e is for elaboration, g is for generation · the basic mechanism is simple-ideas that fire together...
TRANSCRIPT
Elaboration Making memories meaningful
ELABORATION IMPROVES MEMORY by making connections between new information and prior knowledge.
The outcome of the following scenario should be familiar. Sandra discovered a potential client while shopping at the mall. As they parted, the client said, "Call me at 422-8888." Sandra eagerly repeated the phone number while walking back to her car. On the way, a passerby asked for the location of a restroom. After answering the passerby, Sandra realized she had utterly forgotten the phone number she had just been repeating so successfully.
Poor Sandra-she should have tried to make meaningful patterns of the numbers. For example, "This would be my fourth client; 4 divided by 2 makes 2 (422), and adding them up makes 8, of which there are 4 again (8888)." By only repeating the phone number, Sandra did nothing to commit the number to long-term memory. Elaboration helps by actively connecting new ideas to ideas that are already in long-term memory, for example, by associating the phone number with the knowledge that this would have been her fourth client, and that 4 + 2 = 2. It is a basic truth worth memorizing: repeating something over and over keeps it in your mind fleetingly, whereas connecting new information to what you already know creates a memory. So, how will you remember this truth?
52
FROM: Schwartz, D. L., Tsang, J. M., & Blair, K. P. (2016). The ABCs of how we learn: 26 scientifically proven approaches, how they work, and when to use them. WW Norton & Company.
E IS FOR ELABORATION I 53
I. How Works
The human body has many different memory systems, each specializing in a
different type of information. The immune system, for example, has a mem
ory. When people receive bone marrow transplants, doctors destroy the existing marrow. The immune system "forgets" all the diseases it has encountered, and it needs to relearn from (painful) experience. The immune system is the poster
child for the wisdom that it is much more efficient to remember a solution than
figure it out afresh. At the same time, the immune system also cautions us that memory is insufficient for adaptive behavior, because we still need to handle
novel problems (new germs) for which we have no exact memories. Elaboration is a strategy specialized for memorizing declarative informa
tion-things about which one can talk. The amount of declarative information that people know is astonishing. Newspaper articles, movies, a bully
from second grade, sundry math facts, your friend's preferences, alphabeti
cal order, favorite dinners-it's all in there. It is a good thing people do not remember everything at once! The great trick of memory is to remember the
right thing at the right time. Elaboration helps. To understand how elaboration works, we need to consider two of our
many memory systems. One is called working memory. It enables the conscious manipulation of information, for example, when thinking through a
problem. Working memory has only temporary storage. Information moves
ENCODING
Think of password in ~
working memory
Elaborate in long-term memory
RETRIEVAL
Search Memory cue ~ long-term ~ Successful
retrieval memory
Figure E.1. The benefits of elaborative encoding for retrieval. Elaboration makes connec
tions between an idea in working memory and long-term memory. During retrieval, the
increased pathways among ideas improve the chances that the memory will be found.
54 I THE ABCs OF HOW WE LEARN
in and out of working memory depending on the problem at hand. Working
memory cannot hold information very long, so people need to keep refresh
ing the information, for example, by repeating the name of a person they just
met. Refreshing the information keeps it available for immediate processing
in working memory, but it is a poor technique for storing the information for
later use. To file the information away for later use, people need to encode
it in long-term memory. Long-term memory holds information indefinitely.
Encoding is only half the story, however. To use the stored information, peo
ple need to be able to retrieve it from long-term memory to help solve prob
lems and answer questions using their working memory.
Figure E. l shows the distinction between encoding and retrieval. The left
side, which is a massive simplification of what goes on in the brain, shows
how someone might use elaboration to remember the password to a Bank of
America online account. The elaboration connects both the bank's name and
the password (Bald&Bold) to the idea of America, using the associated ideas
of brave and eagles.
The right side of Figure E.1 shows the payoff of elaboration. Retrieval
is a process of spreading activation. Working memory makes a request of
long-term memory, such as "What is my Bank of America password?" This
activates a long-term memory, such as America. The activation in America
cascades across associated ideas. The activation spreads through several con
nections that lead to the desired memories, which become sufficiently active
that rhey become available to working memory.
Think(!{ password in
working memory
ENCODING
' •@) I ~ ' ,
No elaboration in long-term
memory
RETRIEVAL
' ~: '
,
Search
Memory cue -- long-term ~
memory
Failed retrieval
Figure E.2. What happens when people do not use elaboration. During encoding, the
new ideas (Bald and Bold) do not connect with other ideas in long-term memory. During
retrieval, there is no path back to the original ideas, so they cannot be found in memory.
E IS FOR ELABORATION j 55
Figure E.2 shows what happens when people do not elaborate new infor
mation. Even though the information may make its way to long-term memory, it remains unconnected to any other ideas. It is hard to search for the memories, because there are no connections to them. As an analogy, it is much easier to find documents in a well-organized filing system than a box filled with mounds of paper.
The formal representation of memory networks can become very complicated. The basic mechanism is simple-ideas that fire together wire together, and ideas that wire together fire together. Elaboration increases the chances that one idea will fire another idea you want to remember.
to
Elaboration is useful for memorizing meaningful material, including new vocabulary words, sentences, people's names, directions, or even phone numbers. Ironically, elaboration does not need to be very elaborate to make a
difference. In one study, participants had to memorize one hundred words
(Tresselt & Mayzner, 1960). There were three conditions: cross out the vowels, copy the words, and judge the degree to which each word was an instance
of the concept "economic" (e.g., poem would be low, and credit would be high). When asked to remember the words, participants in the judge con
dition did twice as well as the copy condition and four times as well as the cross-out-vowels condition.
If the goal is to memorize, elaborations do not need to be correct. For instance, given the sentence ''A group of woodpeckers is called a descent," people can elaborate the sentence by manufacturing a "because" to finish
it out. A group of woodpeckers is called a descent because ... together they could fell a tree. Of course, it is probably best to elaborate with a true reason
so that one does not build up misconceptions (e.g., a group of woodpeckers may actually be called a descent because woodpeckers like to peck trees from
the top to the bottom). Nevertheless, to simply remember the woodpecker-descent connection, the accuracy of the "because" does not matter.
The basic strategy for improving memory is to make up relevant connections to what one already knows. There are three complementary approaches: precise and relevant elaborations, chunking, and connecting to well-structured knowledge.
PRECISE AND RELEVANT
Relevant and precise elaborations create better retrieval paths. For instance, in one study people read sentences in one four conditions (Stein & Bransford, 1979):
56 I THE AB Cs OF How WE LEARN
(a) The tall man purchased the crackers. (no elaboration) (6) The tall man purchased the crackers. (come up with own elabora
tion) (c) The tall man purchased the crackers that were on sale. (irrelevant
elaboration) (d) The tall man purchased the crackers that were on the top shelf.
(relevant elaboration)
People read 10 sentences like these. Afterward, they took a memory test in which they had to provide the missing adjective for each sentence: The ____ _
man purchased the crackers. The percentages of correctly remembered words were as follows:
(a) 42 percent (No Elaboration condition)
(b) 58 percent (Self-Elaboration condition) (c) 22 percent (Irrelevant Elaboration condition)
(d) 74 percent (Relevant Elaboration condition)
The relevant elaboration provided a precise connection that marked the relevance of the man's height for reaching the crackers. On average, receiving a relevant elaboration was even more effective than generating a
Figure E.3. Meaningful elaborations can help us
remember images. How would you elaborate these
images to improve your chances of remembering them
later? (Reprinted from Schwartz, 1999.)
E IS FOR ELABORATION I 57
self-elaboration. This is likely a result of some people making ineffective elaborations: when people in the self-elaboration condition did generate
their own precise and relevant elaborations, their probability of recall was
91 percent (Stein & Bransford, 1979). Relevant elaboration can also improve memory for pictures. Pretend you
need to remember the images in Figure E.3. They may appear as a bunch
of very hard to remember squiggles. It would help if you could elaborate by connecting the squiggles to something you already know. Here is a clue: the
image in the upper left corner is the head of a baseball player. For fun, we will let you figure out the other three (we provide them below). One handy use of
elaborating visual information is to remember people's names. Find a way to connect a name to its owner's facial features. To make up an example, the two Bs in "Burt Bennett" stand for his black beard.
CHUNKING
Chunking depends on urntmg discrete ideas (see also Chapter D). For instance, to remember the digits 2 6 2 4 2 2 2 0, one can chunk them as 26,
24, 22, and 20. Although there are still eight digits, it is easier to remember
four numbers than eight. (One can go further to elaborate that each chunk is two less than the preceding one.) Chunking depends on elaboration, because
it is prior knowledge that enables people to convert two digits into a single number (and to notice the pattern of subtracting by 2). A second example is
to chunk words into a sentence: house, dog, car, sprint, spring • Last spring, my dog sprinted from the house to chase a car. Here, the list is elaborated by
chunking the words into a single sentence and by connecting the sentence to prior experience.
CONNECTING TO WELL-STRUCTURED KNOWLEDGE
A third elaborative technique associates the new material to well-structured knowledge. The method of loci is an example. If people need to remember a
series of steps, parts of speech, or other sequential information, they can associate each step to a different room of their house or office. Once they have mentally "placed" the memories in their locations, they can later retrieve the
memories by taking an imagined walk through the house. The spatial mem
ory for the house provides a well-known search path for finding the memories. When you try to remember a person's name by working through the
alphabet, you are relying on a well-known sequence to help search memory. Did it start with an A, B, C ... ?
A hierarchy is also an example of a well-organized structure. In a famous study (Ericsson, Chase, & Faloon, 1980), a person had to remember long lists
of numbers. The person happened to be an active marathon runner. He took
58 I THE AB Cs OF How WE LEARN
sequences of numbers and associated them with known race times. For instance,
3 5 9 1 2 becomes 3 hours 59 minures and 12 seconds. He then further orga
nized the elaborated chunks into a hierarchy such as race times early and late in
his career. This way, he only needed to remember "my early career race times,"
and it would connect to the chunked times and the digits within each chunk.
The three remaining elaborations for Figure E.3 are James Dean (lower
left), a baby (upper right), and Santa Claus (lower right) (Schwartz, 1999).
Ill. The Outcomes of Elaboration
Elaboration improves memory for declarative information, especially under conditions of cued recall. There are different conditions under which one
might need to remember something. For instance, one condition merely involves recognition memory: ''I've seen that painting before." Under condi
tions of cued recall, a thought or stimulus cues memory for an associated
idea. Most school-based memory tests rely on cued recall: Given the name of
a vocabulary word, remember the definition. Remember the prime factors of
12. What is an example of an apex consumer? In each case, having more con
nections between the possible cues and the memory will improve the chances of retrieving the answer.
IV. Can People Learn to Teach Themselves with Elaboration?
As presented here, elaboration is primarily an aid for memory. (To improve
understanding, it requires more constrained forms of elaboration, such as self-explanation; see Chapter S.) Elaboration strategies are relatively straight
forward to teach and learn. In one study (Weinstein, 1982), researchers told
adolescents several elaboration strategies, including create a mental picture,
invent a story or sentence, and draw similarities and differences to what they already know. The students practiced with further coaching once a week for
five weeks. On the sixth week, they read a passage without coaching. One
month later, the students showed better memory for the passage than students who had not received elaboration training.
Even five-year-olds can learn simple elaboration strategies (Yuille & Catchpole, 1973). Children played a game in which they had to remember
which objects went together. Researchers trained one group of children to
imagine object pairs interacting, such that children learned to generate their own elaborations. The experimenters showed children ten pairs of objects side
by side and then showed how the objects in each pair could interact, such as
putting a hat on a duck, or a rock on a spoon. After the ten training pairs, the
experimenters told the children they would get new pairs of objects to study.
7
E IS FOR ELABORATION I 59
They encouraged the children to think about how the objects in the new pairs might interact, similar to what they had seen in the training pairs. The chil
dren then saw a sequence of twenty pairs of objects set side by side. This group of children was compared with two groups who were not trained
to imagine objects interacting. In the Side-by-Side condition, children simply saw the twenty pairs of objects and heard that they "go together." In the Inter
acting condition, the experimenters demonstrated the twenty pairs of objects interacting together, so that the children saw experimenter-generated elaborations. All the children then completed a cued-recall test. They saw one of the
objects (e.g., a rock) and had to pick out the paired object (e.g., a spoon).
Elaboration Training
11.6
No Training
Side-by-side
6.1
Demonstrated Interacting
12.2
Table E.1. Memory performance of five-year-olds taught and not taught to elaborate twenty pairs of objects (Data from Yuille & Catchpole, 1973)
Table E.l shows the results of the cued recall test for the twenty study pairs. There are two relevant comparisons. The first is that the children who
received the elaboration training nearly doubled the performance of other
wise equivalent children who did not receive training and also saw each pair just side-by-side. The second is that children who learned to elaborate a con
nection between two objects did nearly as well as the children who saw the experimenter explicitly show the objects interacting. This indicates that the
trained children had learned to elaborate quite well, even when not receiving the support of the experimenter to do so. Whether these children continued to
elaborate for the rest of their lives when nobody was telling them to imagine is unknown (and unlikely). Nevertheless, the study provides a concrete model
for how to help children understand how they can use an elaborative strategy.
\I. Risks Elaboration
Cognitive overload can block opportunities for elaboration. Perhaps you have
had the experience of being introduced to someone and forgetting her name
60 I THE AB Cs OF How WE LEARN
within a minute. You were probably putting most of your cognitive resources
into the social aspects of the introduction (e.g., saying something intelligent
and observing the reaction). You did not have the cognitive resources to spare
for elaborating the name and committing it to long-term memory.
Working memory can work with only a few ideas at once. When there is
too much information simultaneously, people experience high cognitive load
and it is difficult to elaborate, because just keeping all the information in
working memory is so much work. High cognitive load is a problem during
many college lectures. Chemistry and mathematics lecturers are notoriously
inconsiderate, because they present many new equations and ideas one after
another. The sheer number of new ideas introduced per minute makes it dif
ficult to pay attention to all the new information and elaborate it simultane
ously. Professors often forget how hard they worked to make sense of all that
they know~after all, it seems so obvious to them now. Consequently, they
offer lectures that are suitable for their peers, not for students. Compared
with students, their faculty peers can readily follow the lectures, because they
already know much of the information being presented and they have well-or
ganized knowledge structures for quickly connecting a new equation or idea
to what they know.
A second elaboration risk is that people can misidentify what they need
to remember. Imagine that you need to remember the following words: bun,
stew, bee, boar, chive, ticks. A reasonable strategy is to elaborate the words
into food and creature categories. But what if the recall task demanded
remembering the order of the words? Then it would be better to use the peg
word method: one-bun, two-stew, three-bee, fuur-boar, five-chive, six-ticks. It
is important to define one's memory goals before elaborating.
A common practice asks students to construct sentences using new vocab
ulary words. When done well, it entails elaboration and improves memory
for the vocabulary words and their meaning. When done poorly, it is useless.
Here is a series of possible sentences for the word gloaming (twilight) ranked
from worst to best with respect to elaboration:
Gloaming is spelled g-1-o-a-m-i-n-g. 0 Not meaning focused
Gloaming means twilight. 0 Rehearsal, not elaboration
Gloaming could be a name for my dog. Sit, Gloaming, sit. 0 Irrelevant elaboration to prior knowledge
E IS FOR ELABORATION I 61
Gloaming is when the horrible mosquitoes finally went to bed during my
recent hike. • Relevant elaboration to prior knowledge The lovers glowed as they stole a kiss under the gloamingfringe of the night's
cloak. • Precise, double elaboration comprising the sound of the word and
its meaning
VII. Reforem::es
Ericsson, K. A., Chase, W. G., & Faloon, S. (1980). Acquisition of a memory
skill. Science, 208(4448), 1181-1182. Schwartz, D. L. (1999). The productive agency that drives collaborative learn
ing. In P. Dillenbourg (Ed.), Collaborative learning: Cognitive and computational
approaches (pp. 197-218). New York: Pergamon. Stein, B. S., & Bransford, J. D. (1979). Constraints on effective elaboration: Effects
of precision and subject generation. journal of Verbal Learning and Verbal Behav
ior, 18(6), 769-777. Tresselt, M. E., & Mayzner, M. S. (1960). A study of incidental learning. Journal of
Psychology, 50(2), 339-347. Weinstein, C. E. (1982). Training students to use elaboration learning strategies.
Contemporary Educational Psychology, 7(4), 301-311. Yuille, J.C., & Catchpole, M. J. (1973). Associative learning and imagery training in
children. journal of Experimental Child Psychology, 16, 403-412.
62 I THE AB Cs OF How WE LEARN
E IS FOR ELABORATION
What is the core learning mechanic? The process of elaboration involves explicitly connecting new information to
what one already knows. Elaboration increases the chances of remembering the material later.
1s an v.o,nn10 and what is it good for? A student needs to memorize a problem-solving cycle that has the follow
ing elements: identify problems, define goals, explore strategies, anticipate outcomes, look back to learn. An elaboration strategy is to find a way to
connect each of these steps and relate them to one's prior knowledge. One solution is to generate the acronym IDEAL and connect it to the idea of an ideal problem solver.
Why does it work? Human memory is vast. Remembering depends on finding the right memory at the right time. Elaboration makes connections among memories when
learning, so it is easier to find a path to the stored information later. For instance, when asked how a "good" problem solver operates, one might think
good • ideal • IDEAL • identify problems, define goals ....
What problems does the core mechanic solve? Students forget too much.
• They have trouble remembering vocabulary words and their definitions. 0 They cannot remember the steps in a procedure. Teachers cannot remember student names.
• A teacher rereads the student roll sheet but cannot remember who is who.
of how to use it To learn vocabulary words.
· Create a sentence that reflects the meaning of the word precisely. To remember a long speech or a long sequence of actions. 0 Associate each section/step with way stations on a route that one
travels frequently.
To memorize a set of rules. 0 Make up an acronym that summarizes the rules. The acronym
FOIL (first, outer, inner, last) helps people remember the rules for multiplying binomials.
Risks
E IS FOR ELABORATION I 63
Teachers (and videos) may move so quickly that students do not have
time to elaborate. People may fail to identify what they are supposed to remember and
elaborate.
for
Generation
GENERATION IS A memorization technique that relies on the fact that remem
bering something makes it easier to remember the next time.
Flash cards are the canonical example. You read a cue on one side, and you try to remember the target on the other side. With practice, you get better each time through the flash cards. The invention of flash cards surely
belongs in the pantheon of breakthroughs in learning technologies-simple, effective, and available to all.
Flash cards work because of the generation effect. The expression comes
from a famous study in which people learned word pairs (Slarnecka & Graf, 1978). You can get a feel for the study by looking at the three columns of word
pairs below. For the finished pairs, your task is to read each pair silently. In the cases where there are missing letters, you should generate the word. For instance, if you know that the second word is the opposite of the first, and you receive happy : s _ d, the appropriate generation is sad.
SYNONYMS ANTONYMS RHYMES fast: r _p_ d flavorful: bland rain · g_
pain : ache sleep: a __ k_ dime: time
witty:c __ v_ give: t stink: link
jump: leap leave: come graph:I ____
78
G IS FOR GENERATION I 79
If you wait a few minutes and then try to write down all the words (without peeking back), you will likely remember about 25 percent more of the
words that you had to generate compared with words you directly read. For
instance, you should remember rapid and clever better than ache and leap. You are also likely to remember the word rapid better than the word fast, even
though they are part of the same pair. This is because you had to generate rapid, whereas you only read fast.
Generation is very useful for improving recall. It is also very general-it works for memorizing motor movements as well as equations. Knowing some
of its subtleties can help avoid common memorization mistakes by enforcing two simple constraints: make sure that students generate the target memory {not
read it), and space the memorization practice over time (don't cram).
Generation works on the retrieval side of memory. It is not a technique for
encoding, or getting information into one's head (see Chapter E). You already knew all the words in the preceding example. Instead, generation is a tech
nique for making it easier to retrieve the memories. The distinction between
encoding and retrieval is implicit in much classroom instruction. Teaching promotes the encoding of information; homework promotes its retrieval.
Homework typically involves some form of generation whereby people practice remembering the information by accessing it to solve relevant problems or simply through rote practice.
How does generation work? The analogy to strengthening a muscle works well when considering the generation effect. Generation involves practicing
exactly what you need to do in the future-exert effort to retrieve a memory from cues. For instance, try to remember a single word that can go with each
the following words: cottage, cake, Swiss. The words are the cues. The correct memory, which takes some effort to find, is cheese. The successful effort to
find the word strengthens your memory for cheese (lucky you). Now, if we had just told you cottage cheese, cheesecake, and Swiss cheese, you would not have
needed to do any mental heavy lifting, and your memory for cheese would not have improved.
Also like muscles, memories weaken over time with disuse, so they become harder to retrieve. A well-defined forgetting function describes how memories
decay and become more difficult to retrieve. Memories fade the most in the
first few days and weeks after their acquisition but then flatten out to a very slow decay pattern, so they never fully disappear (forgetting follows a power law; see Chapter D).
The time course of memory acquisition influences the forgetting
80 I THE ABCs OF How WE LEARN
Caffeine
cl3 l~ltlal Encoding (~J" ,, '
H_iC O CH3
cl3
1st Practice <::() ' ' 1-1_ic rn3
2nd Practice ~:() ,,
3rd Practice ':() <, "
'""""'"" • Figure G.1. Expanded practice: generating more and
more from memory overtime.
function through an important mechanism called the spacing effect. Spac
ing your memory acquisition over time will make the memory last longer in the future. It is better to practice vocabulary words ten minutes a day
for two days (spaced practice) than to study the words once for twenty minutes (massed practice). Spacing one's practice over two days buys about a 10 percent improvement in memory a month later (Cepeda, Pashler, Yul,
Wixted, & Rohrer, 2006). There are competing explanations for the spac
ing effect, but a simple one is that it takes more effort to retrieve a memory a day later (spaced practice) than one second later (massed practice). The
greater effort creates a stronger memory trace, which in turn takes longer
to decay. The spacing effect helps explain a common experience. Once upon a
time, you must have crammed for a midterm test. You did all the study
ing the night before the test in a big burst of effort. You remembered the material for the test the next day and felt very clever about the whole thing.
But now you cannot remember very much, and you may even have forgot
ten most everything by the time of the final exam. This is because your memory coded the information as something you do not need to retrieve very often, and the memory decayed quickly. Spaced practice solves this
problem.
G IS FOR GENERATION I 81
it How to Uso Generation lo Enhance Learning
Tasks that require retrieving a memory based on a partial cue improves the strength of the memory. There are ways to optimize the generation effect.
The first is to keep the task at a desirable difficulty level (Bjork, 1994). You
want people to succeed at remembering, but you do not want it to be so easy that they do not have to work to remember. One nice approach is to use
expanded practice, where each subsequent round of practice requires remembering all the prior information plus a new piece of information. For example,
witb a deck of flash cards, it works well to do card 1, then cards 1 and 2, then cards 1 and 2 and 3, and so on. Figure G.l provides a concrete example.
The goal is to remember the molecular composition of caffeine. In rhe first round, people study the molecule. Ideally, they would use elaboration to help
encode the molecule; for instance, H shows up three times with three Hs (see Chapter E). When people move to retrieval practice, they only need to
remember and draw a few of the missing atoms and bonds on the first try. In
the second round, they need to remember everything from before plus a new set of removed atoms and bonds. The process of expanding the memory task
continues until finally, given the cue "caffeine," people can remember the full molecular structure.
A second important consideration is that people increase the strength of
the memory trace that they generate but not the cue that triggers the memory. (People remember the target rapid better than the cue fast in this chapter's initial example.) The implication is that sometimes it is important to swap
the cue and the target memory. If you want to remember a definition for a
word, you should use the word as the cue and generate the definition from memory. However, if you want to remember the word given a definition, then
you should use the definition as the cue and generate the word from memory. People often forget that they need to practice in both directions.
The third important consideration is temporal spacing. As described above, it is better to practice over several sessions than to cram it all into one
session. There is also a second temporal issue: sleep. People consolidate their memories during sleep (see Chapter Z). It is a good idea to do memorization
practice, get a good sleep, and then practice again to build on top of the con
solidation. It is another reason that practicing for two days is better than one.
Ill. The Outcomes of Generation
Generation works for all types of memories and tasks, but it is especially useful for memory tasks that require free recall. Free recall refers to tasks where
I,
• 82 I THE ABCs •F How WE LEARN
there is not a strong external stimulus cuing your memory, for instance, if you
are trying to remember what somebody told you in a conversation last month,
or if you have a final exam where you need to write down all the formulas you
can remember from algebra.
One natural implication of the generation effect is that simply taking a test
will improve memory (Karpicke & Blunt, 2011). Taking a test requires retriev
ing memories and therefore improves the accessibility of those memories later,
for example, on a future test. Of course, the implication is not that students
should only take repeated tests; rather, they should practice remembering what
they know. Much school-based instruction focuses on encoding and elabo
rating information to make it meaningful. This is one important side of the
memory equation, but people also need to practice the retrieval side.
IV. Can People Learn to Teach Themselves with Generation?
People learn all sorts of memorization techniques, so it is not difficult to
get them to learn generation strategies. People often think that wanting to
remember something will help. There is little evidence that the desire to
remember, for example, to do well on a test, improves memory (Hyde &
Jenkins, 1973). Instead of wishing themselves to a good memory, people must
invoke deliberate memorization strategies such as generation and elaboration.
V. Risks of Generation
The primary risk of generation is that people can generate the wrong thing,
which will strengthen an "incorrect" memory. A common experience is driv
ing up to an intersection and not remembering whether to turn left or right.
After some deliberation, you take a turn only to realize that (a) it is the wrong
direction, and (b) it is the direction you turned the last time you were at
the same intersection. Blame it on the generation effect! Because you gener
ated the turn last time, you were more likely to remember it again this time.
In fact, by taking the wrong turn yet again, you strengthened the incorrect
memory trace further!
VI. Examples of Good and Bad Use
Imagine you have used your yellow marker to highlight a sentence in a text.
Let us imagine you highlighted Generation works on the retrieval side of mem
ory. When you go hack to study your text, you make a point of rereading the
highlighted sentence. This is a poor, but common, strategy. By rereading the
sentence, you are not practicing remembering, because it is right there for you
,
G IS FOR GENERATION I 83
to read. It would be much better if you tested your memory by only reading
the firsr half of the sentence and generating the rest from memory: Gener
ation works on the .... Perhaps people would remember more if they only
highlighted part of tbe sentence and used that as a cue to remember the rest.
Bjork, R. A. (1994). Memory and metamemory considerations in the training of
human beings. In J. Metcalfe and A. Shimamura (Eds.), Metacognition: Knowing
about knowing (pp. 185-205). Cambridge, MA: MIT Press.
Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed
practice in verbal recall tasks: A review and quantitative synthesis. Psychological
Bulletin, 132(3), 354-380.
Hyde, T. S. & Jenkins, J. J. (1973). Recall for words as a function of semantic,
graphic, and syntactic orienting tasks. Journal of Verbal Learning and Verbal
Behavior, 12(5), 471-480.
Karpicke, J. D., & Blum, J. R. (2011). Retrieval practice produces more learning
than elaborate studying with concept mapping. Science, 331, 772-775.
Slamecka, N. J., & Graf, P. (1978). The generation effect: Delineation of a phenom
enon. journal of Experimental Psychology: Human Learning and Memory, 4(6),
592-604.
84 I THE AB Cs OF How WE LEARN
G IS FOR GENERATION
What is the core learning mechanic? Practicing the retrieval of target memories given partial cues or hints improves future retrieval.
IS an nv,,mn!n and what is it good for? Flash cards are the original example. On one side a card says jocund, and on the other side it says, cheerful and lighthearted. To work on memorizing the
definition, people read the vocabulary word (the cue) and practice generating the definition (the target) without looking at the other side. This improves
memory for the definition. If people simply flip the card over to read the definition, instead of trying to remember the definition first, there will be little
improvement in memory.
does it work? Retrieving a memory increases the strength of the memory, so it is easier to retrieve later. Spreading out memorization practice over several days increases memory strength compared with memorizing in only one session.
What problems does the core mechanic solve? Students have trouble memorizing arbitrary facts and conventions. " They cannot remember correct spelling.
" They cannot remember the names of the presidents. Students have trouble recalling information without strong reminders. " They cannot remember the definition of a word without giving
them multiple hints.
Students forget too easily.
" Students do well on a weekly test but forget the information for the final exam.
LAaii<1J1~, of how to use it To learn vocabulary words and their definitions. " Use flash cards going both directions. Given the word, remember
its definition. Given a definition, remember the word. To remember an organic molecule. " Show all of the molecule except a couple of atoms and bonds, and
ask students to remember the missing pieces. Show the molecule again, but remove additional atoms and bonds. Ask students to
Risks
G IS FOR GENERATION ! 85
remember all the missing pieces. Continue until the students can remember the complete molecule when only hearing its name.
People may generate the wrong thing, which will strengthen the "incorrect" memory.
People tend to read the answer before trying to generate it, which under
mines the effect.