224 • CHAPTER 8 Everyday Memory and Memory Errors
MPI as Replacing the Original Memory Loftus explains the misinformation effect by
proposing the memory trace replacement hypothesis, which states that MPI impairs
or replaces memories that were formed during the original experiencing of an event.
According to this idea, seeing a stop sign creates a memory trace for a stop sign, but
presentation of MPI that a yield sign was present causes the memory for the stop sign
to be replaced by a new memory for a yield sign. The process of reconsolidation, which
we described in Chapter 7 (page 197), could provide a physiological mechanism for
this replacement. According to the idea of reconsolidation, reactivating a memory can
create the potential for forming new memory traces.MPI as Causing Interference Another explanation proposes that the original information
is forgotten because of retroactive interference, which occurs when more recent learning (the
misinformation in this example) interferes with memory for something that happened in the
past (the actual event). For example, retroactive interference would be involved if studying
for your Spanish exam made it more diffi cult to remember some of the vocabulary words
you had studied for your French exam earlier in the day. This explanation is similar to the
memory trace replacement hypothesis in that the new information affects the old informa-
tion. However, in this case, the old information isn’t eliminated; it is simply interfered with.MPI as Causing Source Monitoring Errors Another explanation for the misinfor-
mation effect is based on the idea of source monitoring, which we discussed earlier.
According to source monitoring, a person incorrectly concludes that the source of his
or her memory for the incorrect event (yield sign) was the slide show, even though
the actual source was the experimenter’s statement after the slide show. The following
experiment by Stephen Lindsay (1990) investigated source monitoring and MPI by
asking whether participants who are exposed to MPI really believe they saw something
that was only suggested to them. The answer to this question would be “yes” if the
participant is making a source monitoring error.
Lindsay’s participants fi rst saw a sequence of slides showing a maintenance man
stealing money and a computer. This slide presentation was narrated by a female
speaker, who simply described what was happening as the slides were being shown.
Two days later, participants returned to the lab for a memory test. Just before the test,
they listened to a story, without slides, by the same female speaker. This story was simi-
lar to the one they had heard and seen 2 days earlier, but with a few details changed. For
example, a pack of Marlboro cigarettes in the original became Winstons in the retelling
of the story, and a can of Maxwell House coffee became Folgers.
Before the participants heard the second telling of the story, they were informed
that there were some incorrect details in it, so they should ignore what they heard
in the second story when taking the memory test. In the memory test, participants
were asked questions such as “The man had a pack of cigarettes. What brand
of cigarette was shown in the slides?” Three of the questions were about misled
items, for which they had received incorrect information in the second story, and
three were about control items, for which they had received correct information.
The results, shown in ● Figure 8.17a, indicate that for the misled items,
27 percent of the responses corresponded to the incorrect information in the sec-
ond story. This compares to only 9 percent of incorrect responses for the control
items. These responses to the misled items would be source monitoring errors if
the participants were confusing the information from the second story with the
information from the fi rst story.
The results for another group of participants, who heard a male voice tell
the second story, are shown in Figure 8.17b. In this case, misled items received
only 13 percent incorrect responses, which was not signifi cantly different from the
10 percent incorrect responses for the control items. This lack of source monitor-
ing errors occurred because the male voice was different from the female voice, so
it was easier to distinguish which information came from which story. Thus, using
the same female voice for both stories created source monitoring errors that led
participants to believe they had seen something that they hadn’t seen.Control items“Misled” itemsPercent “suggested”responses3020100
MC
(b) Male voiceMC
(a) Female voice
●FIGURE 8.17 Results of Lindsay’s
(1990) source monitoring experiment.
(a) Eff ect of misleading information
provided by the female voice on
responses to misled (M) and control
(C) items. (b) There was no diff erence,
compared to control, when the male
voice presented misleading information.Copyright 2011 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part.