Society for the Scientific Study of Reading,
May 31-June 3, 2001, Boulder, Colorado
Abstracts, in alphabetical order by first author
P. G. Aaron
(firstname.lastname@example.org; Indiana State University) and R.M. Joshi
. Outcome of a component-model-based
remedial reading project.
The Reading-component Model of instruction is developed from the perspectives of classroom instruction. It proposes
that reading is composed of two major components: word recognition and comprehension. The model was used to
remediate 60 at-risk readers from grades 1 through 5. On the basis of their scores obtained from standardized tests, the
children were divided into two groups: poor decoders and poor comprehenders. Poor decoders were instructed in
decoding skills starting with phoneme awareness; poor comprehenders were ta
ught comprehension strategies. The
children were met 30 minutes a day, 4 days a weak, for a year. Post tests showed that poor decoders, when compared
to control groups, gained significantly in both decoding and comprehension. Poor comprehenders also showed gains
but the gains were not uniformly significant in all areas.
Marilyn Jager Adams
(email@example.com; Harvard Graduate School of Education), Hollis Scarborough, Scott Masten, Julie
Smurda, Maria Prokop, and Linda Fidell.
First-year findings from an independent evaluation of California’s K-3 results
The K-3 Results program, a grassroots initiative of California's state-funded consortium of Reading & Literature
teachers (CRLP), is an experiment on using informal, teacher-administered assessments to inform professional
development and instructional progress and priorities. Since 1996 when the program began, the number of
participating teachers has grown from 23 to more than 6,000. In this session, we will present first-year findings from
our independent analyses of student information archived by the program and of its influence on the professional
development and classroom practices of participating K-3 teachers.
Stephanie Al Otaiba
(stephanie.d.al.otaiba@Vanderbilt.edu; Vanderbilt University), and Michelle K. Hosp.
teaching and assessment of literacy for individuals with Down’s syndrome.
Little is known about the effective teaching and assessment of literacy for individuals with Down’s syndrome despite
evidence supporting practices in instruction and assessment of literacy for yo
ung typically developing children (e.g.,
explicit phonological awareness and decoding training, curriculum-based measurement). The purpose of this poster
session is to report findings from 4 case studies examining the effects on reading performance of one-to-one tutoring
program for Down’s syndrome. In addition, the relationship between language ability and growth in letter-sound
correspondence and sight word reading for these children will be reported.
Stephanie Al Otaiba
(stephanie.d.al.otaiba@Vanderbilt.edu; Vanderbilt University), Douglas Fuchs, and Lynn S. Fuchs.
Success for many, but not for all: A review of the literature describing characteristics of children unresponsive to early lite
Recent research indicates that as many as 30% of children at-risk for reading difficulties and as many as 50% of
children with special needs, may not benefit from effective and well-implemented early literacy interventions. This
poster session will present a review of the literature that describes children unresponsive to early literacy intervention.
Findings suggest unresponsiveness was associated with one or more characteristics: low initial phonological skill
(particularly segmentation), slow naming, low verbal ability, poor phonological encoding, attention deficits or
behavior problems, or additional disabilities such as mild mental retardation. Implications for research and practice
will be discussed.
(firstname.lastname@example.org; University of Massachusetts) and Keith Rayner.
Eye movements and reading skill:
Differential effects of word frequency and predictability.
Eye movement research has demonstrated effects of word frequency and predictability on the fixation times of
proficient readers. Those effects are generally assumed to be consistent among college-age subjects irrespective of
overall reading ability. This experiment monitored the eye movements of three groups of college readers as they
silently read sentences containing predictable and unpredictable words of high or low frequency. For the most
proficient readers, frequency influenced initial reading times while predictability did not. Only the lower group
showed strong predictability effects in initial reading times. Results suggest that expert readers engage qualitatively
different processes to support their comprehension of text.
Egbert M. H. Assink
(email@example.com; Utrecht University, The Netherlands) and Sonja van Well.
Contrasting effects of
age of acquisition and word frequency in visual and auditory lexical decision.
An intriguing outcome of recent lexical decision experiments manipulating Word Frequency and Age of Acquisition
(AOA) is that in visual lexical decision the effects of both word frequency and age of acquisition were found, whereas
in auditory lexical decision only an age of acquisition effect was found (Turner, Valentine and Ellis, 1998). A series
of follow-up experiments were designed to shed more light on this paradoxical finding. How can we explain this
remarkable absence of frequency in the auditory condition ? An important feature of the present experiments is that
contrary to the stimulus word set used in the Turner et al. (1998) study, frequency and acquisition age were factorially
manipulated, controlling for concreteness. Reference: Turner, J.E., Valentine, T., & Ellis, A.W. (1998
effects of age of acquisition and word frequency on auditory and visual lexical decision
. Memory & Cognition, 26,
Heather J. Bachman
(firstname.lastname@example.org; Loyola University-Chicago).
Phonological skills and early reading: Sources of
influence and performance discrepancies.
The present investigation examined: age-and schooling-related effects on phonological segmentation tasks (syllabic,
subsyllabic, and phonemic), the influence of phonemic complexity on segmentation skills, and whether growth in
phonemic segmentation skills paralleled children's decoding of increasingly phonemically complex words. Children
whose birthdates fell two months before or after the cutoff date for school entry were followed for two years. A unique
effect of schooling (first grade) emerged only for the phonemic segmentation task, but exclusively for two or three
phoneme words. By the end of second grade, children could accurately decode words with five or more phonemes,
while able to segment only two and three phoneme words.
Caroline E. Bailey
(email@example.com; University of Southern California), Franklin R. Manis, Mark S. Seidenberg, William C.
Pedersen, and Laurie Freedman.
Variation among developmental dyslexics: Evidence from a novel word learning task.
A nonsense word learning task was used to investigate variation in word reading processes among developmental
dyslexics. Phonological dyslexics, surface dyslexics and younger normal readers were taught to pronounce nonsense
words. Half the nonsense words were assigned a regular pronunciation. The remaining half were assigned an irregular
pronunciation. Print exposure and feedback were controlled. Subjects received 6 training trials and two days of
testing. Progress was assessed across training trials, a spelling test, and a post-test of word pronunciation. As
predicted, PD's demonstrated a distinctive developmental pattern. The SD profile was similar yet depressed when
compared to younger normal readers.
Megan M. Bakan
(firstname.lastname@example.org; Florida State University) and Richard K. Wagner.
Orthographic models of geminate
Even good readers can have difficulty spelling words with doubled letters. One problem with geminates is that they
cannot be discerned from the phonological representation, an orthographic representation is required. Two recent
models of the orthographic representation of words have been developed by Caramazza and Miceli (1990) and
McCloskey, Badecker, Goodman-Schulman and Aliminosa (1994) based upon case studies of individuals with
acquired dyslexia. This study tests the accuracy of these models in explaining the geminate spelling errors of
individuals without dyslexia. Findings from both college student and elementary student samples support the
McCloskey et. al. model. Interaction between orthographic and phonological representations are discussed.
Virginia W. Berninger
(email@example.com; University of Washington).
A simple story for the simple view of reading
The 96 lowest-achieving 2nd grade readers from 7 schools were randomly assigned to time-equated treatments over 4
months: word reading; comprehension; combined word reading and comprehension; and repeated reading. Of the
treatments, combined was the most superior to the control (effect sizes, 1.24, real word, 1.1, pseudoword reading). All
treatments improved in word reading and comprehension, and accuracy and rate of word reading (or text reading)
uniquely predicted each of 5 measures selected for text-based and situation-based comprehension, which loaded on
one factor (cf1=1.00). Conclusion: Word reading accuracy and rate constrain comprehension, but word reading and
comprehension also exert reciprocal influences that enable each other.
(firstname.lastname@example.org; City University of New York) and Linnea C. Ehri.
Syllable reading practice
improves decoding and spelling in disabled readers.
Triplets of adolescent disabled readers matched on word reading score were randomly assigned to one of two
instructional treatments or a no-treatment control condition. Syllable-treatment participants practiced reading 100
multisyllabic words by breaking them into syllables. Whole word participants practiced reading the same words as
wholes. Results showed that the syllable treatment produced superior gains in reading and spelling compared to the
word treatment and no-treatment control. Results suggest that disabled readers’ reading and spelling deficits can be
remediated through explicit instruction and practice segmenting and blending multi-letter units in words. Results
support Ehri’s (1992) theory of sight word learning.
(email@example.com; Institute of Child Study, Canada).
The relationship between vocabulary
assessed with picture vocabulary methodology, same words with sentence context method, root word inventory, and reading
I tested the hypothesis that young children (SK and grade 1) may do less well than older children on a vocabulary test
requiring explaining word meaning for words presented in sentences. Items from the Peabody Picture Vocabulary Test
were easier than similar Peabody words given in sentences. However, there was no age discontinuity. Words from the
Root Word Inventory (SSSR 2000) continued to show a very large increase in vocabulary in grade two. The Root
Word Inventory also proved to be more strongly correlated with reading comprehension.
Benita A. Blachman
(firstname.lastname@example.org; Syracuse University), Chris Schatschneider, and Jack M. Fletcher.
phonologically-based tutoring: Do the benefits last?
Second and third grade children with reading disabilities were randomly assigned to an intensive tutoring program that
emphasized the phonologic and orthographic connections in words or to the remediation program provided by the
school. After eight months of tutoring, treatment children significantly outperformed control children on standardized
measures of word identification, word attack, spelling, and paragraph reading and comprehension. In this presentation,
we will report the results of a one-year follow-up on these same measures and also look at the ability of pretreatment
phonological processing measures to predict reading and spelling achievement at the end of the follow-up year.
Jay S. Blanchard
(email@example.com; Arizona State University), Kristen Eignor DiCerbo, Jill Oliver, and Craig A. Albers.
divided-time administration raise test scores? The relationship between attention and standardized reading comprehension
Quantifying student achievement has gained the interest of policy makers and the public. A significant portion of this
interest focuses on reading achievement tests. The researchers examined the effects of reducing attention demands on
reading comprehension test scores. Third grade students (N=939) completed two forms of the Stanford 9 reading
comprehension test. The state-mandated version was administered following standardized procedures in one session.
The alternate form was administered in multiple, divided-time sessions to lessen attention demands. A repeated
measure ANOVA found significant effects for average and low readers. Implications for assessment of reading ability
will be discussed.
Derrick C. Bourassa
(Derrick.Bourassa@Acadiau.Ca; Acadia University, Canada), and Betty Ann Levy.
skills in children with down syndrome: Use of orthographic analogies.
We compared the effectiveness of Rime versus Whole Word training as a means of building a reading vocabulary in
children with Down Syndrome. Reading vocabulary was rapidly and accurately acquired when the words were
segmented into onset- rime units compared to when the words were repeated as whole word units. Furthermore, the
segmentation technique yielded superior generalization to reading new words. These data point to the superiority of
onset-rime segmentation techniques in teaching children with Down Syndrome to read.
(firstname.lastname@example.org; University of Pittsburgh), Julie Van Dyke, Charles A. Perfetti, and Barbara Foorman.
Decoding skill and orthographic knowledge, perfect together.
The following study looks at reading performance as a function of printed word learning of first grade children and the
development of word decoding skills. By printed word learning, we mean the acquisition of specific word forms
beyond the decodability of their letter sequences. Such acquisition is increasingly observed as children come to read
words with increased fluency. According to Perfetti’s (1992) Restricted-Interactive model, the acquisition of a quality
lexical representation depends upon both the number of exposures to the word and the availability of the
correspondences between the graphemes of the word and the appropriate phonemic representations. Share’s self-
teaching hypothesis (1995) rests on a similar idea that each successful identification of a new word (via decoding)
provides an opportunity to acquire the word-specific orthographic representations on which skilled visual word
recognition is based. Relying upon both the orthographic knowledge of words and upon decoding skill, children
develop a bootstrapping mechanism from which to read new and old words alike. As Perfetti’s developmental model
predicts, increased exposures to a word leads to both accurate visual representations of the word itself, but also to the
sub-lexical components of the word. For example, accurately decoding the word “cat” numerous times results in a
strong visual code for the whole stimulus ‘cat’ and for the onset and rime constituents (c = /k/ and at = /at/). Consistent
with this prediction, Treiman’s (et al, 1990) orthographic-rime hypothesis claims that as children acquire
orthographic/lexical knowledge grapheme-phoneme (GPC) rules early, then develop larger and more reliable units (i.e.
onsets and rimes) are coded. We suggest that word identification and acquisition follows this trajectory as exposure to
novel words increases. That is, knowledge of print-sound correspondences enables early bootstrapping to visual word
forms, as exposure to words increases, both lexical and sublexical representations are strengthened enabling more
accurate novel word reading. We set out to test the acquisition of accurate lexical and sublexical word forms as a
function of decoding skill. We measured children’s decoding skills by accounting for individual knowledge of specific
letter-sound patterns from instruction, having students read stories matched to their reading level, and then testing
isolated word reading. The experiment is carried out in two sessions with individual students. In the first session (Day
1), the child will read two stories that are similar in level of decodability (i.e. having nearly identical spelling-sound
patterns) and they are equivalent to the child’s on reading level (as identified by performance predictive reading lists).
In the second session (Day 2), students receive a word list containing familiar words (from the stories) and novel
words (using the letter sound patterns of words form day 1. Novel words were divided into three categories: items
sharing rimes units, items sharing positional dependent GPC units, and items sharing no units with words from day 1.
If the Restricted-Interactive hypothesis holds true, high skilled readers should have stronger lexical and sublexical
representations than less skilled readers, allowing them to read novel words sharing constituents with familiar words.
We predict that children with high decoding skill should be able to accurately identify familiar words, as well as
accurately identify novel words similar in both rime units and GPC units. Children with less adequate skills will
perform well on familiar words, but show decrements in novel words sharing rime units and in words sharing single
GPC units. If the rime hypothesis holds true, low skilled readers should perform worse for rimes over GPC words
when controlling for number of shared phonemes. Data are being collected from approximately 60 first grade children
in the Houston public schools as part of a larger assessment project.
James R. Booth
(email@example.com; Northwestern University), Charles A. Perfetti, Lesley Hart, and Kathleen Barlo.
Developmental and reading skill differences of consistency effects in visual and auditory word recognition.
Children and adults were administered visual and auditory lexical decision tasks to examine consistency effects in
word recognition. The results showed that younger or less skilled readers showed a larger accuracy difference between
phoneme-grapheme consistent and inconsistent items as compared to older or more skilled readers in visual and
Wim van Bon
(W.vanBon@ped.kun.nl; University of Nijmegen, Netherlands), Ben Maassen, and Rob Schreuder.
Pseudoword repetition by poor and normal readers: An error analysis.
Children with poor reading and spelling scores (mean CA 10;1) and literacy score matched children at normal
performance levels (mean CA 7;11) participated in a pseudoword repetition task. Pseudowords differed in their length
(1 to 4 syllables) and in the frequency (low, high) with which their syllables occur in everyday language. Effects of
length and syllable frequency were found, no interaction appeared to exist with literacy status of the participants. Error
analyses as to place and type of the wrongly reproduced speech sounds and characteristics of the speech products also
did not show literacy group differences. The results are in conflict with some current hypotheses.
Victor H. P. van Daal
(firstname.lastname@example.org; University of Wales, United Kingdom).
Orthographic processing in a
In the present research an orthographic processing test was trialed with English beginning readers. It was examined
whether English beginning readers do rely more heavily on orthographic processing due to the orthographic depth of
their writing system.
Kees P. van den Bos
(email@example.com; University of Groningen, Netherlands).
Reading speed, naming speed,
and general processing speed
A domain-specific theory of decoding speed development predicts (1) an increasing relationship between word reading
speed and naming speed (of alphanumeric symbols especially), and (2) no relationship between word reading speed
and general processing speed. These predictions were tested with random reader samples (n’s around 50) at the age
levels of 8, 10, and 12 years of age. Participants were administered various reading tasks which differed in word type
and degree of item repetitivity, rapid-naming tasks with digits, letters, colors, and pictures, and Visual Matching and
Cross-Out tasks (Woodcock-Johnson TCA) as operationalizations of general processing speed. Results support
prediction (2), and partially support prediction 1. A developmental increase of the reading-naming speed relationship
was found for the reading task with the highest degree of item-repetitivity only.
W. van den Broeck
(firstname.lastname@example.org; University of Leiden, The Netherlands).
The role of contextual
information in the development of word recognition ability.
In this study a prediction derived from Share's hypothesis about the disambiguating role of contextual information in
word decoding was tested. Using a longitudinal design, the facility to use contextual information was only slightly
correlated with decoding accuracy one year later, but not with decoding fluency.
Julie Van Dyke
(email@example.com; University of Pittsburgh), D.J. Bolger, Nicole Landi, Charles A. Perfetti, and Barbara
Contributions of word decodability and text predictability in first grade oral reading and printed word learning.
We suggest that decodability percentage of words in a text is only a rough approximation of readability which ignores
important text characteristics. In this study we seek to understand better what affects "decodability tolerances"--
characteristics of the text and the reader that buffer comprehension processes against specific word obstacles. Data to
address this question includes oral reading of 1st graders in an explicit decoding curriculum and a student-directed
curriculum on texts with contexts making target words predictable or not. Two assessment points allow a comparison
of the relative contributions of decoding skills and text predictability for students' word form acquisition.
Julie Van Dyke
(firstname.lastname@example.org; University of Pittsburgh), Nicole Landi, D.J. Bolger, and Charles A. Perfetti.
Decodability as a text factor: Alternative approaches to characterizing word predictability.
Research has shown that decoding skill, defined by students' ability to apply letter-sound correspondences, is the single
best predictor of reading comprehension (Stanovich, 1990; Vellutino, 1991). Consequently, decodability has been
studied from the point of view of the child (e.g. phonemic awareness skill) and of words (e.g. letter-sound
predictability, consistency, frequency, etc.) Yet words come embedded in texts and thus far, the contribution of text-
level factors as contributors to decodability has been ignored. This work compares several formulations of text-level
predictability, suggesting that texts can be more or less decodable despite a child's letter-sound decoding skills or a
(L.Verhoeven@ped.kun.nl; University of Nijmegen, The Netherlands), Rob Schreuder, and Kirsten Dors.
Units of analysis in reading bisyllabic nonwords.
Two experiments have been conducted in order to explore children’s units of analysis in reading Dutch bisyllabic
nonwords. Although Dutch orthography is highly regular, several deviations from one-to-one correspondence occur. In
polysyllabic words the grapheme E may represent three different vowels: å, e, or ù. In Experiment 1, 33 children in
grade 6 were given 8 word lists of bisyllabic pseudowords: words with two times the grapheme E, the first syllable
being a morpheme (1), a prefix (2), or a random string (3); words with E in the first and another vowel in the second
syllable, the first syllable being a morpheme (4), a prefix (5), or a random string (6); and words with a random string in
the first and E in the second syllable with (7) or without (8) a morpheme in the first syllable. It was found that the both
the pronunciation and stress assignment of pseudowords was dependent on word type, showing that morpheme
boundaries and prefixes are being identified. However, the identification of prefixes could also be explained from the
fact that in the present word set prefix boundaries coincide with syllabic boundaries. In order to exclude this
alternative explanation a follow-up experiment with the same group of children was conducted contrasting
pseudowords with two times the grapheme E, with a prefix in the first part of the word not coinciding with syllable
boundaries versus similar pseudowords with no prefix. The results of the first experiment could be replicated in that
the children identified prefixes and assigned word stress accordingly. The results will be discussed with reference to a
parallel dual-route model of word decoding.
(email@example.com; Queen’s University, Canada), Erin O’Donnell, and John R. Kirby
English-speaking children learning to read in French Immersion classrooms.
In Canada, many English-speaking parents choose to have their children educated both in French and English. In many
cases, formal literacy instruction is conducted only in French until grade three. This paper reports the results of the
first year of a three-year longitudinal project studying acquisition of English reading skills in English-speaking
children enrolled in French immersion classrooms. Phonological awareness, phonological memory, and naming speed
at the beginning of grade one are used to predict English and French reading ability at the end of grade one. Results of
teacher observations and parent interviews are used to account for additional variance.
(firstname.lastname@example.org; University of Pittsburgh), Ying Liu, and Charles Perfetti.
Learning to read a logographic
system by alphabetic readers: The role of visual, phonological and semantic processes.
Learning to read entails mastery of association between the print form and spoken form of the language. Learning to
read in a second language (L2) may present a case where the mapping between spoken form and print form of the L2 is
weak and/or indirect due to the mediation of the first language (L1). This hypothesis is tested in the present study on a
group of English L1 adults learning Chinese as L2. The subjects are tested on English (L1) language and reading
skills, visual-reasoning skill and a set of experimental tasks on Chinese character recognition. These experiments are
aimed at examining the role of visual, phonological, and semantic information in learning to read a logographic system
by alphabetic readers, and comparing it with findings in Chinese L1 reading literature. The similarity and difference
between L1 and L2 reading would inform the relationship between oral language and literacy in reading in general.
Victor L. Willson
(email@example.com; Texas A&M University), William H. Rupley, Ronald D. Zellner, and Malatesha R.
Kindergarten-grade 1 reading development in rural-poor and metropolitan-semiaffluent school districts.
This research reports the results of a longitudinal study the reading development of children entering kindergarten in
four schools from three school districts differing in ethnic composition, affluence, and urban-rural setting. Diagnostic
fall testing was conducted in 1998, summative testing spring 1999, grade 1 testing for the same student pool fall 1999
and again spring 2000. Variables included (for kindergarten): letter recognition, print awareness, letter sounds, word
recognition; and for grade 1, letter sounds and combinations (clusters, diphthongs, etc.), rimes, context clues in word
identification and meaning, directions, and story sequence. Teacher instructional reading practices were documented
for each classroom in each school during kindergarten and grade 1. Developmental processes and instructional
practices are evaluated in a longitudinal model for beginning reading development.
(firstname.lastname@example.org; Universität Salzburg, Austria) and Heinz Mayringer.
Subtypes among German
We expected that the typical reading rate problem of German dyslexic children may be due to an impoverished
orthographic lexicon. To examine this possibility, reading accuracy for exception words (foreign English words) was
related to nonword reading accuracy in a sample of reading rate-disabled 10-year-olds. However, the subtyping
showed that only about 30% of rate-disabled children exhibited specifically impaired exception word reading, whereas
about 40% showed average performance on exception word reading and also on orthographic spelling. We conclude
that slow speed of prelexical phonological processing is the main impairment of German dyslexic children.
(email@example.com; Oxford University, United Kingdom), Joel B. Talcott, Catherine J.
Stoodley, Peter C. Hansen, and John F. Stein.
Auditory and visual dynamic processing in developmental dyslexia.
This study investigated how auditory and visual dynamic processing skills might relate to phonological decoding and
orthographic skills in adult developmental dyslexics and controls. The dyslexics were less sensitive than controls to
visual coherent motion (but not coherent form), to auditory 2-Hz frequency (FM) and 20-Hz amplitude (AM)
modulation (but not 2-Hz AM or 240-Hz FM). Multiple regression analyses showed that coherent motion detection
predicted variance in orthographic skills, and thresholds for 2-Hz FM and 20-Hz AM accounted for independent
variance in phonological decoding skills. Thus, visual and auditory dynamic processing could contribute separately to
variance in component reading skills.
(firstname.lastname@example.org; Tufts University), Tami Katzir-Cohen, Robin Morris, and Maureen Lovett.
luency, phonology, and naming-speed in subtypes of dyslexia.
Two large areas of question motivate this study: first, the nature of the relationships between the best known core
deficits in developmental dyslexia phonological and naming-speed processes -- and reading fluency; and second,
whether subtypes of impaired readers defined by these two deficits have differences in fluency at particular levels (e.g.
letter, word, and connected text). Toward these ends, we will apply the criteria used in the Double-Deficit Hypothesis
(DDH) framework (Wolf & Bowers, 1999) to subtype 163 Grade 2 and 3 severely impaired readers in Boston, Atlanta,
and Toronto. Replicated extensively, this hypothesis has two central premises: that processes underlying naming speed
and phonology represent two independent sources of reading failure; and that three subtypes of readers can be
characterized by the presence, absence, or combination of these two deficits. Children will be assessed on a variety of
fluency measures and reading subskills. For our first set of questions, regression analyses will be used to explore the
relationships of phonological awareness and naming speed measures to two outcome measures-- fluency in word
recognition and in connected reading. The second goal of the study is to determine whether significant differences are
found among reader subtypes on fluency-based measures at the letter, word, and passage
evel. Implications for
diagnosis and intervention will be discussed.
(email@example.com; The Open University, United Kingdom) and Pav Chera.
Animated, multimedia ‘talking
books’ can promote phonological awareness in beginning readers.
Studies of computer software have shown its potential to enhance phonological awareness in children with reading
difficulties, but none have examined its use with young children. This paper describes an intervention study where
software, designed in consultation with teachers and children, was used to promote phonological awareness in children
beginning to read. Children were given ten, ten-minute sessions with the software over four weeks, while a matched
control group completed normal classroom activities. The intervention group showed significantly higher increases in
phonological awareness than the control group did, but there were no significant benefits observed for word reading.
Society for the Scientific Study of Reading,