Bruce, D., Phillips-Grant,
K., Conrad, N., & Bona, S. (2004). Encoding context and false recognition memories. Memory, 12 (5), 562-570.
Using the DRM paradigm, the
authors focused on the activation and source monitoring theory, which states that
participants make inaccurate judgments as to whether a critical word was
activated externally or internally. The
authors wanted to make source monitoring easier for participants by providing
either visual or verbal information at study to attempt to reduce false
memories. They used a within-subject
design to determine whether a distinctiveness
heuristic could provide reference when there is extra context information
available in some lists but not others.
Experiment 1
Eight lists of nine words
each were presented in block format with no pauses at the completion of each
individual list. Participants were told
that some words would have a distinctive visual context and that they should
try to use this to help them remember the words. They were also told that the cues would be
present at test (even though they were not).
Four of the eight sublists were selected to
appear with one of nine visual contextual cues: a border around the word,
background color, different font, a small picture, patterned background,
quadrant where the word was located, lower case font, a colored stripe, or
underlining the word with characters.
All nine dimensions occurred within each sublist. No more than two sublists
with or without visual contexts ran in succession. The test list consisted of 3 targets, the
critical lure, 2 weak lures, and 1 unrelated lure for each of the eight sublists. After list
presentation, participants were to circle old or new, then remember or know for
old answers. They then reported whether
each old item had been presented with a visual context, and if so, what the
context was.
When a visual context was
present, false alarms to critical and weak lures were reduced, but
identification of targets was unaffected.
Visual context reduced the remember judgments for critical and weak
lures that were called old, but not for targets. There was no significant difference in know
judgments for critical lures and targets, but there were more know judgments
for weak lures when there was no visual context. The proportion of times participants correctly
decided a target was presented with a visual context was .43, but they
correctly identified that context only .23 of the time.
Experiment 2
Experiment 2 was the same as
Experiment 1 except for the type of contextual information offered. List words were presented in upper case with
another word to be used as the verbal context cue in lower case to the
left. The verbal context did not
semantically relate to the word with which it was presented. Participants had to say the word pair or the
context word aloud to ensure they paid it attention.
Identification of targets
declined when they were presented with the context word. False alarms to critical
lures, but not weak lures, was reduced.
Verbal context also reduced the number of remember judgments on critical
lures and targets said to be old, but not on weak lures. Know judgments were not affected. The proportion of times participants
correctly identified that an item had been presented with a context cue was
.44, but this time, they correctly identified the cue only .06 of the time.
Discussion
The authors explain the
results of both experiments in terms of the source monitoring and activation
theory. In Experiment 1, the results
emphasized the source monitoring part of the process. Because there was extra visual context,
participants spent more time encoding (and monitoring) the contextual aspects,
thus not activating the critical lure as much.
The results of Experiment 2 emphasize the activation component. Because subjects were encouraged to develop a
relationship between the non-semantically related cue word,
this disrupted meaningful relationship between the target words, not activating
the critical lure.
The results also indicate
that the distinctiveness heuristic can operate when some lists receive extra
cues but not others. Subjects were
reasonably accurate at assessing whether words appeared with verbal or visual
cues.