Brainerd, C. J., & Wright, R. (2005). Forward association, backward
association, and the false-memory illusion. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 31, 554-567.
The authors briefly review research on the DRM paradigm
and other literature on semantic false memory. Results of many of the studies
on semantic false memory indicate that paradigms in which the meaning of the
studied items is preserved tend to produce false recall or false recognition.
Increases in semantic overlap (analogous to mean forward associative strength)
produces increases in false-memory responses. This same correlation has not
been found in the DRM paradigm, which does show a relation between MBAS and
false recall and recognition. Further, the authors argued that the lack of a
MFAS-false memory relation may be an artifact of a restriction in range of the
MFAS strength. Thus the goal of the present study was to examine this situation
in more detail.
Two reasons, one theoretical and one strategic were
proffered for the experiment. Strategically, the DRM paradigm is very widely
used, and it would be reasonable to determine if it is representative of the
larger class of false-memory paradigms, or if it is an anomaly. Theoretically,
they felt that different theoretical approaches depended on finding either
differential effectiveness of MBAS and MFAS, or similar effectiveness. In
particular, FTT requires that MBAS and MFAS have similar effects on false
memory production.
The authors manipulated MBAS and MFAS independently by
choosing sublists from the DRM lists in which they could independently
manipulate MBAS and MFAS. They also utilized a conjoint-recognition paradigm to
be able to estimate the parameters specified by the FTT model.
Method
Subjects.
Subjects were 170 out of 180 undergraduates who participated to fulfill
a course requirement. Of the original 180 subjects, sixty were randomly
assigned to each of the three conditions, V, G, VG, specified by the
conjoint-recognition paradigm.
Materials. Sixteen DRM lists were chosen for use in the
study, and from each of the 15-item lists the authors developed 4-item sublists
which fit one of the four conditions from the factorial combination of MBAS (high,
low) and MFAS (high, low).
Procedure. Subjects were presented an audio recording of
the 16 lists at a 3-s rate per word, and with a 10-s pause between sublists.
Following the last list, subjects had 2 min to read the conjoint-recognition
instructions. These instructions were detailed, and contained examples of the
types of items that would be presented. In V conditions subjects were told to
select only targets; in G conditions they were told to select semantically
related distractors only, while those in the VG condition were told to select
both targets and semantically related distractors.
The test list was composed of 128 words, and consisted of
32 targets, two from each of the 16 DRM lists; 16 critical distractors, one
from each list; 32 semantically related distractors; 48 unrelated distractors.
Results
Critical Distractors. Both FAS and BAS both affected the
probability of accepting critical distractors in all three instruction
conditions. In the V condition the main effects of BAS and FAS were both
significant, with high levels of each producing more false acceptances. In the
G condition the BAS X FAS interaction was significant; when BAS was low, FAS
increased false acceptance rates. When FAS was low BAS increased acceptance
rates. Finally, in condition VG the BAS x FAS interaction was also significant.
FAS had no effect when BAS was high, but increasing FAS led to increased
acceptances when BAS was low.
The authors pointed out that the results in the V
condition were contrary to previous correlational findings involving BAS and
FAS and false alarms for critical distractors.
Targets. In
condition V the BAS x FAS interaction was significant with the effect of
increasing on variable only being manifest when the other variable was low. In
condition G, neither factor was significant, and condition VG produced the same pattern of
results as in condition V.
Model analyses.
Analyses using the parameter estimates based on the FTT model were done
to examine the processes-level effects of BAS and FAS. The authors concluded
that both BAS and FAS had broad effects with both FAS and BAS influencing both suppressive and
supportive processes. They concluded
that both BAS and FAS had their primary effects by influencing subjects’
ability to use memory for surface details of target presentations to reject
critical distractors, i.e. through recollection rejection. They also posited a
new finding for “erroneous recollection rejection.”
Important Legal Disclaimer: The preceding are articles we read
together in the Lampinen Lab Fall 2005 false memory reading group. By clicking
on the button next to the 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,
Hugo's, 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).