Abstracts 2005 - The Psychonomic Society

Saturday Evening Posters 5051–5058
ing on the particular MPT model of interest and the type of parameter
variability and correlation expected.
(5051)
The Role of Monetary Incentives for Recognition Memory. JESSICA
R. CHAMBERLAND & RICHARD A. CHECHILE, Tufts University—
Multinomial-processing trees (MPTs) have been used to examine a
number of processes associated with memory, including storage, retrieval,
fractional storage, and guessing. Recently, Chechile (2004)
proposed several new MPT models that can represent recognition
memory. Some validation evidence for these models was also provided
in that study. In the present study, an additional validation assessment
was performed. In two experiments, the role of monetary incentives
was examined with an MPT model from Chechile (2004). In
both experiments, incentives influenced memory performance by
means of changes in guessing measures. The manipulation of incentives
did not change storage, which is additional validation support.
(5052)
Conjunction Memory Errors: The Contribution of Recollection.
HEEKYEONG PARK & LYNNE M. REDER, Carnegie Mellon University—Conjunction
memory errors refer to false recognition of recombined
compound words (e.g., false alarming to blackbird after
studying blackmail and jailbird). Dual-process models of recognition
have been used to account for these conjunction errors; however, it is
unclear whether recollection is involved in avoiding conjunction errors
or exacerbating these errors. Previous studies showed that the manipulations
known to affect recollection (e.g., divided attention, LOP,
and deadline) did not have an effect in the conjunction error rate. We
presented compound words either in a rare font or in a common font
and tested the compound words with conjunction lures in a matched
font or a swapped font. Participants produced more hits to studied
items that were re-presented in studied rare fonts, but they also made
more conjunction errors when conjunction lures were presented in the
font that was used with rearranged parent words during study. These
results raise issues about the role of recollection in spurious recognition
of conjunction lures.
(5053)
Reducing Conjunction Errors in Recognition Memory. MARIANNE
E. LLOYD, SUNY, Binghamton—Three experiments examine the role
of encoding and testing conditions on the rate of conjunction errors
in recognition memory (e.g., endorsing the test item “firefly” as presented
when “fireplace” and “housefly,” but not “firefly,” were presented).
Presenting distinctive information (e.g., a picture) with each
study item was found to reduce conjunction errors during standard,
but not speeded, recognition tests. Furthermore, these results were obtained
even when participants made both a speeded and an unspeeded
response to each test item. Results are discussed in terms of dualprocess
theories of memory and metamemorial influences in recognition
memory.
(5054)
Manipulating Word Frequency Effects by Varying Processing
Demands: Low-Frequency Words are More Difficult to Encode.
RACHEL A. DIANA & LYNNE M. REDER, Carnegie Mellon University
(sponsored by Lynne M. Reder)—Low-frequency words produce
more hits and fewer false alarms than do high-frequency words.
However, a number of experiments using tasks other than recognition
show effects that seem to contradict the low-frequency advantage in
memory. The present research shows that in addition to the advantage
of low-frequency words in recognition tasks, there is a low-frequency
disadvantage during encoding. That is, low-frequency words require
more processing resources at encoding than do high-frequency words.
Under conditions where processing resources are limited, lowfrequency
words show a larger decrement in recognition than do highfrequency
words. We used a divided attention task, a picture–word interference
task, and examined individual differences in working
131
memory to demonstrate this effect. Thus, the advantage for lowfrequency
words in recognition occurs despite an encoding disadvantage.
We also investigated the mechanisms behind this disadvantage
and the implications for encoding in general.
(5055)
Orienting Task, Letter Distinctiveness, and the Word Frequency Effect
in Recognition Memory. AMY H. CRISS, Carnegie Mellon University—In
studies of recognition memory, low-frequency (LF) words
have higher hit rates (HRs) and lower false alarm rates than do highfrequency
words, called a mirror pattern. In five experiments, varying
study time and orienting task, the full mirror pattern was found
only in two cases: the standard condition requiring study for a later
memory test and a condition requiring a judgment about unusual letters.
A variety of other orienting tasks, both perceptual and semantic
in nature and covering a wide range of different performance levels,
resulted in no benefit for LF targets. We hypothesized that the HR portion
of the word frequency effect is driven by letter distinctiveness and
tested this in a final experiment where letter distinctiveness and word
frequency were orthogonally manipulated. As was expected, we found
an advantage for targets composed of distinct letters, regardless of
study task, and we found no effect of word frequency.
(5056)
Encoding Duration, Task Performance, and Task Dissociation.
MARCI A. FLANERY & CRAIG E. L. STARK, Johns Hopkins University—Conventional
wisdom suggests that memory for an item will
be better if subjects are allowed more time to encode the stimulus.
However, Mulligan (1997, 1999) has demonstrated that in the case of
very short, masked encoding conditions (100 msec), single-item, but
not relational, memory is enhanced relative to longer durations. We
further investigated this phenomenon by varying encoding duration by
20-msec intervals. We found that old–new recognition accuracy
peaked for items encoded for 110 and 190 msec, whereas relational
memory accuracy followed a monotonic function. We further examined
whether these patterns were affected by the existence of prior
representation (by testing with pseudowords), by familiarity (by testing
with and without recent exposure), or by task (by testing with declarative
and nondeclarative tasks) and found that these additional factors
were able to eliminate or alter the nonmonotonic pattern in the
single-item memory task.
(5057)
Attention and Memory Retrieval: Effects of Semantic Encoding and
Divided Attention. JEFFREY P. LOZITO & NEIL W. MULLIGAN,
University of North Carolina, Chapel Hill—Research on attention and
memory suggests that semantic encoding leads to retrieval that is
highly susceptible to the dividing of attention. Three experiments
tested this proposition and showed that dividing attention did not selectively
affect semantically encoded items. Participants encoded a
list of words in one of two ways: semantically or phonetically. Later,
memory was assessed using either a standard recognition test (Experiment
1) or a rhyme recognition test (Experiments 2 and 3). Participants
took the memory test either alone (full attention) or while simultaneously
performing a secondary task (divided attention).
Recognition accuracy was reduced by divided attention on both recognition
tests, and semantically and phonetically encoded words were
equally affected.
(5058)
Concurrent Task Effects on Memory Encoding and Retrieval: Further
Support for an Asymmetry. MOSHE NAVEH-BENJAMIN &
ANGELA KILB, University of Missouri, Columbia, JONATHAN GUEZ,
Achva College, Israel, & SUSAN OLD, University of Missouri, Columbia—Several
studies have demonstrated that divided attention at encoding
significantly reduces memory performance. However, divided
attention at retrieval has been shown to affect memory performance
to a lesser degree, although this relative protection of retrieval