Abstracts 2005 - The Psychonomic Society

Posters 1099–1105 Thursday Evening
frequency neighbors will be the word’s strongest competitors, and preactivation
of these neighbors will delay the word’s identification.
Using the masked priming paradigm (Forster & Davis, 1984), Segui
and Grainger (1990) reported that, consistent with this prediction, a
higher frequency neighbor prime delayed responses to a lower frequency
target, whereas a lower frequency neighbor prime did not
delay responses to a higher frequency target. In our experiments, using
English stimuli, we found this was true only when the primes and targets
had few neighbors—when the primes and targets had many
neighbors, higher frequency primes delayed responses to lower frequency
targets and lower frequency primes delayed responses to
higher frequency targets. We offer several possible explanations for
these findings and discuss their theoretical implications.
(1099)
The Effect of a Word’s Orthographic Neighborhood on Eye Movement
Latencies During Reading. MELISSA CROCKER & CHRISTO-
PHER R. SEARS, University of Calgary, & PAUL D. SIAKALUK,
University of Northern British Columbia—Many studies have examined
how the identification of a word is affected by its orthographic
neighbors. This research is motivated by the fact that, for most models
of word identification, the number of neighbors and the existence
of higher frequency neighbors have important processing implications.
Reports of neighborhood size effects (faster responding to
words with many neighbors) and neighborhood frequency effects
(slower responding to words with higher frequency neighbors) are
consistent with the notion that a word’s neighbors play an important
role in the lexical selection process. But most of this research has used
the lexical decision task—surprisingly little is known of how the identification
of a word is affected by its neighbors during normal silent
reading. In our experiments, participants’ eye movements were monitored
while they read sentences, and the effect of a word’s neighbors on
eye movement latencies (first-fixation and gaze durations) was examined.
(1100)
Components of Word Familiarity: Orthography Controls Eye Movements
During Reading. JESSICA R. NELSON, ERIK D. REICHLE,
& CHARLES A. PERFETTI, University of Pittsburgh—The E-Z reader
model of eye movement control (Reichle et al., 2003) posits two
stages of word identification: an early stage (familiarity) that triggers
eye movements to the next word, and a later stage that shifts attention.
We hypothesized that the early stage reflects orthographic processing,
but that the latter stage reflects orthographic, phonological, and semantic
processing. To test these predictions, participants were first
taught the spellings and pronunciations, spellings and meanings, or
pronunciations and meanings of new (unknown) words, and then they
read sentences containing these words while their eye movements
were recorded. The results of this eye-tracking experiment supported
our hypothesis: Orthographic knowledge shortened first-fixation durations
on the trained words, but orthographic, phonological, and semantic
knowledge shortened gaze durations. These results are consistent
with cognitive models of eye movement control (e.g., E-Z
reader) and support the hypothesis that orthographic processing
largely determines when the eyes will move during reading.
(1101)
The Role of Orthographic Neighbors as Parafoveal Previews in
Reading. CARRICK C. WILLIAMS, Mississippi State University,
MANUEL PEREA, Universitat de València, & ALEXANDER POL-
LATSEK & KEITH RAYNER, University of Massachusetts, Amherst—
In two boundary change experiments, we presented parafoveal previews
that were either identical to the target word, a word that was an orthographic
neighbor of the target (e.g., sweet as a preview for sleet),
or an orthographically matched nonword (e.g., speet). In Experiment
1, the low-frequency word in the orthographic neighbor word
pair was the target and the high-frequency word was the preview, and
in Experiment 2, the roles were reversed. In Experiment 1, the highfrequency
orthographic neighbor word preview provided as much ben-
66
efit as the identical preview and greater benefit than the nonword preview,
whereas in Experiment 2, the low-frequency orthographic
neighbor word preview provided no greater benefit than the nonword
preview. These results indicate that the frequency of the preview influences
the extraction of the letter information without setting up significant
competition between the lexical entries of the preview and target.
(1102)
Do Readers Obtain Parafoveal Information Two Words to the Right
of Fixation? BARBARA J. JUHASZ, SARAH J. BROWN, & KEITH
RAYNER, University of Massachusetts, Amherst—Useful information
is obtained one word to the right of fixation during reading (word
n + 1), which aids recognition of the word once fixated. If word n +1
is completely visible in the parafovea, fixation durations once it is fixated
are shorter than when it is masked (see Rayner, 1998). The present
experiment addressed whether useful information about a word
two words to the right of fixation (word n + 2) can also be obtained.
Adjective–noun pairs (e.g., simple problem) were embedded in sentences.
Prior to fixating the noun, a preview was provided of the noun
itself (e.g., problem), a different word (e.g., furlong), or random letters
(e.g., gnahfin). This preview was given when the noun was word
n + 1 or word n + 2. Parafoveal preview effects were obtained for word
n + 1 but not for word n + 2. These findings have implications for the
role of attention during reading and for models of eye movements in
reading.
(1103)
Rereading and Eye Movements. CYNTHIA H. NULL, NASA Langley
Research Center, & TIMOTHY J. SLATTERY, University of Massachusetts,
Amherst—Rereading a text has been shown to increase comprehension
of it. In addition, reading rate increases during subsequent
readings. We explored the nature of rereading benefits as well as how
long they may last by having participants read single sentences once a
week for 3 weeks while tracking their eye movements. Experimental
sentences contained a target word controlled for frequency and length.
In Weeks 2 and 3, the target word in a third of the items changed, but
no item changed more than once. Half of the changes kept the sentence
meaning the same, the other half caused the sentence meaning to
change. Target word and sentence reading times in Weeks 2 and 3 indicated
significant rereading benefits for items that were not changed
and for items that did not change meaning. We discuss these results and
their implications for abstractionist and episodic accounts of rereading.
(1104)
RT Distributional Analyses of Priming and Degradation Effects in
Word Recognition. MELVIN J. YAP & DAVID A. BALOTA, Washington
University—Stimulus quality and semantic relatedness interact
robustly in visual word recognition, with larger priming effects observed
for visually degraded words. According to additive factors logic, this
finding suggests that stimulus quality and semantic relatedness influence
a common early process in word recognition. We manipulated
stimulus quality and semantic relatedness in lexical decision and naming,
and replicated the interaction in both tasks. Distributional analyses
revealed that priming effects in the clear condition reflect mainly
distributional shifting, whereas priming effects in the degraded condition
reflect both shifting and skewing. In contrast with the view that
degradation and relatedness primarily influence a common early
stage, these variables appeared to also influence later processes en
route to word recognition. The findings are discussed within extant
models of semantic priming and current accounts of the interactive effects
of semantic priming and visual degradation.
(1105)
Effects of Repeated Letters and Lexical Status in the Overlap Model.
PABLO GOMEZ, DePaul University, MANUEL PEREA, Universitat
de València, & ROGER RATCLIFF, Ohio State University—We present
an application of the overlap model to two experiments. In one experiment,
the model accounts for a forced choice task in which sub-