Ellgen, Amanda
What Happens When We Remember the Wrong Thing: The Effects of False Recognition on Memory Performance
Faculty Mentor: C. Brock Kirwan, Psychology
Introduction
Our long-term memory system has the impressive ability to form unique representations
of events and later retrieve those distinct events with minimal interference from similar
events. For example, you may park in the same parking lot everyday, but in a different
spot each time. Most days, you can remember specifically where you are parked even
though each encounter is very similar and thus might interfere with all the others.
Computational models of brain functioning propose that this ability depends on the
complementary processes of pattern separation and pattern completion. Pattern
separation is the process whereby overlapping representations (of, for example, two
similar events or stimuli) are made as dissimilar as possible. Pattern completion, on the
other hand, is the process whereby previously encoded representations are retrieved
given a degraded or noisy cue.
While much research has been done examining the effects of pattern separation in a
variety of contexts, little has been done concerning what would happen if pattern
completed stimuli were to be re-encoded. Is a new memory representation formed
alongside the representation of the original stimulus? Is the original memory
representation modified to incorporate the new episode? This research will be important
for establishing what happens when new information is incorporated into older memory
representations and how the brain handles this conflict. This has implications for
applications such as false memory formation and teaching practices.
Methodology
61 participants (23 men and 38 women) were recruited through BYU via flier, SONA,
and word-of-mouth. The participants took a three-part computer test. The first part, S1,
was a study phase in which participants were shown several images of everyday
objects.
The second part was a test phase, T1, in which participants were shown images of
items exactly the same as the previously studied items (repeats) or items similar to the
previously studied items (lures). During the test phase, participants were instructed to
indicate whether the item is old (i.e., exactly the same as an item previously studied) or
similar (i.e. slightly different than an item previously studied).
The third part was another test phase, T2, in which participants were shown images that
contained paired items, one from the S1 phase and one from the T1 phase. During this
second test phase participants were asked to indicate which image was seen in the S1
phase. Additionally in this phase, confidence levels were determined. See figure 1.
Results
Many responses could have been examined, but focus was on placed incorrectly
labeled stimuli during the first testing phase. In the second testing phase, we examined
the proportion stimuli that continued to be labeled incorrectly to the stimuli that were
corrected upon further exposure. The proportion of corrected stimuli in the second
testing phase was significantly greater than the incorrect.
Additionally, the confidence levels reported at the end of the second testing phase were
different between the correct and incorrect stimuli. The responses that were corrected in
the second testing phases were significantly more confident than the incorrect
responses. Finally, reaction times were measured on the responses. The corrected
response that was also indicated and the most confident had a significantly faster
reaction time than any other response.
Discussion
From these results, we can further understand our memory’s ability to correct. It is
interesting that we are more likely to correct memories upon repeated exposure than to
continue making the same mistake. Additionally, greater confidence and faster reaction
times are associated with correction. Our mind seems to have an ability to recognize
mistakes and correct them quickly. This method of testing was just one example of
memory abilities. It focused on visual and object recognition. Further research will be
needed to examine false memory correction in other areas of memory including audio,
spatial, biological, etc.
Conclusion
In conclusion, we examined what happens when our memories make mistakes and we
remember something incorrectly. In regards to pattern separation, it seems that our
minds are more likely to self-correct and improve than to continue remembering false
memories. This new insight shapes how we view our memories and how our minds
work. These findings further confirm the power and ability of our minds to learn and
remember.