Computational modeling and human experimental approaches
We use a combination of approaches to
study human memory processes, including experimental research with
human participants and computational modeling. The latter involves the
development of computer models of human memory processes. Experimental
research in our lab and other labs informs the develiopment of these
models. We have recently reported the first computational model of false recall,
which we have named "fSAM." Unlike other theories, fSAM fully
specifies the theoretical processes assumed to be responsible for false recall.
The model has successfully simulated a broad array of the core findings on false
recall that have been reported in the literature to date. We are
currently pursuing the computational modeling of such phenomena as
forgetting, false recognition, and the effects of cuing eyewitnesses to
recall events they have witnessed. We are also interested in using
computational modeling to simulate other phenomena of interest,
including metamemory processes and memory enhancement.
False recall and recognition.
False memories are memories of events that never occurred, or that occurred
but not as remembered. We are interested in understanding the basic
memory processes underlying false memories, including the processes
used in recall and recognition testing. We are exploring these
processes using both experimental and computational approaches.
Forgetting and false memories.
Most research on forgetting has focused on the loss of access to
previously learned information. Although that issue is important, we
are also interested in exploring how forgetting affects false
memories. Our research has shown that the effect of
forgetting on false memories depends critically on what causes the
forgetting—for example, whether the forgetting is intentional and
goal-directed, or is instead a byproduct of enhancing memory for other
events or facts. We are also interested in exploring other conditions
that affect memory distortions, as well as the extent to which such
distortions affect other cognitive processes, such as inferencing and
reasoning. More generally, we are interested in exploring the ways that
prior experience and knowledge affect our memory for episodes and
Forgetting can serve important goals, such as promoting efficient updating of
memory by reducing interference from information that is no longer
relevant (e.g., one's former telephone number). Our research seeks to
test alternative theories regarding how this forgetting occurs,
including active suppression of to-be-forgotten information, selective
rehearsal of to-be-remembered information, and enhanced contextual
segregation of to-be-remembered and to-be-forgotten information. In
addition, most findings regarding intentional forgetting are based on
laboratory experiments that use lists of words as stimuli. We seek to
determine the extent to which such findings generalize to
real-world stimuli and contexts, such as texts and eyewitness memory.
Part-set cuing is a counter-intuitive memory phenomenon in that it suggests
that, when someone is trying to retrieve information from memory, their
retrieval efforts are actually hurt rather than helped if someone else
provides cues that are part of the information to be retrieved. In
addition to exploring the basic cognitive processes responsible for
this phenomenon, we are interested in exploring whether this
phenomenon generalizes beyond word stimuli to more real-world stimuli
such as scenes and events. In particular, we are interested in how such
cuing affects the volume and accuracy of information retrieved by
eyewitnesses to crimes.
Metamemory versus memory.
Another line of research in our lab involves acquiring a better
understanding of the workings of metamemory-—that is, one's ability to
assess the workings and contents of one's own memory. A critical part
of this research is designed to tease apart behavior that reflects
memory processes from behavior that reflects metamemory processes. A
key example of this issue involves the finding that, when people are
asked to judge which words from a list they are likely to remember,
they are more accurate if they can test themselves after a delay and
then make the judgment. This finding is widely regarded as evidence
that delayed, test-based judgments reflect more accurate metamemory. A
competing view is that this phenomenon is attributable instead to the
operation of typical memory effects rather than any difference in
metamemory accuracy. Our research explicitly tests these competing
explanations and thus seeks to further our understanding of the
cognitive processes that underlie both memory and metamemory phenomena.
Metamemory in the learning process.
We are also interested in exploring the widely held assumption that
better metacognitive accuracy leads to better decisions regarding
allocation of study effort, which in turn leads to better memory
performance. Surprisingly little evidence has been reported supporting
this causal chain. A related area of interest involves developing
methods to train learners to use metacognitive judgments to enhance
learning—training which has been largely neglected in schools and
training programs. We are also interested in exploring the intriguing
scarcity of evidence for a strong relationship between metacognitive
skill and memory performance.
Multiple techniques introducing desirable difficulties.
Another area of interest in our lab is optimization of memory
performance, for both theoretical and practical reasons. One line of
our research has explored interactions among the effects of various
memory enhancement techniques that are designed to introduce desirable
difficulties into the learning process, techniques that make learning difficult
but that yield long-term benefits to memory performance—e.g., generation,
spacing, levels of processing, and variation. Our evidence indicates that
combining techniques does not always result in recall that one would expect by
simply adding together the effects of the individual techniques. This line of
research can inform us about the mechanisms underlying these various
techniques, and it also has important applications to the classroom and
to training programs, where the cost effectiveness of adding techniques
is an issue.
Effects of feedback.
We are also interested in how post-test feedback affects long-term
memory. Research to date on such feedback has provided confusing and
contradictory guidance about the efficacy of feedback in improving
long-term retention. We seek to establish boundary conditions for such
efficacy, in particular for information rich in meaning, by focusing on
variables that are likely candidates to discriminate between
positive, negative, and neutral effects of post-test feedback.
Applications of research to real-world settings
In addition to exploring basic memory processes from a theoretical
viewpoint, we are also interested in understanding how such processes
are used in various real-world settings. One line of research explores
how such phenomena as forgetting, false memories, memory enhancement,
and metamemory judgments apply to eyewitness memory. One goal of this
line of research is to provide recommendations for optimizing not just
the amount of information an eyewitness can provide, but also the
accuracy of that information, in particular by reducing the proportion
of false memories.
Another area to which our research applies is education.
Understanding how cognition works in the classroom and in training
programs is an important step in optimizing the learning and retention
of material presented in such contexts. Accordingly, we seek to
establish the extent to which cognitive phenomena such as forgetting,
false memories, metamemory—and perhaps most importantly—memory
enhancement techniques generalize from the simple stimuli used in
laboratories to more complex information used in real-world educational