed do the job of making you able to prove the theorem
Syllables may be easier to remember than numbers
(stephanie phillips, Jan 14, 2006, 21:40:55)
Question:
Is it easier to remember a sequence of letters (example aw by hu il te be ha qx lm ni) rather than
remembering a sequence of numbers (example 33 57 96 75 29 65 92 01 75)? If so why?
Answer:
No
theory answers such questions better than a practical test. If you put a couple of examples of both sequences to SuperMemo, you will be able to compare performance in a number of ways (for example, an average inter-repetition interval after a longer period of time). Such tests are important as individuals may differ substantially in the way they convert question-answer pairs into specific representations in their mind. Moreover, there is substantial interference between student's knowledge, knowledge currently learned and retained with SuperMemo, and the newly memorized material.
If syllables are easier to recall than numbers then, as in all similar cases, it is related to the complexity and uniqueness of individual memories that are to be consolidated. Numbers are less unique and an average
individual does not store specific representations beyond the ten digits, plus a set of numbers (s)he easily associates with other memory landmarks such as one's own birthdate, the date of landing on the moon, etc. Syllables, on the other hand, are central for language processing and a typical set used by an average individual (including "sounds-like" syllables) will be far larger than the set of landmark numbers.
The ability to store memories in a specific form is related to their evolutionary applicability. Humans have used basic numbers in a limited way for a couple of
millennia. That time-span is too short to equip our brain with specific number circuits. Spoken language has been in use for a period 10-40 times longer. And the best mnemonic tool,
visualization, capitalizes on circuits that have been in use for hundreds of millions of years (i.e. even before the appearance of vertebrates).
The importance of uniqueness of memories is well illustrated by the fact that you can best defeat a mnemonist by asking him or her to memorize strongly interfering patterns. For example, when asking to memorize objects, you can list those that are very similar to each other, e.g. rod, javelin, stick, pole, whip, pale, spear, pike, etc.
Language ambiguities can be tackled with SuperMemo
(G.W., Mar 09, 2007, 03:26:54)
Question:
I
learn Spanish. Spanish words often have multiple English meanings. When all common English meanings are matched to the same Spanish word, failure rate dramatically increases. When single meanings are matched, much is lost
Answer:
When
multiple meanings slow your progress, you could define Spanish words using
Spanish definitions. Thus you can ask for two different meaning of the word
"album" with two distinct questions:
- a book with blank pages for the insertion of collections of stamps
- a set of recordings issued together on cassette tape or compact disc
This approach should effectively cover for ambiguities in active recall. As for passive recognition, you can use synonyms with appropriate context. For example:
As your fluency increases, you can hone your semantic sensitivities with exemplary sentences processed with incremental reading. Here is another meaning of "album" taken from a real life situation:
We sat with my husband browsing our photo [...](book with pictures) from Madagascar.
Pick those sentences from real life situations in which your original active or passive items did not do their job up to your expectations.
If sequence is important, it must be included in a single item (#348)
(Martin Zielinski, Friday, January 21, 2005 12:27 AM)
Question:
How can you memorize a poem if SuperMemo keeps randomizing the
entries I input
Answer:
If you need to recite the entire sequence in one go, you must simply include the entire sequence in the item's answer.
It is economical to use SuperMemo for poems only if you need to be able to recall them for long periods of time. If you learning process is sparse (few repetitions, few postpones), you might consider using SuperMemo for poems if you need to retain them for months. However, a heavily loaded learning process may extend that minimum period to years. Learning poems is simply "workload intensive" and will stifle progress in all other areas.
As for the extraction of memory bottlenecks, the optimum strategy will depend on your goals, available time, and the required fluency. The higher the fluency required, the greater the redundancy needed. What makes learning poems costly is the fact that redundancy grows exponentially with fluency. For high levels of fluency, traditional poem learning techniques may be necessary and frequent recitation unavoidable.
If you want to apply SuperMemo, one item must include the call to reproduce the entire poem and must use a forgetting index that will determine the probability of recalling the entire poem. Without this item, you will always be able to reproduce fragments of the poem, but there is a possibility of the disruption in the chain. That cumulative probability will be very high for a large number of paragraphs (each carrying its own probability of forgetting).
In addition to the main poem item, you should include individual items in which the preceding line evokes the succeeding line. Those items will assist automation of the recitation and plug up weak-link bottlenecks. The hardest fragments thus will be repeated more often.
For example:
Q: Look, how spry she still is, how well she holds up:
A: hatred, in our century.
Q: Look, how spry she still is, how well she holds up: hatred, in our century.
A: How lithely she takes high hurdles.
Q: hatred, in our century. How lithely she takes high hurdles.
A: How easy for her to pounce, to seize
and so on.
If it appears that one of the items above still causes frequent problems (e.g. wrong recall, substitutions, wrong word order, etc.), you will need to generate more specific queries.
For example:
Q: Look, how [...] she still is, how well she holds up: hatred, in our century.
A: spry
Q: Look, how spry she [...], how well she holds up: hatred, in our century.
A: still is
Q: Look, how spry she still is, [...]: hatred, in our century.
A: how well she holds up
The fastest way to generate individual items is to use incremental reading tools. In addition, if you are not in a hurry, the entire process should be incremental. This means that you should generate detailed items only from fragments of the poems that you frequently stumble on. Processing the entire poem at once will generate an astronomical number of items and be very costly. Such gained redundancy will improve recall, but may make the entire process uneconomic