Brainerd, C. J. , Reyna,
V.F> and Estrada, S. (2006). Recollection Rejection of False Narrative
Statements
This article concerns Brainerd and Reyna’s conception
of recollection rejection. Recollection
rejection occurs when verbatim recall of a studied item decreases false
memories for gist consistent lures. The
article examined five factors that theoretically should influence recollection
rejection. These were (1) effects of
repetition, (2) effects of recency of the study
materials, (3) effects of long time delays, (4) effects of degree of semantic
and surface overlap between the related lures and the studied items and (5)
effects of repeated testing.
These issues were examined using a narrative memory
procedure. Each narrative involved three
sentences. The first two sentences (premise sentences) set up relations between
three objects. The third sentence was a filler that mentioned one of the objects but not the
relation. For example:
The coffee is hotter than
the tea (Premise 1)
The tea is hotter than the
cocoa (Premise 2)
The cocoa is sweet
(Filler)
Participants then took a recognition test with
targets, related lures, and unrelated lures.
The related lures varied in terms of the degree of semantic and surface
overlap with the studied items.
For the above example, test items would include:
Targets:
The coffee is hotter than
the tea TPO1 TRUE
PREMISE WITH OLD WORDS
The tea is hotter than the
cocoa TPO2 TRUE
PREMISE WITH OLD WORDS
The cocoa is sweet (Filler
target)
Gist Consistent Distractors:
The cocoa is cooler than
the tea (Gist consistent lure) TPN – TRUE PARAPHRASE WITH NEW WORD
The coffee is hotter than
the cocoa (Gist consistent lure) TIO – TRUE INFERENC OLD WORDS
The cocoa is cooler than
the coffee (Gist consistent lure) TIN – TRUE INFERENCE NEW WORD
Gist Inconsistent Distractors:
The cocoa is hotter than
the tea. (Control lure) FPO – FALSE PARAPHRASE OLD WORDS
The tea is cooler than the
cocoa. (Control lure) FPN – FALSE PARAPHRASE NEW WORD
The cocoa is hotter than the coffee.(Control lure) FIO
– FALSE INFERENCE OLD WORD
The coffee is cooler than
the cocoa (Control lure) FIN – FALSE INFERENCE NEW WORD
To estimate recollection rejection, the conjoint
recognition model was used. Conjoint recognition estimates a set of underlying
mental processes that can lead to true or false recognition. The memory processes estimated by conjoint
recognition are:
Identity Judgments for Targets (It): When presented with a target on a recognition memory
test, participants retrieve a verbatim trace that is identical with the test
item.
Erroneous Recollection Rejection of
Targets (Rt): When presented with a target on a recognition
test, participants experience a phantom recollection (i.e., compelling false
memory) for a related item, leading to an experience of non-identity.
Similarity Based Acceptances of Targets (St): When presented with a target on a recognition memory
test, participants retrieve a gist trace is similar in meaning with the test
item.
Phantom Recollection of Related Lures (Pr): When presented with a related lure on a recognition
memory test, participants experience an illusory recollection leading them to
confidently conclude that that exact item was studied.
Recollection Rejection of Related Lures
(Rr): When presented
with a related lure on a recognition test, participants experience a
recollection of an item that actually was studied, leading to an experience of
non-identity.
Similarity Based Acceptances of Related Lures (Sr): When
presented with a related lure on a recognition memory test, participants
retrieve a gist trace is similar in meaning with the test item.
Conjoint recognition estimates these memory processes
using a multivariate processing tree model that uses three sets of
instructions. In addition to the memory parameters, response bias parameters
are estimated for each set of instructions.
Verbatiim Instructions: Participants are told to only say “YES” when an item they studied is
presented. Under these conditions the probability of responding YES to targets,
related lures, and unrelated lures are provided by the following equations:
P(YES |TARGET) =
It + (1-It)(1-Rt)bv
P(YES | RELATED) = (1-Rr)Pr + (1-Rr)(1-Pr) bv
P(YES | RELATED) = bv
Gist
Instructions: Gist instructions tell
participants to say YES only when a gist preserving lure is presented and to
say NO when a target is presented or an unrelated lure is presented. Under
these conditions the probability of responding YES to targets, related lures,
and unrelated lures are provided by the following equations:
P(YES |TARGET) = (1-It)Rt
+(1-It)(1-Rt)
bg
P(YES | RELATED) = Rr + (1-Rr)(1-Pr) bg
P(YES | RELATED) = bg
Verbatim
plus Gist Instructions: The VG
instructions tell participants to say YES both to targets and gist preserving
lures and to say NO when a t an unrelated lure is presented. Under these
conditions the probability of responding YES to targets, related lures, and
unrelated lures are provided by the following equations:
P(YES |TARGET) = It + (1-It)Rt +(1-It)(1-Rt) bvg
P(YES | RELATED) = Rr + (1-Rr)Pr +
(1-Rr)(1-Pr) bvg
P(YES | RELATED) = bvg
These equations specify nine
empirical probabilities which can be used to estimate the model’s
parameters. For the present research the
most relevant parameter being estimated is recollection rejection of related lures
(Rr).
The Experiments
Participants were presented
with narratives like the one listed above. The narratives were presented in
pairs. For some subjects, the narratives
were repeated twice. For other subjects the narratives were only presented once. For other subjects, either the first member of the
pair or the second member of the pair was presented twice. After each pair, participants took a
recognition test on one of the members of the pair. This was repeated for six pairs of
stories. One week later, participants
returned and took recognition tests on the narratives that hadn’t previously
been tested. Each test consisted of the
10 item types identified above. And participants responded to these tests after
being given verbatim, gist or verbatim plus gist instructions.
Central Findings
Immediate recognition testing….
·
Across
conditions there were very high rates of recollection rejection
·
There was more recollection rejection for the second
narrative than for the first narrative
·
There was more
recollection rejection for narratives that were presented twice than for
narratives that were presented once
·
Recollection
rejection was greater for the inferences that involved new words (TIN) than for
the inferences that contained only old words (TIO) and for the paraphrases that
contained new words (TPN). This makes
sense because the TIN narratives are different from what was experienced in two
ways, that exact proposition was never explicitly stated and a word was used
that hadn’t previously been used
·
Phantom
recollection was higher for the second narrative as was related lure
similarity.
Delayed recognition testing….
·
For the
previously untested narratives, recollection rejection was much lower after the
delay
·
Recollection
rejection was greater for the second narrative and there was more recollection
rejection for the inferences that involved new words than for the paraphrases
that contained new words and for the inferences that contained old words
·
For the
previously tested untested narratives phantom recollection and similarity
dropped to very low levels.
·
Prior testing
served to inoculate true memory processes.
For instance, for the narratives that were being tested for the second
time, the estimates of recollection rejection were much larger than for the
previously untested narratives. The same thing held true for identity
judgments. Similarity judgments also increased. The only downside of prior
testing was that erroneous recollection rejection also increased.
DISCUSSION
Argued that the variability in verbatim was mainly a
matter of accessibility at retrieval not storage
Argued that dual process account provides best
explanation
Consistent with recollection
rejection not the distinctiveness heuristic.
Important Legal
Disclaimer: The preceding are articles we read together in the Lampinen Lab
Fall 2006 false memory reading group. By clicking on the authors’ names of each
article you can see the summary of that article. The summary was prepared by
the student presenting that article and it is of course the case that the views
expressed in the summary do not necessarily represent the views of the reading
group as a whole, Dr. Lampinen, the Lampinen Lab, the University of Arkansas,
the Razorback Football or Basketball teams (although we're not sure about cross
country), people living down the street from us, Bob Dylan, Jack Fate, our
extended families, or anyone else for that matter except for the student who
wrote the summary (and they don't necessarily believe what they wrote either).