🚀 go-pugleaf

Attached below is a slightly revised schedule for the Mini-Convention.
Note that the talks at 11:35 am (Butcher) and 2:15 pm (Gregory) have been
switched from the original schedule.  Hope to see everyone there on Monday!

Seventeenth Annual Ekstrand Memorial
Psychology Department Mini-Convention

Date:  Monday, May 4, 1998

Location:  Room E214

8:30-9:00 am

Coffee and Donuts

9:00-9:05 am

Introductory Remarks

Alice Healy

9:05-9:30 am

Title:  Character and Line Spacing Effects on Reading Time and Comprehension

Name:  James P. Van Overschelde

Advisor: Alice F. Healy

Abstract:

Participants read prose passages that varied both in line spacing (single
or triple) and in character spacing (normal or two spaces between each
character), while controlling for the number of words per line. Significant
main effects on participants' reading times were found for both character
and line spacing. Participants were slower to read passages printed with
single line spacing than triple line spacing and with two spaces between
each character than normally spaced characters. No main effects were found
on either set of comprehension questions, but a significant interaction was
found for inferential comprehension questions.  The reading time data
provide additional support for Healy's (1994) unitization theory, because
it is difficult to unitize words with two spaces between each character.

9:30-9:55 am

Title:  The Effects of Training Speed on Foreign Language Speech Perception

Name:  Anita R. Bowles

Advisor:  Alice F. Healy

Abstract:

Research has shown that flight simulator training can be enhanced by
presentation of stimuli "above real speed."  The present study examines
whether a similar advantage can be found for foreign language training.
Participants were divided into training groups based on speed of stimulus
presentation: slow, medium, or fast.  Stimuli were 48 Spanish sentences
pronounced by a native speaker.  Participants attempted to type each
sentence and then received feedback.  Each participant was trained on a
subset of the stimuli and at test heard all 48 sentences, one third at each
speed.  We hypothesized that learners would adapt to the rate of training,
and that those trained with rapid speech would find it easier to perceive
foreign words presented at conversational speed.

9:55-10:20 am

Title:  Memory Seeding: Training and Transfer of Population Estimates

Name:  N. Noelle LaVoie

Advisor: Lyle E. Bourne, Jr.

Abstract:

Brown and Siegler (1996) found that training on a subset of country
populations improved estimations for a larger set.  Their results also
suggested that this improvement, called the seeding effect, depended upon
information abstracted during learning to form a general framework, instead
of on the memory of the specific populations.  We conducted two experiments
designed to replicate and clarify their results and to examine the
contributions that specific and general information make to the seeding
effect.  Experiment 1 replicated the seeding effect, but did not provide
conclusive evidence for either a general or a specific information
explanation.  In Experiment 2, the provision of specific or general
information led to equally accurate population estimates, probably because
participants only used a small selection of the information.

10:20-10:45 am

Title:  Effect of Ambiguous Information on State Anxiety and Person Impression

Name:  Patricia B. Sikora

Advisor:  John Forward

Abstract:

An experiment exploring the role of ambiguous information predicted that 1)
ambiguous information would heighten state anxiety and result in a negative
target impression and 2) high power status of a target would amplify this
effect.  Participants read two scenarios, one work-based and one
school-based. No interactions were found; impact of main effects were not
as expected and varied markedly by scenario.  Ambiguous information was
associated with lower levels of anxiety and a more positive target
impression in the school scenario; however, no effects were found in the
work scenario. High power targets were perceived favorably in the school
scenario, but unfavorably in the work scenario.  Availability of schemas is
suggested as one reason for these unexpected results.

10:45-11:10 am

Title:  Effects of a Contact Situation:  How Perceptions of Latinos Held by
White Americans Might be Changed

Name:  Jonathan Bachelor

Advisor:  Bernadette Park

Abstract:

The present study is designed to test the contact hypothesis of attitude
change on White Americans stereotype of Latinos.  Participants were given a
questionnaire assessing their stereotypes of Latinos and of White
Americans.  After 1-2 weeks, participants returned and were put in groups
of 2-5 for a contact situation with a Latino confederate who acted either
confirming or disconfirming of the Latino stereotype, and either identified
or did not identify with the Latino community, depending on group
condition.  Immediately following the contact situation participants filled
out a second questionnaire, again assessing attitudes towards Latinos and
White Americans.  Dependent measures of perceived stereotypicality,
similarity, variability, ethnocentrism and warmth were derived.  Results
indicate several significant changes in attitude towards Latinos.

11:10-11:35 am

Title:  The Influence of Gender Stereotypes on Memory:  An Extension of the
False Memory Paradigm

Name:  Alison P. Lenton

Advisor:  Irene Blair

Abstract:

The purpose of this research was to determine whether memory errors arise
from stereotype associations. Using the Roediger and McDermott (1995) false
memory paradigm, we examined the effect of exposure to gender stereotypic
roles on false alarms in a subsequent recognition test. As expected,
exposure to stereotypically feminine roles increased false alarms to
feminine but not masculine roles, whereas exposure to stereotypically
masculine roles increased false alarms to masculine but not feminine roles.
These results were replicated in Experiment 2, which possessed a more
stringent criterion for role inclusion. In addition, it was found that
participants exposed to feminine roles were more likely to notice that
gender was a theme within the memory list than participants exposed to
masculine roles.

11:35-12:00 am

Title:  Effects of Trained Study Strategies in Comprehension of Expository Text

Name:   Kirsten R. Butcher

Advisor: Walter Kintsch

Abstract:

Effects of training students to use one of two specific strategies in the
study of expository text were examined.  After studying an initial text
using self-selected strategies, 45 students were trained in outlining, in
forming questions about a text, or remained untrained.  Students used the
strategy consistent with their training condition (outlining, margin note
questioning, or self-selected strategy) in studying a text about the human
heart.  Relationships between training condition and study performance were
assessed using pre- and posttests of domain knowledge, as well as Latent
Semantic Analysis cosines of study artifacts to the target text.  Further
analysis assessed whether strategy use was implicated in measures
reflecting representational levels of Kintsch's (1992)
Construction-Integration model of text comprehension.

12:00-1:00 pm

Lunch Break

1:00-1:25 pm

Title:  Differential Roles for Spatial and Verbal Working Memory in the
Comprehension of Spatial Descriptions

Name:  Naomi P. Friedman

Advisor: Akira Miyake

Abstract:

Two experiments investigated the roles of spatial and verbal working memory
(WM) in the comprehension of spatial descriptions.  In Experiment 1,
participants answered spatial and verbal questions about texts varying in
spatial and verbal complexities, then completed measures of spatial and
verbal WM capacities.  Experiment 2 incorporated a revised verbal
manipulation and probes to assess on-line processing.  In both experiments,
spatial complexity affected spatial but not verbal comprehension, and
spatial performance generally correlated with spatial but not verbal WM.
In Experiment 2, verbal complexity affected verbal but not spatial
comprehension, and verbal performance correlated with verbal but not
spatial WM.  These results suggest that representations of spatial and
verbal information were differentially supported by spatial and verbal WM,
respectively.

1:25-1:50 pm

Title:  Involvement of Working Memory in a Visual Search Task

Name:  Tor Wager

Advisor:  Akira Miyake

Abstract:

This study investigated the possible involvement of working memory in an
embedded figures type task.  Thirty-three participants performed a visual
search task in which they searched for a target figure embedded within a
subsequently-presented stimulus figure.  The task was designed to vary
demand on spatial working memory and complexity of the stimulus figure.
The study also examined correlation of performance on this task with
several measures of working memory, field dependence-independence, and
perceptual speed.  A significant interaction between working memory demand
and stimulus complexity was found for target-figure sensitivity (Az),
suggesting that there may be processes common to keeping track of shapes
and dealing with complexity in a visual search space.  Correlational data
support this conclusion.

1:50-2:15 pm

Title: Individual Differences in Reading: Persistence Across Word and Text
Levels

Name:  Jacqueline M. Hulslander

Advisor: Richard K. Olson

Abstract:

Differences in paragraph reading and comprehension between proposed
subtypes of Reading Disability (RD) were investigated.  Subtypes of RD have
been based on phonological and orthographic skills in word-level reading
tasks.  The current study investigates the differential contribution of
these skills to error rate variation in paragraph reading within the RD
group of the Colorado Learning Disabilities Research Center sample.
Results controlling for word recognition and age suggest that individual
differences in these skills at the word level persist at the paragraph
level and that phonological skill, as opposed to orthographic skill, is
primary in predicting reading performance.  In addition, the relationship
between phonological skill and reading in context is strongest in the
younger half of the RD sample.

2:15-2:40 pm

Title:  Verb Subcategorization Frame Biases in Sentence Processing

Name:  Michelle L. Gregory

Advisor:  Dan Jurafsky

Abstract:

This study tested the generality of the claim in the constraint based model
of sentence processing that verb subcategorization frame biases affect
processing.  Previous studies compared only pairs of frames, and only in
garden path sentences, in which participants encountered a temporary
structural ambiguity.  The current study employed a cloze task, with main
verb missing and no possible garden paths, to compare the statistically
most preferred subcategorization frame to a less preferred frame for 24
verbs and 9 subcategorization frames.  Participants were significantly
better able to guess a verb given its most preferred frame than a less
preferred frame.  These findings suggest that verb frame biases are not
limited to the processing of temporary ambiguity nor to specific
subcategorization frames.

2:40-3:05 pm

Title:  Context Effects on Memorization:  Implications for Inference and
Guilty Knowledge

Name:  Beth S. Heywood

Advisor:  Dr. Thomas K. Landauer

Abstract:

Subjects learned lists of words in shared and different contexts, read one
of the two passages from which the word lists were generated via latent
semantic analysis (LSA), and performed free-recall and recognition tasks.
Subjects recalled more words from different-context presentation and more
words related to the passage they read, and in recognition made more hits
and fewer false alarms on words related to the passage they read.  The
results imply that experimenters can determine which passage subjects read
from free-recall and recognition data, suggesting the use of this method as
a forensic test.  The findings will be discussed in terms of LSA and its
implications concerning inferential memory processes.

3:05-3:30 pm

Title:  Effects of Association Distance in Latent Semantic Analysis

Name:  Brian DeCamp

Advisor: Tom Landauer

Abstract:

Traditional Latent Semantic Analysis (LSA) treats all words within a
paragraph as equally associated.  Ignoring the order of words within a
paragraph, LSA generates a multi-dimensional space for semantic
relatedness.  Based on some early results, it was assumed that all words in
a paragraph should be equally associated when generating the LSA semantic
space.  This study tests the assumption that a paragraph of text is the
optimum unit for associating words.   Four models with different
association units (i.e., two-word, phrase, sentence, and paragraph-wide
association windows) performed about as well as traditional LSA techniques
on vocabulary tests and essay grading.  This finding calls into question
previous assumptions about the limitations of co-occurrence matrices.

3:30-3:55 pm

Title: Learning by Exploration: Latent Semantic Analysis as a
Characterization of the Label-Following Strategy

Name: Rodolfo Soto

Advisor: Peter Polson

Abstract:

This study examined the recall of tasks that had been performed once after
a one week retention interval.  Forty eight subjects performed twelve graph
modification tasks using Microsoft Excel.  Participants received
step-by-step instructions in four of the tasks, and they learned the
remaining eight tasks by exploration.  Latent Semantic Indexing was used to
compute the semantic distance between the task descriptions and the labels
in the menus.  Overall, learning and recall was better for tasks with
descriptions that were semantically highly related to the labels in the
menu.  Additionally, there was a tendency to recall better those tasks that
received step-by-step instructions.  This later result has important
implications for theories of skill acquisition.

3:55-4:00 pm

Alice F. Healy

Concluding Remarks

4:00-5:00 pm

Reception