Abstracts 2005 - The Psychonomic Society

Saturday Evening Posters 5080–5085
sity of Texas, El Paso—Two experiments with 120 fluent Spanish–
English bilinguals evaluated the contributions of word comprehension,
word retrieval, and articulation to repetition priming in translation.
Experiment 1 combined picture drawing (based on word stimuli)
and picture naming as encoding tasks meant to facilitate word comprehension
and word retrieval/articulation processes, respectively.
These tasks facilitated spoken translation additively, but their combination
fell short of the identical repetition effect, indicating that drawing
does not trigger exactly the same comprehension processes as
translation. In Experiment 2, written translation did not fully prime
spoken translation, indicating a small contribution of articulation. Facilitation
effects from written translation and picture naming interacted
subadditively, as was expected, presumably due to overlapping
word retrieval processes. The second word retrieval repetition increased
facilitation, but by a smaller increment than the first repetition.
Overall, practice of word comprehension, word retrieval, and articulation
processes facilitated spoken word translation.
(5080)
A Test of Functional Equivalence of Automatic Memory Processes in
Recognition, Conceptual Implicit, and Perceptual Implicit Memory.
CHRISTOPHER N. WAHLHEIM & DAWN M. MCBRIDE, Illinois
State University—Dual-process models of recognition posit that recollection
is mediated by conceptual processes; however, there is controversy
as to which processes mediate familiarity. The present study
was designed to provide a direct test of the effects of conceptual processing
on familiarity, conceptual automatic memory, and perceptual
automatic memory, using a levels-of-processing (LoP) manipulation.
LoP effects were predicted for recognition familiarity and conceptual
automatic memory, but not for perceptual automatic memory. The
process dissociation procedure (Jacoby, 1991) was used to estimate
automatic memory in a recognition task (i.e., familiarity), a word association
task (conceptual), and a word stem completion task (perceptual).
LoP effects were found for recognition familiarity and conceptual
automatic memory; however, LoP effects were not found for
perceptual automatic memory. These findings suggest that recognition
familiarity may be mediated by conceptual processes.
• COGNITIVE CONTROL •
(5081)
Stimulus-Driven Cognitive Control: Abstract Task Set Selection or
Episodic Retrieval? MATTHEW J. CRUMP, SANDRA J. THOMP-
SON, & BRUCE MILLIKEN, McMaster University—Cognitive control
is often described as involving high-level voluntary shifts of set that
shape lower level perceptual processing. However, recent studies of
task-switching costs (Allport & Wylie, 2000; Waszak, Hommel, &
Allport, 2003) demonstrate that cognitive control is also imparted by
the involuntary retrieval of similar prior processing episodes, which
can either facilitate or interfere with performance. We examined this
conceptual issue further using a different cognitive control tool, the
item-specific proportion congruency (ISPC) effect. Jacoby, Lindsay,
and Hessels (2003) demonstrated that Stroop interference is sensitive
to proportion congruency even when proportion congruency is tied to
item type, varying from trial to trial. Yet it is not clear whether this effect
tells us that abstract cognitive control sets can be adapted remarkably
quickly or that cognitive control is inherent to memory
episodes that are retrieved, of course, quickly. We examine this issue
further in the context of both Stroop and attention capture studies.
(5082)
The Cognitive Representation and Integration of Task Set Components.
ANDREA M. PHILIPP & IRING KOCH, Max Planck Institute
for Human Cognitive and Brain Sciences—The present study examined
the cognitive representation of tasks (task sets), using the task-switching
paradigm. Generally, task sets are thought to contain several components.
In two experiments, we explored different task set components
135
in order to identify those components that can distinguish one task
from another. The experiments showed that task set components, such
as stimulus categories and response modalities, play the same crucial
role in cognitive task representations. In further task-switching experiments,
we manipulated two task set components (stimulus categories
and response modalities) orthogonally. We found that both
components were not represented independently but were integrated
into a single task representation. On the basis of these results, we developed
a new model, the task integration model. The model proposes
that all task set components are equally important for the cognitive
representation of tasks and that different task set components have to
be integrated into one single task representation before subjects can
perform a task (i.e., select a response).
(5083)
Neural Correlates of Cue Encoding and of Task Set Reconfiguration.
STEPHANIE V. TRAVERS & ROBERT WEST, University of Notre
Dame—We used event-related potentials (ERPs) and Stroop stimuli to
examine the processes engaged during a task-switching paradigm in
which the task switched, the cue switched but the task remained the
same, or no switch occurred (cf. Mayr & Kliegl, 2003). Analyses of
activity during the cue-to-target interval revealed differential processing
for pure and mixed blocks and for task switches, relative to
other types of trials. Analyses of posttarget processing distinguished
between pure and mixed blocks and, again, differentiated task switch
trials from cue switch and no-switch trials. These data provide evidence
that task-switching costs arise from an active reconfiguration
of the task set in response to the cue and, possibly, from processes that
serve to regulate proactive interference from previous trials.
(5084)
Instruction-Induced Feature Binding. DORIT WENKE, PETER A.
FRENSCH, DIETER NATTKEMPER, & ROBERT GASCHLER,
Humboldt University, Berlin (sponsored by Peter A. Frensch)—In order
to test whether or not instructions specifying the stimulus–response
(S–R) mappings for a new task suffice to create bindings between
specified S- and R-features, we developed a dual-task paradigm of the
ABBA type, in which participants saw new S–R instructions for the
A-task in the beginning of each trial. Immediately after the A-task instructions,
participants had to perform a logically independent B-task.
The imperative stimulus for the A-task was presented after the B-task
had been executed. The present data show that the instructed mappings
influence performance on the embedded B-task, even when they
(1) have never been practiced and (2) are irrelevant with respect to the
B-task, at least when (3) overlapping features are relevant for both
tasks. These results imply that instructions can induce bindings between
S- and R-features without prior execution of the task at hand.
(5085)
Action Effects in the PRP Paradigm: Which Codes Require Central
Resources? MARKO PAELECKE & WILFRIED KUNDE, Martin
Luther University, Halle-Wittenberg—Ideomotor theories of action
control assume that actions are represented and accessed by codes of
their sensorial effects. In three experiments, we investigated whether
the activation of effect codes is subject to central capacity-limited
mechanisms. Participants made two choice reactions in response to
stimuli presented in rapid succession at variable stimulus onset asynchronies
(SOAs). In Task 2, we varied the compatibility between responses
and forthcoming sensorial effects (Experiments 1 and 2) or
between responses and stimuli partly resembling those effects (Experiment
3). With forthcoming effects, the influence of compatibility
was additive with influence of SOA, whereas with perceptual stimulation
of effect codes, an underadditive interaction with SOA was
found. These results suggest that an endogenous, but not an exogenous,
activation of effect codes occurs during the capacity-limited response
selection stage. The results are discussed with respect to current
models of action control.