Transcript Cognitive Science - University of California, Irvine

Cognitive Science
Jose Tabares
Psych 202B
January 23, 2006
Creating False Memories:
Remembering Words
Not Presented in Lists.
Henry L. Roediger III
and
Kathleen B. McDermott
Rice University
Journal of Experimental Psychology:
Learning, Memory, and Cognition
1995, Vol 21, No.4, 803-814
False Memories
 Remembering events that never
happened.
 Or, remembering them quite differently
from the way they happened.
Reconstructing Memories
 Sir Frederic Bartlett, a prominent English
psychologist.
 Had his subjects read “War of the Ghosts”
 Subjects read the story twice and waited
15 minutes.
 Then they were asked to write down the
tale as best they could recall it.
War of the Ghosts
 So what did Bartlett find?
 Subjects frequently changed the story.
 Bartlett concluded that the subjects
reconstructed the story to fit with their
established schemas.
 Reproductive memory
 Reconstructive memory
Elizabeth Loftus
Misinformation Effect
Definitions
 Hits (Hit) – number of words from List A
recognized (out of 15)
 Correct Rejections (CR) – number of words
not on List A that were correctly identified as
not being on the list.
 Misses (Miss) – number of words on the list
that were not recognized.
 False Alarms (FA) – number of words
indicated as being on the list that were, in
fact, not.
Roediger and McDermott
 Two experiments
 Modeled after J. Deese’s 1959 study.
 Sweet: sour, candy, sugar, bitter, good,
taste, tooth, nice, honey, soda, chocolate,
heart, cake, tart, and pie.
 First recall test
 Then recognition test.
Summary
 Studies of memory suggest that people
can have the subjective experience of
remembering an event that never actually
occurred. This happens either when
people mistakenly associate a memory
with the wrong source, or when they have
seen so many things consistent with the
event that the event that did not happen is
somehow represented as well.
Contextual Prerequisites for
Understanding:
Some Investigation of
Comprehension and Recall
John D. Bransford and Marcia K. Johnson
State University of New York, Stony Brook
Journal of Verbal Learning and Verbal Behavior, 11, 717-726.
Tree structure of a sentence
Sentence
Noun phrase
Adjective
Visiting
Verb phrase
Noun
Relatives
Modal
can
Verb
Noun phrase
article
noun
a
nuisance
be
Comprehension
 Involves the recovery and interpretation of the
abstract deep structural relations underlying
sentences.
 One type of error occurs because the act of
comprehending sentences often includes
plausible inferences, and the results of these
inferences may be indistinguishable in memory
from information actually given (Bransford,
Barclay, and Frank, 1972).
Example
 Three turtles rested on a floating log and a
fish swam beneath it.
 Subjects create semantic products that are
a joint function of input information and
prior knowledge.
Experiment 1
 “Would subjects who receive appropriate
prerequisite knowledge be able to
comprehend the passage quite easily, and
hence would subsequently be able to
recall it relatively well.”
 Acquisition phase followed by two tasks—
comprehension rating and recall.
Methods
 Five groups
-No Context
-No Context 2
-Context After
-Partial Context
-Context Before
Results
 Context Before Group had the best recall
and rated the passage easier to
comprehend than the other groups.
Experiments 2
 Experiment 2 heard one a shorter version
of a passage on washing clothes.
 Three groups
-No Topic
-Topic Before
-Topic After
Experiment 3
 Experiment 3 heard a passage similar but
longer than the groups in Experiment 2.
 Only two groups
-Topic Before
-Topic After
Experiment 4
 The groups in Experiment 4 heard a
passage on making a kite
 Again, three groups
-No topic
-Topic before
-Topic After
Results
 Both in Experiment 2 and 3, the Topic
Before groups rated the passage easier to
comprehend than the other group(s).
 Also, then showed better recall.
 Similar results occurred in Experiment 4.
Discussion
 Prior knowledge for a situation does not
guarantee its usefulness for
comprehension.
 For maximum benefit, the appropriate
information must be present during the
ongoing process of comprehension.