Abstracts 2005 - The Psychonomic Society

Thursday Evening Posters 1008–1013
stimuli. Participants performed object-based and perspective-based
judgments for rooms and bodies in different combinations of reference
frame congruence, defined by the position of the person and the
computer monitor. Rather than showing one dominant reference
frame, results revealed that upright coincided with any two congruent
reference frames, suggesting flexibility in reference frame use for spatial
problem solving.
(1008)
Perspective and Instruction Effects on Mentally Representing a
Virtual Environment. HOLLY A. TAYLOR, Tufts University, &
FRANCESCA PAZZAGLIA, University of Padua (sponsored by
Holly A. Taylor)—How do instructions to focus on particular aspects
of an environment and learning from different spatial perspectives affect
one’s cognitive map? Participants learned an urban virtual environment
from either a survey or a route perspective and were instructed
to focus either on landmarks or on intersections. The route
group learned the environment by watching a virtual person walking
through it, whereas the survey group learned by watching a dot moving
through a map. While learning, participants were stopped at critical
points, and their attention was focused on either landmarks or intersections.
After learning, all participants performed several spatial
tasks: navigation, map drawing, and pointing. Individual differences
in the cognitive style of spatial representation were recorded. Results
showed that spatial perspective, instructions, and individual differences
in spatial representations interacted to affect performance.
These results will be discussed in the context of spatial mental models
and their influences.
(1009)
Encoding Direction During the Processing of Proximity Terms.
AARON L. ASHLEY & LAURA A. CARLSON, University of Notre
Dame—A target’s location may be described by spatially relating it to
a reference object. Different types of spatial terms emphasize different
aspects of this spatial relation, with projective terms (e.g.,
“above”) explicitly conveying direction (but not distance), and proximity
terms (e.g., “near”) explicitly conveying distance (but not direction).
It has been suggested that only aspects of the relation that are
explicitly conveyed by the spatial term are encoded when interpreting
a spatial description. However, recent research has demonstrated that
distance information is encoded during the processing of projective
spatial relations, because such information is important for finding
the target. In the present research, we demonstrate that direction information
is similarly encoded during the processing of proximity
terms that convey a close distance (“near,” “approach”), but not for
those conveying a far distance (“far,” “avoid”), in support of the idea
that multiple aspects of the spatial relation are encoded when they assist
locating the target.
(1010)
The Effect of Recipient Perspective on Direction-Giving Processes.
ALYCIA M. HUND & KIMBERLY M. HOPKINS, Illinois State University—Getting
to unfamiliar destinations often involves relying on
directions from others. These directions contain several cues, including
landmarks, streets, distances, and turns. The goal of this project
was to understand the cues people use when giving directions for navigation.
In particular, does the information provided depend on
whether a route or survey perspective is employed? Sixty-four participants
provided directions to help a fictitious recipient get from starting
locations to destinations in a fictitious model town. On half of the
trials, the recipient was driving in the town (a route perspective). On
the remaining trials, the recipient was looking at a map of the town (a
survey perspective). As predicted, people included significantly more
landmarks and left/right descriptions when addressing a recipient driving
in the town. In contrast, they used significantly more cardinal descriptors
when addressing a recipient looking at a map. These findings
suggest that perspective affects direction-giving processes.
53
• COGNITIVE SKILL ACQUISITION •
(1011)
The Separation of Words and Rules: Implicit Learning of Abstract
Rules for Word Order. ANDREA P. FRANCIS, Michigan
State University, GWEN L. SCHMIDT & BENJAMIN A. CLEGG,
Colorado State University—Artificial grammar learning studies have
implied that people can learn grammars implicitly. Two studies using
word strings, rather than traditional letter strings, examined the incidental
learning of three-word orders. English speakers practiced unfamiliar
strings ordered as either “verb noun noun” or “noun noun
verb.” Despite possible prior associations between the words and the
“noun verb noun” order, self-timed reading speed decreased following
exposure to the unfamiliar rule. This pattern generalized beyond
the specific instances encountered during practice, suggesting that
learning of the structure was abstract. A second experiment found
learning when nouns were replaced with pseudowords, showing that
learning was possible in the absence of preexisting meaning and
meaningful relationships between items. These findings suggest that
word orders can be learned implicitly and that words and orders can
be dissociated during learning. These results extend artificial grammar
learning to more ‘language-like’ materials and are consistent with
accounts emerging from structural priming research.
(1012)
Implicit Learning of Artificial Grammars: Under What Conditions?
ESTHER VAN DEN BOS & FENNA H. POLETIEK, Leiden
University—Numerous artificial grammar learning (AGL) experiments
have shown that memorizing grammatical letter strings enables
participants to subsequently discriminate between grammatical and
ungrammatical strings at least as well as does looking for underlying
rules. The present study examined the circumstances triggering implicit
learning. We suggest that implicit learning occurs during memorizing,
because structure knowledge facilitates this task. In general,
we propose that implicit learning occurs whenever structure knowledge
contributes to fulfilling a person’s current goal. This goal directedness
hypothesis was tested in an AGL study. Adults and children
performed an induction task to which knowledge of the grammar
could be more or less functional. Both groups showed the same pattern
of performance on a subsequent grammaticality judgment test:
Functional conditions (identifying semantic referents, memorizing)
outperformed nonfunctional conditions (identifying different semantic
referents, rating likeability, computing values associated with sentences).
These results suggest that implicit learning is goal directed,
occurring whenever structure knowledge facilitates one’s current task.
(1013)
Movement Matters: Enhancing Artificial Grammar Performance
With Animation. BILL J. SALLAS, ROBERT C. MATHEWS, &
SEAN M. LANE, Louisiana State University, & RON SUN, Rensselaer
Polytechnic Institute—When learning abstract material, one approach
involves exposure to many examples of the corpus (memorybased),
while a second approach involves learning the underlying
structure (model-based). Research (Domangue et al., 2004) using an
artificial grammar task has found that memory-based processing leads
to fast but relatively inaccurate performance and model-based processing
leads to slow but accurate performance at test. Attempts to integrate
memory- and model-based training to facilitate fast and accurate
performance were unsuccessful. The present experiment utilized
a computer-animated training task, whereas previous research used a
pen-and-paper task. Training with an animated representation, or diagram,
of the grammatical rules led to fast and accurate performance
at test. Animation without this explicit representation led to fast but
inaccurate performance. Our results suggest that it is possible to integrate
memory- and model-based processing to enhance performance.
In addition, our results bear on the current debate on the utility of animation
for learning.