Gallo, D.A. & Seamon,
J.G. (2004), Are nonconscious processes sufficient to
produce false memories?”
Introduction:
Seamon, Luo, and Gallo (1998) found evidence for false memories in
the DRM paradigm when conscious processes were minimally involved. They restricted conscious processing through
extremely brief presentation rates (20 ms) with a simultaneous cognitive
load. Hit rates were only 13% above base
rate false alarms to control lures, but false alarms to critical lures was .23
with the same correction (these differences were significant). Based on this evidence, they argued that conscious
processes were not necessary for the
false memory effect in the DRM list paradigm.
Zeelenberg,
Plomp, and Raaijmakers
(2003) challenged Seamon et al.’s results, however,
arguing that their methodology was flawed due to the use of a monitor whose
refresh rate that led to an inconsistent presentation rate about the target
rate of 20ms. Also, they interpreted the
argument posed by Seamon et al. as claiming unconscious
perception, though the authors did not claim such. They argued instead that conscious processing of the word list was minimal,
not conscious perception, and that the processing was not necessarily nonexistent. Zeelenberg et al. failed to replicate the results and
argued the null results showed that Seamon et al.’s
argument lacked experimental support.
Gallo & Seamon challenge this by arguing
that the results are ambiguous and do not counter the original claims of Seamon et al. They
accept that the issue needs to be readdressed, a propose
a second experiment with the following key questions:
1.
Is conscious
activation of the related lure during study necessary to cause subsequent false
recognition (via a source monitoring error)?
2.
Is conscious
recollection of the list words at test necessary to cause subsequent false
recognition?
Method:
Subjects: 28 Harvard undergraduates participated for $10 or
course credit
Materials: 36 DRM lists, 15 items each divided into 2 lists (A
and B) counterbalanced across subjects.
List presentation order was constant.
18 recognition test trials consisting of 1 critical lure from each
presented DRM list and 1 critical lure from each non-presented list in a
forced-choice format. These were
randomly presented to each subject and position of targets was counter-balanced
(right vs. left).
Procedure: Subjects were presented each list on computer screen
with words in descending associative strength at presentation duration of 20ms
per word. After each list they were
instructed to write down as many words as they could remember. After all lists were recalled, the surprise
forced-choice recognition test was given.
Subjects were told that one word of the pair had been presented and one
had not. They had to indicate which had
been.
Results:
Immediate
perception/recall data were liberally scored to ensure any semantically related
variant of a list item would be counted.
This still showed an extremely low recall rate of .07. Recall of first list items was highest at a
rate of .29. Recall of the related lure
occurred at a rate of .03.
Recognition
of the related lures was significantly greater than chance
at a rate of .58. Excluding those trials
where the related lure was immediately recalled, recognition of the related
lure was still recognized significantly more often than chance at a rate of
.57.
Recognition
of related lures was significantly greater than chance
when at least one item was perceived/recalled at test (.66) than when none was
(.49). There was a trend that the more
words recalled the higher the likelihood that the related lure would be
recognized.
Conclusions:
The
results were consistent with Seamon et al. (1998),
showing that conscious processes are not necessary for the false memory effect
to occur in the DRM paradigm. Further,
this study shows that the results of Seamon et al.
(1998) cannot be explained through rehearsal of those list items that were
recalled because the related lures were tested on a surprise test. Unlike Seamon et
al., the current study showed no recognition of related lures when there was no
memory for list items. There was some
minimal processing of the lists, and this led to the false recognition
effect. The authors propose two possible
interpretations: (1) recognition of related lure is dependent on the subsequent
recollection of at least some of the list items; (2) additional processing of
the list items boosts the nonconscious activation of
the related lure. The authors also place
their findings in the body of evidence regarding the automatic processes
involved in false memories that are not under conscious awareness or conscious
control.