LongTerm Memory

Percentage of items recalled
216
Chapter 6
100
90
80
70
60
50
40
30
20
10
3
6
9
12
15
18
Recall interval (in seconds)
Source: Data from Peterson & Peterson (1959).
FIGURE 6.4
Forgetting in Short-Term Memory
This graph shows the percentage of items
recalled after various intervals during
which rehearsal was prevented. Notice
that forgetting was virtually complete
after a delay of eighteen seconds.
Brown-Peterson procedure A method
for determining how long unrehearsed
information remains in short-term
memory.
long-term memory (LTM) The stage of
memory for which the capacity to store
new information is believed to be
unlimited.
Memory
How long does information remain in short-term memory if you don’t keep rehearsing it? John Brown (1958) and Lloyd and Margaret Peterson (1959) devised the BrownPeterson procedure to measure the duration of short-term memory when no
rehearsal is allowed. In this procedure, participants are presented with a group of three
letters, such as GRB. They then count backward by threes from some number until they
get a signal. Counting prevents the participants from rehearsing the letters. At the signal, they stop counting and try to recall the letters. By varying the number of seconds
spent counting backward, the experimenter can determine how much forgetting takes
place over time. As you can see in Figure 6.4, information in short-term memory is
forgotten rapidly: After only eighteen seconds, participants can remember almost nothing. Evidence from other such experiments also suggests that unrehearsed information
can be held in short-term memory for no more than about eighteen seconds. However,
if the information is rehearsed or processed further in some other way, it may be
encoded into long-term memory.
Long-Term Memory
When people talk about memory, they are usually referring to long-term memory.
Long-term memory (LTM) is the part of the memory system whose encoding and
storage capabilities can produce memories that last a lifetime.
Encoding in Long-Term Memory Some information is encoded into long-term
memory even if we make no conscious effort to memorize it (Ellis, 1991). However,
putting information into long-term memory is often the result of more elaborate and
conscious processing that usually involves semantic coding. As we mentioned earlier,
semantic encoding often leaves out details in favor of the more general, underlying
meaning of the information.
In a classic study, Jacqueline Sachs (1967) demonstrated the dominance of semantic encoding in long-term memory. Her participants first listened to tape recordings of
people speaking. She then showed them sets of similar sentences and asked them to say
which contained the exact wording heard on the tape. Participants did well at this task
when tested immediately, using mainly short-term memory. However, after twentyseven seconds, they couldn’t be sure which of two sentences they had heard. For example, they could not remember whether they had heard “He sent a letter about it to
Galileo, the great Italian scientist” or “A letter about it was sent to Galileo, the great
Italian scientist.” They didn’t do as well after the delay because they had to recall information from long-term memory, where they had encoded the general meaning of what
they had heard, but not the exact wording.
Perhaps you are thinking, “So what?” After all, the two sentences mean the same thing.
Unfortunately, when people encode the general meaning of information they hear or
read, they can make mistakes about the details (Brewer, 1977). This can be a problem
when recalling exact words is important—such as in the courtroom, during business
negotiations, and in discussions between students and teachers about previous agreements. Later in this chapter we show that such mistakes occur partly because people
encode into long-term memory not only the general meaning of information but also
what they think and assume about that information (McDermott & Chan, 2006).
Counterfeiters depend on the fact that people encode the general meaning of visual
stimuli rather than specific details. For example, look at Figure 6.5, and find the correct
drawing of a U.S. penny (Nickerson & Adams, 1979). Research shows that most people
from the United States are unsuccessful at this task. People from other countries do poorly
at recognizing their country’s coins, too (Jones, 1990). This research has prompted the U.S.
Treasury to begin using more distinctive drawings on the paper currency it distributes.
The capacity of long-term memory
is extremely large. In fact, many psychologists believe that it is literally unlimited
(Matlin, 1998). There is no way to prove this, but we do know that people store vast
Storage Capacity of Long-Term Memory
217
Storing New Memories
WE
TR
U
doing
2
learn
by
G
T
OD
WE
TR
U
N
I
LIBERTY
Which is the correct image of a
U.S. penny? (See p. 218 for the
answer.)
S
LIBERTY
(A)
S
G
T
OD
WE
TR
U
N
LIBERTY
(B )
S
G
T
OD
WE
TR
U
N
S
T
I
OD
I
G
N
I
FIGURE 6 .5
Encoding into Long-Term Memory
LIBERTY
(C )
(D)
(E)
Source: Nickerson & Adams (1979).
quantities of information in long-term memory that can be remembered remarkably
well after long periods of time. For example, people are amazingly accurate at recognizing the faces of their high school classmates after having not seen them for over
twenty-five years (Bruck, Cavanagh, & Ceci, 1991). They also do surprisingly well on
tests of a foreign language or algebra fifty years after having formally studied these subjects (Bahrick & Hall, 1991; Bahrick et al., 1994).
But long-term memories are also subject to distortion. In one study illustrating this
point, students were asked to describe where they were and what they were doing at
the moment they heard about the verdict in the O. J. Simpson murder trial (Schmolck,
Buffalo, & Squire, 2000). The students first reported their recollections three days after
the verdict and then again after either fifteen or thirty-two months. Only half the recollections reported at fifteen months were accurate, and 11 percent contained major
errors or distortions. Among those reporting after thirty-two months, 71 percent of
their recollections were inaccurate, and just over 40 percent contained major errors or
distortions. For example, three days after the verdict, a student said he heard about it
while in a campus lounge with many other students around him. Thirty-two months
later, this same person recalled hearing the news in the living room of his home with
his father and sister. Amazingly, most of the students whose memories had been greatly
distorted over time were unaware of the distortion; they were very confident that the
reports were accurate. Similar findings have been reported in relation to people’s memories of the September 11, 2001, terrorist attacks on New York and Washington, D.C.
Even when these memories changed as time went by, people remained confident in
their accuracy (Talarico & Rubin, 2003). Later, we see that such overconfidence can also
appear in courtroom testimony by eyewitnesses to crime.
doing
2
learn
A REMARKABLE MEMORY
Using only his long-term memby
ory, Franco Magnani created
amazingly accurate paintings of his
hometown in Italy even though he had
not seen it for more than thirty years
(Sacks, 1992). People like Magnani display
eidetic imagery, commonly called photographic memory. About 5 percent of
school-age children have eidetic imagery,
but it is extremely rare in adults (Haber,
1979). You can test yourself for eidetic
imagery by drawing a detailed picture or
map of a place that you know well but
have not seen recently. How did you do?
Magnani's painting
Photo of the same scene