False
Memories: Young and Older Adults Think of Semantic Associates at the Same Rate,
but Young Adults Are More Successful at Source Monitoring, Dehon and Bredart
(2004)
Several
studies have shown that older adults are more prone to memory distortions than
younger adults, namely higher rates of false recall/recognition, even when
rates of correct responses are similar.
The DRM paradigm used in the study has been shown to illicit high rates
of false recall and false recognition.
False memories are thought to occur due to spreading activation in the
associate network leading to increased accessibility of the CL. To make a correct rejection of the CL, the
subject must correctly identify this accessibility, or familiarity, with their
own thoughts and not the item’s presentation on the study list. Another reason the CL may not be endorsed,
however, is a failure during study to activate the CL at all.
Bredart (2000) created a modified DRM paradigm to assess whether a CL
rejection was due to proper source monitoring or failed activation. In this setup, subjects are asked to add any
words they thought of recording on the initial recall phase but had not. Failure to produce a CL on both the initial
recall and added phase can be interpreted as failed activation, while
production of the CL on the latter phase only can be seen as proper source
monitoring.
Prediction:
Older
adults are expected to recall fewer studied items and more CL’s during initial
recall task than younger adults; older adults will be less likely to recall
CL’s during added phase than younger adults; sum of CL’s from both recall
portions should be equal for younger and older adults
Participants:
30
younger adults (mean age 22.57) and 30 older adults (mean age 70.67); even
proportions of men and women
Procedure:
In
this part of the study, participants are subjected to 6 phases:
1. (Study Phase) S’s are read a
list of words from a French-form DRM list (six lists, 15 items each) at a rate
of 1 word/1.5 sec. After they were read
one list, they then counted backward by 3’s for 30 sec., then
moved on to the next phase.
2. (Phase 1) Participants are
asked to write down all the words they can remember from the list. They have 90 sec. to complete this.
3. (Phase 2) Participants rated
how confident they were about the recall of each word on a 1-5 scale where the
larger the number, the more confident they are that the word HAD been presented
by the experimenter.
4. (Phase 3) After repeating
1-3 for all 6 word lists, the subjects are asked to add to each of the lists
any other words they had thought of during recall but had not written down.
5. (Phase 4) Each of these
added words was rated on a scale of 1-5 with the higher number indicating
greater confidence the experimenter had NOT produced the word.
6. (Phase 5) Each participant
fills out a French form of the Mill-Hill Vocabulary Scale
Results:
Phase
1—Initial Recall
1. Participants in both groups
recalled higher proportions of studied items than CL’s
2. Older adults recalled fewer
studied items than younger adults
3. Proportions of studied items
recalled were significantly higher than proportion of CL’s in younger adults,
whereas they were equal in older adults
4. Older participants assigned
globally higher confidence ratings in their responses than younger participants
5. Confidence ratings assigned
to studied items were significantly higher than for CL’s
6. Younger adults assigned
higher confidence ratings to studied items than CL’s, whereas in older adults
these ratings did not differ
Phase
3—Added Recall
1. Older adults recalled fewer
CL’s during added phase than younger adults
2. Both younger and older
adults were highly confident CL’s had not been presented by experimenter
3. Total proportion of
activated CL’s across Phase 1 and Phase 3 is equal for younger and older adults
4. Younger adults showed
greater recall of CL’s on Phase 3 than Phase 1, while older adults had the
reverse pattern
Experiment 2
Prediction:
Source
monitoring in older adults may be helped with specific, strong warnings
intended to compensate for any failure to self-initiate monitoring strategies
Participants:
56
younger adults (mean age 23.10; 32F, 24M) and 56 older adults (mean age 71.90;
38F, 18M)
Procedure:
Materials
and procedure were same as Experiment 1 with following exceptions
1. List readings were performed
by a female and recorded and stored on a computer for playback. Word list duration ranged from 34-37 seconds
with same word rate as Experiment 1
2. Half of participants in each
age group were given strong warning that included the following: lists were not
constructed randomly but associated with a theme word, this theme word would
never be presented on list, people sometimes mistakenly remember theme word
though told not to, to figure out theme word and remember to NOT recall it
Results
Phase
1—Initial Recall
1. Overall, higher recall of
studied items than CL’s
2. Proportions of studied items
recalled were significantly higher than proportion of CL’s in younger adults,
whereas they were equal in older adults
3. Warnings had no effect on
proportions of studied items recalled in younger or older adults
4. Warnings significantly
reduced the proportion of CL’s recalled by young adults, but had no effect on
older adults’ recall
5. Older adults gave globally
higher confidence ratings
6. Unwarned participants gave
higher confidence ratings overall
7. Confidence ratings for
studied items were significantly higher than those for CL’s in younger adults,
whereas they did not differ in older adults
Phase
3—Added Recall
1. Younger adults produced more
CL’s during 3rd phase than older adults
2. Strong warnings improved
performance for young adults in this phase but did not influence older adults
3. Tendency to assign higher
confidence ratings in young adults than older adults
4. In warned condition, younger
participants rated higher confidence than older participants; no difference in
unwarned condition
5. In warned condition, total
proportion of activated CL’s was higher for younger adults than older
6. In unwarned condition, total
proportion of activate CL’s was equal for both age groups
The
results replicated findings that older adults recalled fewer studied items and
more CL’s on initial recall than younger adults. Also, confidence ratings were higher for
studied items than CL’s, indicating that these ratings may be useful in
discriminating correct performance.
Because older adults were equally confident on studied items as CL’s,
this implies they were relying more on indistinct, thematic information than
younger adults were. Since CL’s have
high similarity to studied items, they are rated as equally confident as
studied items. Evidence supports the
hypothesis that false recall on DRM lists is due to inefficient monitoring, and
that older adults are more prone to this kind of memory distortion. Warnings did not help older adults’
performance, suggesting that memory distortions are not due to a deficit in
spontaneous use of monitoring processes.
A possible explanation for older adults’ lower performance is that they
have difficulty encoding specific details leading to a similarity of perceived
source features for both studied items and CL’s.