S1 (FriAM 1-65) - The Psychonomic Society

Friday Evening Posters 3010–3015
Northwestern University (sponsored by David N. Rapp)—Research
has demonstrated that readers rely upon fictional stories as sources of
information (Marsh, Meade, & Roediger, 2003). Fiction, though, can
vary in the degree to which it describes settings, characters, and events
that conform with the real world. We investigated the ways in which
unrealistic fiction might influence readers’ reliance on information
embedded in stories, and their attributions of the source of any resulting
knowledge. Stories were modified so contexts were farfetched
and fantastical, to examine whether contexts might dissuade
readers from relying on information in the stories. Participants completed
a test of general world knowledge containing questions about
embedded facts; some of these facts were entirely inaccurate or unfamiliar.
Despite the contexts, participants’ responses suggested that
they learned from the texts. Additionally, participants frequently attributed
facts that they learned, including misinformation, to their
prior knowledge, replicating Marsh et al. This demonstrates the
strength with which readers rely on stories as sources of information.
(3010)
Anticipatory Processes During Comprehension of Poetry. JIM L.
WOEHRLE & JOSEPH P. MAGLIANO, Northern Illinois University,
& DAVID N. RAPP, Northwestern University (sponsored by Joseph P.
Magliano)—Recent research suggests that comprehenders use phonological,
semantic, and pragmatic information to anticipate upcoming
lexical items in discourse. This experiment examined anticipatory
processes during comprehension of rhymed quatrains with fixed
rhyme structures (ABAB). Participants read rhymed quatrains that
were presented without the final word or without the final one or two
lines. Participants had to predict the poem-final word. When participants
were presented with the first several lines of the poem, phonological
constraints (e.g., number of rhyme possibilities) were the primary
determinants of prediction accuracy. However, when presented
with the entire poem (missing the final word), both phonological and
discourse-level constraints (e.g., semantic overlap of the poem-final
word with prior context) influenced prediction accuracy. These results
suggest that while anticipatory processes in poetry comprehension
can be influenced by both surface-level text properties and discourselevel
information, surface-level properties may be of initial import
until further discourse is available to the comprehender.
(3011)
How Nonnative Speakers Make Phonetic Adjustments to Partners
in Dialogue. JIWON HWANG, SUSAN E. BRENNAN, & MARIE K.
HUFFMAN, Stony Brook University—Ambiguities arise when nonnative
speakers fail to make phonetic contrasts that are absent in their
native language. Korean speakers lack the voicing contrast b/p (“mob”
vs. “mop”) and the vowel contrast ae/E (“pat” vs. “pet”), leading to
ambiguous pronunciation of English words. In two referential communication
experiments, Korean speakers of English spontaneously
pronounced target words (e.g., “mob”). A confederate partner either
primed the target words (e.g., asking “What is below ‘hob’?”) or not,
or needed to be able to pragmatically distinguish two contrasting
words (“mob” adjacent to “mop” in the array) or not. Korean speakers
produced more English-like phonetic targets in both the priming
and pragmatic conditions (vowel duration was used to signal both contrasts).
Moreover, Korean speakers were primed to make the disambiguating
contrast when interacting with an English speaker but not
with another Korean speaker of English. Implications for audience design
and adaptation in dialogue are discussed.
(3012)
Can Giraffes Go Without Water Longer Than Zhiraffes? The Influence
of Accent on Truth Judgment. SHIRI LEV-ARI & BOAZ
KEYSAR, University of Chicago—Studies have shown that people
infer various aspects of stimuli from their relative ease of processing.
For instance, stimuli that are easier to process are judged as more familiar,
clear, and truthful. Foreign-accented speech is harder to process
than unaccented speech. Therefore, we investigated whether it is per-
88
ceived as less truthful than unaccented speech. Participants rated the
truthfulness of trivia statements, such as Giraffes can go without water
longer than camels can, read by native speakers of American English
and nonnative speakers with a mild or a heavy accent. Participants
were informed that the speakers read statements that the experimenter
provided. Statements were rated as less truthful when read by nonnative
as opposed to native speakers. Furthermore, the effect depended
on participants’ expectations regarding the difficulty of processing accents
and the heaviness of the speakers’ accents. These results demonstrate
that accent has wide implications for the interaction between native
and nonnative speakers.
(3013)
Individual Differences in Comprehension Monitoring. YASUHIRO
OZURU, CHRISTOPHER A. KURBY, & DANIELLE S. MCNAMARA,
University of Memphis—This experiment investigated whether and
how readers monitor expository text comprehension. Participants read
a 62-sentence passage on brain functions 1 sentence at a time in three
different conditions: judgment of learning, judgment of sentence difficulty,
or no judgment. Participants’ reading time of each sentence
was measured. After reading the passage, participants answered 59
multiple choice questions about the passage content based on memory.
The results indicated: (1) high-skill readers’ reading time is correlated
with objective measures of sentence difficulty; (2) high-skill
readers’ judgments (learning and difficulty) were correlated with their
reading time; (3) readers’ judgments do not correspond to actual performance
on multiple choice questions based on that sentence; and
(4) performing overt metacognitive judgment tasks did not increase
accuracy on comprehension questions in comparison with a nojudgment
group. The results suggest that comprehension monitoring
is a skilled process, but may not contribute to better comprehension
in itself.
(3014)
Reading Comprehension: The Relative Contributions of Four
Sources of Individual Differences. BRENDA A. M. HANNON, University
of Texas, San Antonio—Although there has been considerable
research demonstrating the separate contributions of lower level
processes, higher level processes, working memory, and learning
strategies/metacognitive skills to reading performance, at present, no
study has compared all four sources simultaneously. As a result, we
know little about how these four sources interact (see, e.g., Cornoldi,
Beni, & Pazzaglia, 1996; Perfetti, Landi, & Oakhill, 2005) or whether
one or all of the sources make separate and important contributions
to reading (e.g., Cornoldi et al., 1996; Perfetti, Marron, & Foltz, 1996).
The present study begins to address this issue. The results replicated
a number of previous findings; however, they also revealed some interesting
relationships, such as a negative relationship between repetition
and higher level processes. The results also showed that although
three of the sources made unique contributions to reading
performance, lower level and higher level processes are the most important
contributors.
(3015)
The Impact of Visual Availability on Referential Communication
in Younger and Older Adults. KATYA A. OTIS & WILLIAM S.
HORTON, Northwestern University—In referential communication
tasks, partners work together to negotiate shared perspectives on task
referents. Moreover, prior work has shown that older adults require
more negotiation to achieve successful reference than do young adults
(Horton & Spieler, in press). Typically, however, partners are prevented
from seeing each other, even though eye gaze, head nods, and
facial expressions may be important sources of information about understanding,
particularly for older adults. We tested the impact of visual
availability on communication by manipulating, for separate
blocks of trials, whether pairs of younger and older adults could see
each other as they matched sets of cards containing both abstract tangrams
and caricatures of human faces. Lack of visual availability im-