ArticlePDF Available

Phonological Segmentation Assessment Is Not Enough: A Comparison of Three Phonological Awareness Tests With First and Second Graders


Abstract and Figures

Despite extensive research on phonological awareness and reading, there has been little effort to study practical questions that would assist practitioners regarding the choice and interpretation of the phonological awareness tests available to them. This study examined the relationship between decoding (real and pseudowords) and three phonological awareness tests (segmentation, blending, and manipulation) taken from the Comprehensive Test of Phonological Processing (CTOPP) with an unselected population of first grade (n = 67) and second grade (n = 49) students. Segmentation displayed the weakest correlation with reading and accounted for no statistical variance in reading beyond what was found in the blending test. It also failed to account for a substantial amount of variance in reading that is captured by the manipulation test. Despite its popularity in educational contexts, phonological segmentation may be less useful than phonological manipulation or blending in assessing the phonological substrates of reading at these grade levels.
Content may be subject to copyright.
Canadian Journal of School Psychology
27(2) 150 –165
© 2012 SAGE Publications
Reprints and permission:
DOI: 10.1177/0829573512438635
atrickCanadian Journal of School Psychology
1State University of New York, College at Cortland, Cortland, NY, USA
Corresponding Author:
David A. Kilpatrick, State University of New York, College at Cortland, Department of Psychology,
SUNY Cortland, P.O. Box 2000, Cortland, NY 13045, USA
Phonological Segmentation
Assessment Is Not Enough:
A Comparison of Three
Phonological Awareness
Tests With First and
Second Graders
David A. Kilpatrick1
Despite extensive research on phonological awareness and reading, there has been
little effort to study practical questions that would assist practitioners regarding the
choice and interpretation of the phonological awareness tests available to them. This
study examined the relationship between decoding (real and pseudowords) and three
phonological awareness tests (segmentation, blending, and manipulation) taken from
the Comprehensive Test of Phonological Processing (CTOPP) with an unselected population
of first grade (n 67) and second grade (n 49) students. Segmentation displayed the
weakest correlation with reading and accounted for no statistical variance in reading
beyond what was found in the blending test. It also failed to account for a substantial
amount of variance in reading that is captured by the manipulation test. Despite its
popularity in educational contexts, phonological segmentation may be less useful than
phonological manipulation or blending in assessing the phonological substrates of
reading at these grade levels.
Malgré des recherches approfondies sur la conscience phonologique et la lecture, peu
d'efforts ont porté sur les questions pratiques qui pourraient guider les intervenants
dans le choix et l'interprétation des tests de conscience phonologique qui leur sont
Kilpatrick 151
offerts. Cette étude examine la relation entre le décodage (mots et pseudomots)
et trois tests de conscience phonologique (segmentation, fusion et manipulation)
tirés du Comprehensive Test of Phonological Processing (CTOPP) chez des élèves de
première (n 67) et de deuxième année (n 49) du primaire. La segmentation affiche la
corrélation la plus faible avec la lecture et n'apporte aucune contribution significative
au modèle de régression linéaire au-delà de celle associée à la mesure de fusion. Elle
n'arrive pas non plus à rendre compte de la variance en lecture contrairement au
test de manipulation. Malgré la popularité de cet indicateur dans le milieu scolaire,
la segmentation phonologique pourrait être moins utile que la manipulation ou la
fusion phonologiques dans l'évaluation des substrats phonologiques de la lecture au
phonological awareness assessment, phonological segmentation, phonological
blending, phonological manipulation, Comprehensive Test of Phonological Process-
ing (CTOPP)
Based on extensive evidence, researchers have determined that phonological aware-
ness is strongly associated with the development of word-level reading skills (Perfetti,
Beck, Bell, & Hughes, 1987; Vellutino, Fletcher, Snowling, & Scanlon, 2004;
Wagner, Torgesen, & Rashotte, 1994). Phonological awareness includes the ability
to notice that spoken words can be divided into smaller units such as syllables, onsets,
rimes, and phonemes. Students who develop phonological awareness to the phoneme
level are able to quickly and easily map printed words to permanent memory (Høien,
Lundberg, Stanovich, & Bjaalid, 1995; Laing & Hulme, 1999), while those who do not
typically struggle in reading (Bruck, 1992; Greenberg, Ehri, & Perin, 1997; Vellutino
et al., 2004).
The goal of the present article is to address a practical question that has been rarely
addressed in the research literature. Simply put, Which phonological awareness test or
tests will be most helpful in determining the presence of phonological awareness dif-
ficulties in educational contexts? Some practitioners may assume, based on the
popularity of phonological segmentation tasks (e.g., DIBELS, AIMSweb, PALS,
Yopp-Singer), that such popularity stems from a body of best practice research. This
is not the case. The construct of phonological awareness has been evaluated in multi-
ple ways, such as segmentation, blending, categorization, and manipulation. It has yet
to be established whether one phonological awareness test or task is better than another
at determining if a student’s reading progress is being affected by poor phonological
awareness skills.
While best practice cannot be established by a single study, the goal here is to take
an important step toward raising this issue as well as encouraging further inquiry.
Answering this best practice question is difficult for two reasons: (a) the sheer number
152 Canadian Journal of School Psychology 27(2)
of tasks that have been used to assess the construct of phonological awareness, and
(b) the lack of research that directly addresses this question.
Measuring the Construct of Phonological Awareness
Researchers have measured the construct of phonological awareness in many ways,
such as rhyming, segmentation, blending, isolation, categorization, and manipulation
(Anthony, Lonigan, Driscoll, Phillips, & Burgess, 2003; Chafouleas, Lewandowski,
Smith, & Blachman, 1997; Høien et al., 1995; Lenchner, Gerber, & Routh, 1990;
Schatschneider, Fletcher, Francis, Carlson, & Foorman, 2004; Schatschneider,
Francis, Foorman, Fletcher, Mehta, 1999; Stanovich, Cunningham, & Cramer, 1984;
Vloedgraven, & Verhoeven, 2009; Yopp, 1988). In addition, each of these ways of
measuring of phonological awareness has been examined using multiple tasks. For
example, rhyming can involve rhyme recognition, rhyme matching, or rhyme produc-
tion (Stanovich et al., 1984; Yopp, 1988) while manipulation can involve deletion,
substitution, or reversal (Lenchner et al., 1990; Wagoner et al., 1999). Also, across
the various tasks, other factors have been considered such as levels of linguistic com-
plexity (syllables, onset-rimes, & phonemes), position of phonemes within words,
whether a phoneme is part of a blend, or whether it is voiced or unvoiced (Anthony
et al., 2003; Seymour & Evans, 1994; Stahl & Murray, 1994). These factors affect
performance on phonological awareness tasks, making best practice questions rather
The Paucity of “Best Practice” Research
Despite hundreds of studies on the relationship between phonological awareness and
reading, there has been no concerted effort devoted to determining the most practical
and effective way of evaluating phonological awareness in schools, given the assess-
ment instruments available to educational professionals. Numerous studies have
incorporated multiple phonological awareness tasks (e.g., Anthony et al., 2003; Høien
et al., 1995; Schatschneider et al., 1999, 2004; Seymour & Evans, 1994; Stahl &
Murray, 1994; Vloedgraven, & Verhoeven, 2009; Wagner, Torgesen, Laughon,
Simmons, & Rashotte, 1993; Wagner et al., 1994; Yopp, 1988). However, these stud-
ies made no attempt at directly comparing tasks for clinical utility. Rather they used
multiple measures to either determine the factor structure of phonological awareness,
or to create a phonological awareness factor that is then used to study its relationship
with reading. Two studies, however, are welcome exceptions, Chafouleas et al. (1997)
and Swank & Catts (1994). In 1994, Swank and Catts could say, “it remains unclear
which measures of phonological awareness will be the most effective in clinical prac-
tice for identifying children who lack sufficient phonological awareness” (p. 10). This
comment is as relevant today as it was in 1994. Unfortunately, these two studies have
not had a substantial impact on the field, based on how rarely they have been cited by
later researchers (according to the citations forward feature in PsychINFO).
Kilpatrick 153
The National Early Literacy Panel (2008) conducted a meta-analysis of studies
looking at the relationship between reading and different aspects of phonological
awareness. This is a welcome step because they raised the question of differing
strengths of correlation between reading and differing aspects of the construct of pho-
nological awareness. However, they examined the correlations between reading and
two dimensions of the phonological awareness construct (level of linguistic complex-
ity & analysis vs. synthesis) abstracted from various tasks, rather than a direct com-
parison between the tasks themselves. Thus, the NELP report does not provide
assistance in terms of selecting among the various tests available to practitioners.
The Context for the Study
A phonological awareness test can be used as a screening (e.g., DIBELS). But once
reading skills begin to develop, direct reading-related tasks (word identification, pho-
nic recoding) tend to parallel or eclipse phonological awareness as a predictor of
future reading (NELP, 2008). Phonological awareness tests are also administered to
students referred for an evaluation of reading difficulties. In such instances, they are
not used to predict future reading skills but to determine the likelihood that the stu-
dent’s level of phonological awareness development is affecting his or her reading
progress. The question in this context is whether some phonological awareness tests
or tasks (e.g., segmentation, blending, manipulation) are more closely associated with
early reading skills than others, and therefore presumably better at addressing the
question that prompted the assessment.
A related question is how to interpret the profile of a battery of phonological aware-
ness tests, such the CTOPP. There is simply no specific research literature designed to
assist with this practical issue. If a child’s performance on the phonological manipula-
tion, blending, and segmentation subtests are all consistently low or high, one may feel
confident in deciding whether phonological awareness training is needed. However, it
is quite common for a student to display a mixed profile with these subtests. What can
be concluded from such a profile? Should teachers invest valuable instructional time
in phonological awareness training with a student who displays a mixed profile, or
only provide intervention for those with consistently low phonological awareness
scores? Can such decisions be data driven? At the present time, they cannot, because
other than the studies mentioned above (Chafouleas et al., 1997; Swank & Catts,
1994), there has been no effort to investigate this specific question. It would seem that
a direct comparison between these phonological awareness tests and reading might
begin to address this issue. The present study was designed to make such a direct
The impetus for this article was the author’s experience of administering the
CTOPP to hundreds of students referred for reading difficulties in an elementary
school context. It became clear that approximately half of these students performed at
or above the 50th percentile on the segmentation test, while an estimated 80% of these
same students performed low average to below average on the manipulation task
154 Canadian Journal of School Psychology 27(2)
(which involves deleting sounds from words). This suggested that these two types of
phonological awareness tests were not equally well suited for determining whether a
student’s reading difficulty was the result of weak phonological awareness skills.
Phonological segmentation assessment is the lone measure of phonological awareness
within several popular tests and batteries (e.g., DIBELS, AIMSweb, PALS, Yopp-
Singer, Sawyer STAS). The concern is that if weak readers do well on a segmentation
test, it may be assumed poor phonological awareness is not involved, while a test of
phonological manipulation might have suggested otherwise. On the other hand, it may
be that segmentation tests more accurately reflect the phonological substrates of read-
ing acquisition but that manipulation tasks overidentify phonological awareness dif-
ficulties. A better understanding of this practical issue prompted the literature review
summarized below and the present empirical study.
Addressing the Issue of Best Practice
This article is designed to take an initial step toward examining which test(s) might
be the most clinically useful in determining if children have phonological awareness
difficulties. It uses a subset of the many phonological awareness assessment
approaches described above. It therefore represents an early step in addressing the
best practice questions raised here. Also, most previous research on various phono-
logical awareness tasks examined just that: phonological tasks. Many, if not most, of
the tasks used were researcher-designed and do not represent actual tests available to
school-based evaluators. Even the two studies referred to above that attempted to
address clinical usefulness (Chafouleas et al., 1997; Swank & Catts, 1994) used
experimenter-designed tasks that predate the CTOPP and are not commercially avail-
able in the form used in those studies. Practitioners need information regarding the
relative usefulness of actual tests available to them, an issue that has received virtu-
ally no attention within the extensive phonological awareness literature. Therefore,
the present study looks at three subtests from the commercially available Comprehensive
Test of Phonological Processing (CTOPP; Wagner, Torgesen, & Rashotte, 1999).
Each subtest represents a different phonological awareness task (segmentation, blend-
ing, and manipulation).
Segmentation, blending, and manipulation were selected based on empirical and
practical considerations. Rhyming was excluded because studies have shown it has
little or no discriminant validity beyond kindergarten and all participants in this
study were beyond kindergarten. Phoneme categorization and phoneme isolation
are not as commonly found in tests available for educators. Also, Oakhill and Kyle
(2000) found that phonological categorization is confounded with working mem-
ory, but phonological manipulation is not. This leaves segmentation, blending, and
Manipulation versus segmentation. While little attention has previously been drawn
to this fact, numerous research reports include data to show that from first grade and
beyond, manipulation tasks display higher correlations with reading measures than
Kilpatrick 155
segmentation tasks (Backman, 1983; Kroese, Hynd, Knight, Hiemenz, & Hall, 2000;
Lenchner et al., 1990; Perfetti et al., 1987; Swank & Catts, 1994; Wagner et al., 1993).
Authors rarely mention this difference. One must discover these differences by exam-
ining reported correlation tables. Rare exceptions include Catts, Fey, Zhang, & Tom-
blin (2001) who said phonological manipulation “ranks highly among phonological
awareness tasks in predicting reading achievement” (p. 40) and Lenchner et al. (1990)
who stated that their manipulation task had a higher correlation with decoding (r .78
and r .74) than any segmentation task reported in the literature.
The following study compares three tests representing segmentation, blending, and
manipulation with the reading abilities of first and second graders. It is in these grades
that reading difficulties are commonly discovered, so they seemed appropriate grade
levels to consider. An unselected sample was used rather than a clinical sample to
provide an examination of how these phonological awareness tests correlate with the
continuum of reading skill levels among first and second graders. A clinical sample
might skew the correlations and be more appropriate for addressing different aspects
of the question of clinical utility than are being addressed here. The goal in this article
is to take a first step toward addressing best practice questions by seeking explicit
confirmation of what is implicitly reported elsewhere, which is that different phono-
logical awareness tests more closely parallel the development of early reading skills
than others.
Based on the correlations reported in previous studies, it is predicted that at both
grade levels, the segmentation test will have lower correlations with word-level read-
ing and provide minimal help in clinical assessment beyond what can be gleaned from
tests of manipulation and blending.
Participants were 67 first-grade (30 female, 37 male) and 49 second-grade (23 female,
26 male) students from a lower-middle class suburban elementary school in Upstate
New York. No specific data was collected on race, but districtwide, more than 94%
of the students are White. All students in first and second grade were recruited and
there were no preselection criteria other than the absence of any visual, hearing, or
cognitive disabilities. All students were native speakers of English. Five of the first
graders and eight of the second graders had already been identified as having a read-
ing disability.
Reading tests. All participants received the Word Identification and Word Attack
subtests from the Woodcock Reading Mastery Test–Revised (WRMT-R; Woodcock,
1999). In the Word Identification subtest, students are asked to read a graded word list.
156 Canadian Journal of School Psychology 27(2)
The Word Attack subtest involves reading pseudowords (e.g., seeg, trast) of increasing
difficulty. The participant’s scores on these tests were the total number of items read
Phonological awareness tests. Three phonological awareness tests were administered.
These tests were taken from the CTOPP: Segmenting Words, Blending Words, and
Elision, which evaluate phonological segmentation, blending, and manipulation,
respectively. Segmenting Words involves separating words into their individual sounds
(e.g., “Say sat one part at a time” /s/ /æ/ /t/). Blending Words involves identifying a
word from its parts (e.g., “What word do these sounds make: /t/ /æ/ /n/?” tan). The
CTOPP Elision subtest involves deleting a sound from a word (e.g., “Say drive with-
out the /r/” dive). Both Elision and Blending Words take students through the con-
tinuum of linguistic complexity, starting with syllable items, progressing to onset-rime
items and phoneme-level items. Segmenting Words uses only phoneme-level items,
but begins with two phoneme words and progresses to words with more phonemes.
The CTOPP measures were administered first, in standard CTOPP order (Elision,
Blending Words, Segmenting Words). Because the goal was to determine the practical
usefulness of these tests, preserving the order that evaluators would actually use was
deemed essential. The WRMT-R subtests were administered afterward. Students were
pulled from independent work time and tested in a hallway outside their classrooms.
Each session lasted about 15 min. All data were gathered from December to March
by a certified school psychologist.
Table 1 includes the means and standard deviations for the raw scores and standard
scores for each of the reading and phonological awareness tests, at both grade levels.
Raw scores were used in all analyses but standard scores are also reported in Table 1
to provide a normative comparison of these unselected samples. Table 2 presents the
intercorrelations among the measures. All phonological awareness tests correlated
significantly with both Word Identification and Word Attack at both grade levels. At
both grade levels, Segmenting Words had the weakest correlation with both reading
measures. Also at both grade levels, the three phonological awareness subtests were
significantly intercorrelated, except for Elision and Segmenting Words at second
To explore the relationship among these measures, hierarchical multiple regression
analyses were conducted separately on the first and second grade samples (Table 3),
one analysis using Word Identification as the dependent variable, and the other using
Word Attack. At both grade levels, Segmenting Words accounted for no unique vari-
ance beyond Blending Words (Model 2). It did contribute unique variance with first
graders when the model included Segmenting Words and Elision (Model 3), though it
Note: First grade was December/
January; Second was February/March
Kilpatrick 157
just failed to reach significance at second grade. In the model that included all three
tests, an impressive 55% of the variance in reading was accounted for by these phono-
logical awareness measures. Blending Words and Elision each contributed a substan-
tial amount of unique variance in this model, though Segmenting Words does not.
Using Word Attack as the dependent variable, the pattern of results was nearly identical,
with only slight variations in magnitude.
Table 1. Means and Standard Deviations: Grades 1 and 2
Grade 1 (n 67) Grade 2 (n 49)
Measure Mean (SD) Mean (SD)
Raw scores
Age 6 years 6 months (3.7 m) 7 y10m (3.7 m)
WRMT-R Word Identification 28.18 (15.84) 50.88 (16.37)
WRMT-R Word Attack 12.31 (8.00) 24.77 (9.15)
CTOPP Elision 7.69 (3.32) 10.86 (4.68)
CTOPP Segmentation 8.12 (2.79) 8.18 (3.19)
CTOPP Blending Words 12.97 (3.56) 14.24 (2.93)
Standard scores/scaled scores
WRMT-R Word Identification 114.43 (12.95) 106.87 (12.34)
WRMT-R Word Attack 113.29 (9.91) 112.51 (14.89)
CTOPP Elision 11.21 (2.22) 10.28 (2.68)
CTOPP Segmentation 9.46 (1.41) 9.53 (1.57)
CTOPP Blending Words 13.27 (2.53) 11.68 (1.98)
Note: (1) WRMT-R Woodcock Reading Mastery Test–Revised; CTOPP Comprehensive Test of
Phonological Processing. (2) The standard scores for the first graders on the WRMT-R subtests likely
reflect an inflated representation of actual skills. Participants in this study are from New York State,
which has a later cut-off for kindergarten entry than most other states (age 5 by December 1; September
1 is most common). Therefore, the national norms of the WRMT-R compare these students with
students who, on average, have completed 3 to 4 fewer months of schooling.
Table 2. Subtest Intercorrelations: Grade 1 (n 67) and Grade 2 (n 49)
WID .88*** .56*** .31* .64***
WA .80*** .67*** .33* .51***
EL .60*** .59*** .20 .29*
SEG .47*** .42*** .26* .35*
BW .65*** .57*** .47*** .55***
Note: Grade 1 is below the diagonal; Grade 2 is above. WID WRMT-R Word Identification; WA
WRMT-R Word Attack; EL CTOPP Elision; SEG CTOPP Segmenting Words; BW CTOPP Blending
*p .05. **p .01. ***p .001.
158 Canadian Journal of School Psychology 27(2)
At both grade levels, Segmenting Words displayed lower correlations with both word
identification and phonic decoding tasks. Also, it contributed no unique variance to
these reading tests once its overlap in variance with Blending Words has been par-
tialled out. With first graders, and with a nonsignificant tendency in second graders,
Segmenting Words and Elision each account for unique variance when the model
includes just those two (Model 3). This finding seems to suggest that because Elision
was more highly correlated with reading but contributed much variance beyond
Segmenting Words, Segmenting Words failed to capture a large portion of variance
that could be attributed to the construct of phonological awareness. To a lesser extent
(based on Beta scores), something similar could be said about Elision, at least with
the first grade sample. Elision failed to account for a portion of variance in reading
ability captured by Segmenting Words. This suggests that neither test can stand alone,
particularly at first grade. Elision and Blending Words each account for unique vari-
ance in word-level reading tests (Model 4), and together account for an impressive
Table 3. Regression Analyses: Grade 1 (n 67) and Grade 2 (n 49)
Grade 1 Grade 2
Variable Model Independent variables p R2p R2
WRMT-R Word Identification
1 CTOPP Segmenting Words .47 .001 .22 .31 .03 .097
2 CTOPP Segmenting Words .17 .15 (ns) .11 .37 (ns)
CTOPP Blending Words .56 .001 .44 .57 .001 .38
3 CTOPP Segmenting Words .34 .001 .21 .09 (ns)
CTOPP Elision .51 .001 .47 .51 .001 .35
4 CTOPP Segmenting Words .17 .10 (ns) .06 .58 (ns)
CTOPP Blending Words .37 .002 .48 .001
CTOPP Elision .38 .001 .55 .42 .001 .54
Word Attack
1 CTOPP Segmenting Words .44 .001 .20 .33 .02
2 CTOPP Segmenting Words .17 .16 (ns) .19 .18 (ns)
CTOPP Blending Words .50 .001 .37 .42 .004 .26
3 CTOPP Segmenting Words .31 .002 .21 .06 (ns)
CTOPP Elision .53 .001 .46 .62 .001 .48
4 CTOPP Segmenting Words .18 .10 (ns) .12 .28 (ns)
CTOPP Blending Words .29 .02 .29 .01
CTOPP Elision .42 .001 .50 .56 .001 .55
Note: WRMT-R Woodcock Reading Mastery Test–Revised; CTOPP Comprehensive Test of
Phonological Processing.
Kilpatrick 159
amount of variance in both real word and pseudoword decoding. This also suggests
that for the population of students in the present study, none of the three phonological
awareness subtests stands alone but that Segmenting Words accounts for no variance
in word-level reading beyond what is found in the other tests.
These results suggest that Segmenting Words may be the least helpful in determin-
ing whether a student’s word-level reading difficulties stem from phonological aware-
ness deficits. It accounts for no unique variance beyond Blending Words and fails to
account for a large amount of variance in word reading captured by Elision, which is
a test that is much more highly correlated with reading.
To understand the differences in findings regarding Segmenting Words between the
first- and second-grade samples, it may be instructive to note that the raw scores of
the Segmenting Words subtest were nearly identical between the first- and second-
grade samples (see Table 1). This implies that phonological segmentation skills plateau
between the first and second grades, a finding reported by others (e.g., Vloedgraven &
Verhoeven, 2009; Wagner et al., 1993). By contrast, the mean raw scores of the sec-
ond graders on the Elision and Blending Words subtests were higher than the first-
grade sample. This appears to indicate that phonological awareness skills continue to
develop beyond first grade but that a task like Segmenting Words loses some of its
strength of correlation with reading growth and development after first grade.
The data presented above suggests that while phonological segmentation is commonly
incorporated into popular test batteries, it may not be best practice to use it alone to
determine whether a student may have difficulties with phonological awareness. Both
the CTOPP Elision and Blending Words subtests, and particularly the combination of
the two, appear to be superior in accessing the phonological substrates of reading than
Segmenting Words. Though other studies may not have made the direct comparisons
this study made, the present findings are similar to what has been reported in the cor-
relation tables of previous reports (e.g., Backman, 1983; Kroese et al., 2000; Lenchner
et al., 1990; Perfetti et al., 1987; Wagner et al., 1993).
From this it might be reasonable to assume that if a weak reader does poorly on a
manipulation task or a blending task but does well on a segmentation task that the
student is likely to have phonological awareness difficulties. In such a case, the seg-
mentation task is simply not as helpful in detecting these difficulties. Scatter plots
from the first- (Figure 1) and second-grade (Figure 2) samples may help illustrate this
issue. Figure 1a shows that there are many first graders who are among the lower read-
ers in this sample who are at or above the group’s median in their segmentation skills.
By contrast, no children from among the lower first-grade readers performed at or
above their group’s median on the Elision subtest (Figure 1c). This confirms the clini-
cal observations that prompted this investigation. However, there were numerous stu-
dents who appear to be doing well in reading that had a comparatively weak score on
the Elision subtest,1 while no such pattern emerged with either Segmenting Words or
Blending Words.
160 Canadian Journal of School Psychology 27(2)
One implication of the findings above is that a segmentation task may fail to recog-
nize phonological awareness difficulties in a struggling reader. Figure 1b indicates that
Blending Words also appears to display a similar pattern as Segmenting Words, despite
its stronger correlation with reading. It must be noticed that for Blending Words, and to
a slightly lesser extent Segmenting Words, students who display a relatively weak per-
formance on these tasks are almost invariably weaker readers. The clinical implication
of this is that weak performance on either of these subtests seems indicative of genuine
phonological awareness difficulties, while a weak performance on Elision might not.
However, good performance on either Segmenting Words or Blending Words cannot be
relied on to accurately rule out phonological awareness difficulties.
Educators are under pressure to use “research-based” approaches when assessing
and teaching reading. Nearly all phonological awareness tasks can be called “research
based” because they all have been shown via research to correlate with reading at
some age level or another. The question, however, is about best practice.
Figure 1. Grade 1 scatter plots (n 67)
Note: WRMT-R Woodcock Reading Mastery Test–Revised; CTOPP Comprehensive Test of Phonological
Kilpatrick 161
Among weak readers, an important clinical question that evaluators seek to answer
is whether those reading difficulties are affected by deficits in phonological aware-
ness. Such deficits in phonological awareness can hinder reading skills throughout
adolescence (Lenchner et al., 1990) and into adulthood (Bruck, 1992; Greenberg et al.,
1997). So when students present themselves with reading difficulties, it is imperative
that practitioners use a test most likely to determine if phonological awareness is a
factor contributing to the reading difficulties. This means that if educators rely on
segmentation or blending to assess phonological awareness in reading assessments,
they may fail to recognize phonological awareness difficulties in a meaningful per-
centage of students with such difficulties. If phonological awareness difficulties are
“ruled out” based on average performance on a segmentation or blending task, stu-
dents are unlikely to get the phonological awareness training they need to assist them
in their reading progress.
While these data need further confirmation, the implications are that educators
should reexamine the common practice of relying on segmentation assessment. Rather,
Figure 2. Grade 2 scatter plots (n 49)
Note: WRMT-R Woodcock Reading Mastery Test–Revised; CTOPP Comprehensive Test of
Phonological Processing.
162 Canadian Journal of School Psychology 27(2)
a combination of manipulation and blending tasks is likely to provide a better assess-
ment of a student’s phonological awareness skills. Each of these tasks captures unique
variance in reading skill and together they account for a sizeable amount of that vari-
ance. It would seem advisable that if subtest performance is low on both manipulation
and blending tests, then phonological awareness intervention is indicated. If both are
average or better, it may be reasonable to assume the student’s reading difficulties are
unrelated to phonological awareness. The difficulty comes when the performance is
split. Here, further investigation will be needed to explore this question more fully.
This study has its limitations. Only the first- and second-grade levels were evalu-
ated. It would be ideal to investigate preschoolers through third graders, as well as
older, struggling students. Also, this study used concurrent measures of phonological
awareness and reading while longitudinal studies would also help address “best-
practice” questions. However, in terms of strength of correlation, longitudinal studies
are consistent with the present results (Wagner et al., 1994; Swank & Catts, 1994).
Another important follow-up would be to compare a clinical sample with a typical
sample and evaluate classification accuracy of the various phonological awareness tests.
A potential threat to the validity of these results was the lack of counterbalancing,
which is designed to guard against practice effects or any other effect that a given
order of administration may produce. However, three factors suggest that the order of
administration did not likely have an impact on the present results. First, this study was
ultimately an examination of specific phonological tests rather than more generically
the tasks that make up those specific tests (i.e., segmentation, blending, manipulation).
As an applied study, preserving the order of administration of the CTOPP was essen-
tial because practitioners would use that order in clinical practice. It could thus be
argued that counterbalancing might have threatened the validity of the results. Second,
the correlations between the specific phonological awareness tests and reading found
in this study are consistent with findings from other studies. Third, the empirical find-
ings suggest there was no practice effect. Rather, there was a decreasing correlation
with reading across the three subtests administered. Segmenting Words was the third
test administered and it consistently displayed weaker correlations with reading than
the two prior tests. The only possibility here is that some sort of “reverse practice
effect” occurred. The consistency of our results with previous studies into these types
of tasks suggests that this is not likely.
Despite its limitations, the present article was designed to do two things. The first
was to bring to the attention of researchers and practitioners what has already been
reported but unheralded in existing research: From first grade and beyond, phonologi-
cal segmentation tasks have weaker correlations with word-level reading skills than
phonological manipulation tasks. Second, the present study provides explicit confir-
mation of what has already been previously reported implicitly. Educators need rec-
ommendations regarding which test best captures the construct of phonological
awareness. It is therefore important to direct research attention explicitly at compari-
sons between differing phonological awareness tasks and tests to determine which
one(s) would be most practical and most highly recommended for educators. While
Kilpatrick 163
the present study fell short of establishing best practice, it took a step in that direction
by suggesting what is not likely to be best practice, which is relying exclusively on
phonological segmentation, despite its popularity in the schools. This is not to say the
skill of phonological segmentation is unimportant for reading, because it is very
important. Rather, the point is manipulation and blending tasks appear to do a better
job of accessing the construct of phonological awareness than a simple segmenta-
tion task.
It may be too obvious to state that we do not need another study showing that pho-
nological awareness correlates with (or predicts) word-level reading skills. However,
we do need more research on the relationship between phonological awareness and
reading that addresses questions of clinical utility. The goal of this article was to
(a) alert practitioners that it would be inadvisable based on current evidence to rely on
phonological segmentation to assess the construct of phonological awareness beyond
kindergarten; and (b) serve as a catalyst for researchers to address the best practice
questions. It will take numerous studies to develop an empirical base to address the
best practice questions. This study was designed to take an important step in that
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship,
and/or publication of this article.
The author received no financial support for the research, authorship, and/or publication of this
1. The pattern displayed with the Elision subtest might be an artifact of its administration and
not reflective of the degree to which phonological manipulation approximates the construct
of phonological awareness. The Elision subtest unfortunately includes a significant shift in
task demands at Item 9, but feedback on incorrect responses stops after Item 5. Item 9 is
the first to include phoneme deletion from within the middle of a two-syllable word (e.g.,
“Say tiger without saying /g/”). There are no instructions or sample items for this type of
manipulation. The next two items are similar. Because the ceiling is reached following three
incorrect items in a row, it is common for students to reach the ceiling between Items 9 and
11. When considering the pattern in Figures 1c and 2c, it is quite possible the “spikes” in the
upper left are a reflection of this. It may be that some students simply were not clear about
the sudden change in task demands. By contrast there are no “spike” patterns in the scatter
plots of the other two subtests nor are there any shifts in task demands. This might mean that
Elision has the potential of being an even more powerful test of phonological awareness
were it not for this apparent artifact of test administration.
164 Canadian Journal of School Psychology 27(2)
Anthony, J. L., Lonigan, C. J., Driscoll, K., Phillips, B. M., & Burgess, S. R. (2003). Phono-
logical sensitivity: A quasi-parallel progression of word structure units and cognitive opera-
tions. Reading Research Quarterly, 38, 470-487.
Backman, J. (1983). The role of psycholinguistic skills in reading acquisition: A look at early
readers. Reading Research Quarterly, 18, 466-479.
Bruck, M. (1992). Persistence of dyslexics’ phonological awareness deficits. Developmental
Psychology, 28, 874-886.
Catts, H. W., Fey, M. E., Zhang, X., Tomblin, J. B. (2001). Estimating the risk of future reading
difficulties in kindergarten children: A research-based model and its clinical implication.
Language, Speech, and Hearing Services in Schools, 32, 38-50.
Chafouleas, S. M., Lewandowski, L. J., Smith, C. R., & Blachman, B. A. (1997). Phonologi-
cal awareness skills in children: Examining performance across tasks and ages. Journal of
Psychoeducational Assessment, 19, 216-226.
Greenberg, D., Ehri, L. C., & Perin, D. (1997). Are word-reading processes the same or differ-
ent in adult literacy students and third-fifth graders matched for reading level? Journal of
Educational Psychology, 89, 262-275.
Høien, T., Lundberg, I., Stanovich, K. E., & Bjaalid, I.-K. (1995). Components of phonological
awareness. Reading and Writing: An Interdisciplinary Journal, 7, 171-188.
Kroese, J. M., Hynd, G. E., Knight, D. F., Hiemenz, J. R., & Hall, J. (2000). Clinical appraisal of
spelling ability and its relationship to phonemic awareness (blending, segmenting, elision,
and reversal), phonological memory, and reading in reading disabled, ADHD, and normal
children. Reading & Writing: An Interdisciplinary Journal, 13, 105-131.
Laing, E., & Hulme, C. (1999). Phonological and semantic processes influence beginning read-
ers’ ability to learn to read words. Journal of Experimental Child Psychology, 73, 183-207.
Lenchner, O., Gerber, M. M., & Routh, D. K. (1990). Phonological awareness tasks as predictors
of decoding ability: Beyond segmentation. Journal of Learning Disabilities, 23, 240-247.
National Early Literacy Panel. (2008). Developing early literacy: Report of the National Early
Literacy Panel. Washington, DC: National Institute for Literacy.
Oakhill, J., & Kyle, F. (2000). The relation between phonological awareness and working mem-
ory. Journal of Experimental Child Psychology, 75, 152-164.
Perfetti, C. A., Beck, I., Bell, L., & Hughes, C. (1987). Phonemic knowledge and learning to
read are reciprocal: A longitudinal study of first grade children. Merrill-Palmer Quarterly,
33, 283-319.
Schatschneider, C., Francis, D. J., Foorman, B. R., Fletcher, J. M., & Mehta, P. (1999). The
dimensionality of phonological awareness: An application of item response theory. Journal
of Educational Psychology, 91, 439-449.
Schatschneider, C., Fletcher, J. M., Francis, D. J., Carlson, C. D., & Foorman, B. R. (2004).
Kindergarten prediction of reading skills: A longitudinal comparative analysis. Journal of
Educational Psychology, 96, 265-282.
Seymour, P. H. K., & Evans, H. M. (1994). Levels of phonological awareness and learning to
read. Reading and Writing: An Interdisciplinary Journal, 6, 221-250.
Kilpatrick 165
Stahl, S. A., & Murray, B. A. (1994). Defining phonological awareness and its relationship to
early reading. Journal of Educational Psychology, 86, 221-234.
Stanovich, K. E., Cunningham, A. E., & Cramer, B. B. (1984). Assessing phonological aware-
ness in kindergarten children: Issues of task comparability. Journal of Experimental Child
Psychology, 38, 175-190.
Swank, L. K., & Catts, H. W. (1994). Phonological awareness and written word decoding.
Language, Speech, and Hearing Services in Schools, 25, 9-14.
Vellutino, F. R., Fletcher, J. M., Snowling, M. J., & Scanlon, D. M. (2004). Specific reading
disability (dyslexia): What have we learned in the past four decades? Journal of Child Psy-
chology and Psychiatry, 45(1), 2-40.
Vloedgraven, J. M. T., & Verhoeven, L. (2009). The nature of phonological awareness through-
out the elementary grades: An item response theory perspective. Learning and Individual
Differences, 19, 161-169.
Wagner, R. K., Torgesen, J. K., Laughon, P., Simmons, K., & Rashotte, C. A. (1993). Develop-
ment of young readers’ phonological processing abilities. Journal of Educational Psychol-
ogy, 85(1), 83-103.
Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (1994). Development of reading-related
phonological abilities: New evidence of bidirectional causality from a latent variable longi-
tudinal study. Developmental Psychology, 30(1), 73-87.
Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (1999). The Comprehensive Test of Phono-
logical Processing (CTOPP). Austin, TX: Pro-ED.
Woodcock, R. W. (1999). Woodcock Reading Mastery Test–Revised. Circle Pines, MN: American
Guidance Services.
Yopp, H. K. (1988). The validity and reliability of phonemic awareness tests. Reading Research
Quarterly, 23, 159-177.
David A. Kilpatrick, PhD, is an assistant professor of psychology for the State University of
New York, College at Cortland. In addition, he is a part-time school psychologist with the East
Syracuse-Minoa School District in Upstate New York. His research interests include the devel-
opment of word recognition skills and the phonological processes that underlie reading disabili-
ties. He received his doctorate from Syracuse University.
... Past research has established the critical role of phonemic awareness in the development of beginning reading (Dee Nichols, Rupley, Rickelman, & Algozzine, 2004;Edwards, & Taub, 2016;Manyak, 2008;Kelley, Roe, Blanchard, & Atwill, 2015;Kilpatrick, 2012;Noe et al., 2014;Yopp & Yopp, 2000). For example, Manyak (2008) said: ...
... 531). However, Kilpatrick (2012) considered phoneme segmentation skills as the sole measure of phonemic awareness on reading instruments. ...
Full-text available
This study investigated the impact of reading storybooks and writing journal activities on print and phonemic awareness of Jordanian kindergarten children. Subjects participated in book-reading sessions with a print focus, and writing journals. A total of 50 children were recruited for the study from one kindergarten in Irbid City, Jordan. Two intact sections of 25 children each served as experimental and control groups. Pre-test measures of children’s print and phonemic awareness were administered. Subsequently, children in the experimental group participated in 24 small-group reading sessions that included a print focus, and 14 writing journals over a 14-week period. As an alternate condition, control-group children participated in conventional instruction methods only. Post-testing indicated that children who participated in print-focused reading and writing journal sessions outperformed their control group peers on four measures of print awareness (words in print, print concepts, alphabet knowledge and letter discrimination, and literacy terms), and on phonemic awareness (letter sound identification, rhyme, phoneme blending, phoneme segmentation, and phonemic manipulation), as well as overall performance. Implications and future research directions are discussed.
... The PAST assessment is a phonological screening assessment in which students are asked to manipulate the sounds or word parts within words by orally deleting or replacing sounds after given prompts. This skill underlies a students' ability to work with phonemes and word parts and is a strong early predictor for future reading development (Kilpatrick, 2012;Suggate, 2016;Wagner et al., 1993). The IDI presents assessors with distinct syllable patterns in which to determine curriculum placement . ...
... When assessed, PA has the ability to predict future reading ability for Pre-K and Kindergarten students and assist in determining possible underlying contributing factors for older students experiencing reading difficulties. Kilpatrick (2012) argues that PA must be assessed past first grade segmentation tasks such as those available in DIBELS. He describes PA tasks typically studied in assessments as phoneme segmentation tasks, the ability to separate or blend individual phonemes in a word such as /l//o//t/ for "lot" and phoneme manipulation tasks which may contain verbal word work in deleting, substituting, or reversing individual phonemes in words. ...
Full-text available
Conference Paper
Through time-series graphs, both special education and general education teachers often evaluate progress monitoring data to make both low- and high-stakes decisions for students with and at risk for disabilities. The construction of these graphs–specifically the presence of an aimline and the data-points per x- to y-axis ratio (DPPXYR)–may impact decisions teachers make. The purpose of this study was to evaluate the impact of graph construction manipulations on preservice teachers’ accuracy with instructional decision making. Participants included 94 preservice teachers enrolled in an introductory course focused on students with disabilities at two universities. Following instruction on progress monitoring, students evaluated 48 graphs representing eight data sets with six manipulations (i.e., with and without aimline; DPPXYR set at 0.05, 0.10, 0.15). Results suggest the presence of an aimline increased accuracy; whereas, the manipulation of the DPPXYR led to mixed findings. Implications for future research and practice are discussed.
... Dyslexia is characterized by lifelong difficulties in learning or using skills involved in accurate and fluent word reading and spelling (Rose, 2009). Phonological processing is the ability to discriminate speech sounds into smaller components which, in turn, supports learning the associations between phonemes and print (i.e., orthographic mapping; Kilpatrick, 2012). Irrespective of age, individuals with dyslexia tend to exhibit a core deficit in phonological processing that manifests as inefficiency in decoding and learning the orthography of unfamiliar words (Miller-Shaul, 2005). ...
... Namely, administering a comprehensive battery may create extra and unnecessary burdens (e.g., expending additional time and money on assessment) for adults, which could be circumvented. For example, psychologists could reduce the number of phonological processing tests that they administer to adults by selecting only tests that efficiently measure phonemic proficiency (e.g., substitution, deletion) rather than basic/active tasks (e.g., blending, segmenting) that poorly correlate with word reading ability past Grade 2 (Kilpatrick, 2012). These findings provide a rationale for the widespread development of dyslexia assessment guidelines (e.g., DfES, 2005) that help ensure adults are administered a catered battery consisting of only necessary tests. ...
Objective This study explored how psychologists in Australia assess and diagnose adults with dyslexia. Psychologists’ understandings about dyslexia were recorded alongside the tools used for diagnosing adults with dyslexia (and how these differ from practices with young people). Training experiences and influential factors on diagnostic decision-making in assessing adults for dyslexia were identified. Method An online survey, based on the study aims, was distributed to various professional associations. Participants were 32 registered psychologists in Australia who had assessed adults for dyslexia. Survey responses were analysed using descriptive statistics and non-parametric analyses. Results Most participants reported: (1) that dyslexia is attributable to deficits in phonological processing and rapid automatized naming, (2) assessing adults for dyslexia involves the administration of Wechsler cognitive and achievement tests, (3) familiarity with test tools, diagnostic criteria, and clinical judgement informed adult-focused dyslexia assessment practices, (4) adults and children are assessed for dyslexia in similar ways, and (5) there is little post-registration training specific to assessing adults for dyslexia. Conclusions Psychologists extrapolated child-focused dyslexia assessment practices to adults. This may result in a missed dyslexia diagnosis and/or expending unnecessary resources on assessment. There is a need for increased training in developmentally-sensitive methods for assessing adults for dyslexia.
... Phonemic awareness has been generally conceptualized as the ability to be aware of and/or manipulate phonemes within words. It is a latent construct that has been assessed in many ways with a variety of tasks, including segmentation, isolation, categorization, deletion, and substitution (Kilpatrick, 2012a(Kilpatrick, , 2012b. Only recently has any effort been made to examine whether some phonemic awareness tasks are better suited than others for assessing the phonemic substrates of reading (Kilpatrick, 2012a(Kilpatrick, , 2015. ...
... Only recently has any effort been made to examine whether some phonemic awareness tasks are better suited than others for assessing the phonemic substrates of reading (Kilpatrick, 2012a(Kilpatrick, , 2015. It turns out that phoneme manipulation tasks, the most common being phoneme deletion and substitution, correlate more strongly with reading than phoneme segmentation and blending tasks do (Catts, Fey, Zhang, & Tomblin, 2001;Kilpatrick 2012aKilpatrick , 2012bKilpatrick , 2015Swank & Catts, 1994;Wagner, Torgesen, & Rashotte, 1999). Phoneme manipulation "ranks highly among phonological awareness tasks in predicting reading achievement" (Catts et al., 2001, p. 40). ...
This book chapter explores the implications of assessing word-level reading skills given the scientific findings regarding reading acquisition and reading disabilities.
... Dyslexia is characterized by lifelong difficulties in learning or using skills involved in accurate and fluent word reading and spelling (Rose, 2009). Phonological processing is the ability to discriminate speech sounds into smaller components which, in turn, supports learning the associations between phonemes and print (i.e., orthographic mapping; Kilpatrick, 2012). Irrespective of age, individuals with dyslexia tend to exhibit a core deficit in phonological processing that manifests as inefficiency in decoding and learning the orthography of unfamiliar words (Miller-Shaul, 2005). ...
... Namely, administering a comprehensive battery may create extra and unnecessary burdens (e.g., expending additional time and money on assessment) for adults, which could be circumvented. For example, psychologists could reduce the number of phonological processing tests that they administer to adults by selecting only tests that efficiently measure phonemic proficiency (e.g., substitution, deletion) rather than basic/active tasks (e.g., blending, segmenting) that poorly correlate with word reading ability past Grade 2 (Kilpatrick, 2012). These findings provide a rationale for the widespread development of dyslexia assessment guidelines (e.g., DfES, 2005) that help ensure adults are administered a catered battery consisting of only necessary tests. ...
Previous research has investigated how psychologists identify dyslexia in school-aged children. However, it is presently unclear how psychologists diagnose dyslexia in adults. This study aimed to explore psychologists' understandings and experiences in how they assess adults for dyslexia. Nine psychologists in Australia were recruited from professional associations and interviewed using a semi-structured schedule. After member checks, transcripts were analysed using reflexive thematic analysis. The results suggested that participants' assessment practices with adults were similar to those used with children. However, participants were not confident in assessing adults due to a lack of an empirical base and training, and appropriately normed tools. Moreover, participants relied on their clinical judgment to help overcome barriers unique to the assessment of adults including obtaining an accurate developmental history and determining the relevancy of academic intervention for a formal diagnosis. Participants recommended (better) training and accessible research about how to efficaciously diagnose adults with dyslexia. The robustness of current diagnostic tools for equitably identifying adults with dyslexia was questioned by some participants. There is a need for national guidelines in Australia to support psychologists in identifying adults with dyslexia. International research and guidelines have an important role to play in informing this process.
... Even when they address the same component of phonological processing, some tests predict reading skills better than others. For example, Kilpatrick (2012) compared the impact of different phonological awareness tests on reading and reported a better association of the phonological manipulation evaluated by the elision test (deleting a sound from a word, e.g., saying drive without the /r/ -dive) than of the phonological segmentation test (separating words into their individual sounds, e.g., saying the word sat one part at a time -/s/ /ae/ /t/) with reading. One of the most plausible explanations is that phonological tests may differ in their cognitive requirements (Backman, 1983;Vandervelden & Siegel, 1995;Yopp, 1988). ...
... It also explains the significant correlation between the spoonerisms task (featuring a high level of complexity) and reading (Law, Vandermosten, Ghesquière, & Wouters, 2014;Ramus et al., 2003), as well as varying degrees of correlation between reading proficiency and tasks targeting the same traditionally distinguished component (Del Campo, Buchanan, Abbott, & Berninger, 2015). Indeed, the more complex phonological manipulation task predicts reading proficiency better than the less complex phonological segmentation task (Kilpatrick, 2012). ...
Full-text available
Background The important role of phonological processing for reading has been demonstrated by many studies. The purpose of this research was to investigate the role of phonological processing for reading in Russian. Specifically, we tested whether the overall complexity of a phonological task predicts reading fluency and reading comprehension. Method We used seven phonological tests ranked according to the number of linguistic processes involved in each task. We examined the relative difficulty of the tests and the relationship between phonological processing and reading skills (reading fluency and comprehension of simpler and more difficult texts) in 90 typically developing 7 to 11 years‐of‐age Russian‐speaking children. Results Phonological tests that involved more linguistic processes had lower response accuracies. At the individual level, a greater estimated cost of adding a linguistic process to a phonological test was associated with a reliable decrease in reading fluency but not reading comprehension. Conclusions Our findings confirmed the substantial role of phonological processing in reading acquisition while stressing a higher predictive value of more complex phonological tests for reading fluency. The relationship between phonological processing and reading comprehension, in Russian, needs further investigations.
... Furthermore, more precise questions on screen time and it's specifics and content could possibly improve our results: indeed, no significant relationship was found for active screen time variable. And consequently, more accurate understanding of phonological processing and phonological memory, in particular, may also be reflected in future research to consider possible effects of passive and active screen time in preschool children, since it is proved that some test on phonological processing components have better prediction power than others even when measuring the same component (Kilpatrick, 2012). Nevertheless, we believe that the value of this study is to provide preliminary evidence on the possible relationship between screen time and phonological memory in children. ...
Full-text available
The purpose of this study was to fill this gap by examining the relationship between phonological memory in preschool children and their passive (watching TV) and active screen time with using of Smart Screen Technologies such as tablets and phones with a touch screen interface. Study was conducted in two stages: in Time 1, the association between children’s phonological memory, passive and active screen time and family factors was examined; in Time 2 (1 year later) the impact of passive and active screen time on a child’s individual progress in phonological memory development was evaluated. The study enrolled 122 preschool children aged 5–6 years ( M = 5.72, SD = 0.33); boys (54.9%). Information on each child’s average daily passive and active screen time was obtained from a survey with the mother. The survey provided information on how much time each child spent on a typical day with passive (“traditional”) and active (interactive) use of digital devices. For family factors, we included maternal highest educational qualification, family’s financial situation. For children’s characteristics, age, gender and non-verbal fluid intelligence were included. The results indicate that time spent passively with digital devices (watching TV) is negatively related to a child’s ability to process verbal information. In contrast, the interactive time the child spent with Smart Screen Technologies is not significant and does not pose a threat to the development of phonological memory in preschool age. The study also showed that passive and active use of digital devices has no long-term impact on children’s phonological memory development progress over a year. The implications are that use of Smart Screen Technologies, which implies a higher degree of interactivity, is not associated with either short- or long-term negative effects on phonological memory development in preschool age, contrary to passive screen time exposure. The results can be applied in the elaboration of principles and programs on the use of digital devices for the entertainment and education of preschool children.
... Phonological omission errors, signaling the partial alphabetic stage, were highly probable in Phase 1 students who were mostly Kindergartners and perhaps still developing phonological processing skills. This finding is not surprising given that the basic phonological skills (i.e., phoneme blending and segmentation) which influence early spelling often develop throughout kindergarten and first grade, with most students mastering these skills by the end of their first-grade year (Kilpatrick, 2012). Being that our data were collected at the end of the school year, the five first graders and one second grader in Phase 1 may have yet to master the basic phonological skills necessary for learning to read and spell across both transparent and opaque alphabetic orthographies (Melby-Lervåg, Lyster, & Hulme, 2012). ...
This study examined Spanish spelling errors among 166 native Spanish-speaking students from Kindergarten to Grade 3 based on a spelling-to-diction task. Fifteen types of spelling errors were analyzed in a latent class analysis. Results suggested three phases of spellers: Phase 1 students had a high chance of committing almost all types of errors. Phase 2 students had a lower chance of committing vowel-based phonological errors, but still had a high chance of committing consonant-based phonological errors. Phase 3 students had some difficulty differentiating distinct but similar consonant sounds for spelling (e.g., spelling reptile [reptile] as rebtil; Diente [tooth] as viente). Among the word-context factors, syllable complexity and orthography difficulty were the best predictors of phonological and orthographic errors in more proficient spellers, respectively. We suggest that awareness of Spanish vowel and consonant sounds may not grow in tandem, and future research should incorporate word-context features into spelling error analysis.
Purpose The main purpose of this study was to investigate whether the performance on each of seven phonological processing (PP) tests from the Russian Test of Phonological Processing (RuToPP), with their varying levels of linguistic complexity and composite phonological indices, are significant predictors of developmental dyslexia (DD) and can reliably differentiate children with and without reading impairment. Additionally, we examined the general contribution of phonological skills to text reading fluency in children with various levels of reading performance. Method A total of 173 Russian-speaking 7- to 11-year-old children participated in this study: 124 who were typically developing (TD) and 49 who had been diagnosed with DD. We assessed reading fluency with a standardized reading test and PP with the RuToPP. We investigated the potential of phonological skills to predict the presence or absence of a dyslexia diagnosis using multinomial logistic regression, receiver operating characteristic (ROC) curve analysis, and calculations of the sensitivity and specificity of each test and index. The contribution of phonological skills to reading fluency was also assessed in a mixed group of children. Results Six of seven RuToPP tests were significant predictors of dyslexia. However, while the RuToPP correctly identified 93%–99% of TD children, for children with dyslexia, it ranged from 4% to 47% depending on the test. In a mixed group of children with and without dyslexia, performance in the more complex phonological tests was a stronger predictor of reading fluency. Conclusions Our findings are consistent with the literature on predictors of literacy skills and dyslexia while uniquely demonstrating the impact of the complexity level of the phonological tests on the classification outcome. PP is a significant and necessary predictor of reading skills, but it is not sufficient for diagnostic purposes. Supplemental Material
Purpose The purpose of this study was to create and evaluate online learning modules designed to teach speech-language pathologists (SLPs) dialogic reading strategies and phonemic awareness skills. Method School-based SLPs ( n = 28) were assigned to complete one of two online learning modules. One module taught dialogic reading strategies, and the other taught phonemic awareness. Participants in the dialogic reading group served as controls for the phonemic awareness group and vice versa. Participants completed a pretest that assessed phonemic awareness skill and knowledge of dialogic reading strategies to control for prior knowledge and skill, completed their assigned module, and then completed a posttest that assessed phonemic awareness skill and knowledge of dialogic reading strategies. Results Data were analyzed using multiple regression. The independent variables were pretest score, group, and the pretest score by group interaction. The dependent variable in the first model was posttest score on the dialogic reading measure. The dependent variable in the second model was posttest score on the phonemic awareness measure. There was a statistically significant group by posttest score interaction in the dialogic reading model, indicating that the dialogic reading module was effective for improving knowledge of dialogic reading strategies. There was not a statistically significant group by posttest score interaction in the phonemic awareness module, indicating that the phonemic awareness module was not effective for improving phoneme segmentation skill and phoneme manipulation skill. Conclusions Online learning modules may be effective for establishing knowledge needed for evidence-based practice in speech-language pathology. Additional research is warranted to determine whether online learning modules can be used to change clinician intervention behavior.
Full-text available
Purpose: Speech-language pathologists have the skills and knowledge needed to play an important role in the early identification of children who are at risk for reading difficulties. Whereas research has identified language and other factors that may be predictive of future reading problems, studies have not provided the statistical models and classification data needed for the implementation of early identification programs. In this paper, we report the results of a longitudinal study that examined kindergarten predictors of second-grade reading outcome. Method: Six hundred and four children were given a battery of language, early literacy, and nonverbal cognitive measures in kindergarten as part of an epidemiologic study of language impairments in children. Follow-up testing of reading achievement was completed in second grade. Participants were divided into those children with and without reading difficulties. Results: Findings indicated that five kindergarten variables (letter identification, sentence imitation, phonological awareness, rapid naming, and mother's education) uniquely predicted reading outcome in second grade. Clinical implications: A logistic regression formula and classification data based on these results are provided. Suggestions are offered concerning how this information could be used in an early identification and intervention program for children who are at risk for reading difficulties.
This study examined whether word-reading processes operate similarly in adult literacy (AL) students and elementary school students matched for reading level. Comparison of mean performances revealed that adults were severely deficient on phonologically complex tasks (segmentation, deletion, and nonword reading). In contrast, on orthographically complex tasks, adults revealed both strengths (sight-word reading) and weaknesses (spelling). Regression analyses revealed that individual differences in word and nonword reading abilities were explained by the same orthographic and phonological processes in AL students and children, despite differences in their levels of performance. Correlations between word reading and spelling measures were weaker among AL students than among children. Inadequate integration of these skills may explain adults' phonological deficits as well as their reading acquisition difficulties.
Research has shown that explicit awareness of the speech sound structure of language—that is, phonological awareness—is related to early reading development. The purpose of this study was to assess the effectiveness of four measures of phonological awareness in predicting first grade decoding ability. Measures of phonological awareness at the beginning of first grade were found to be correlated with measures of decoding ability at the end of first grade. Correlations between decoding and phonological awareness were generally much higher than those obtained for measures of decoding and verbal and nonverbal intelligence. Discriminant analyses procedures indicated that several tasks identified good and poor decoders, with approximately 80% to 90% accuracy. The clinical implications of these data for the speech-language pathologist are discussed.
The purpose of this study was to investigate the role of speech-sound segmentation, blending, and discrimination in reading acquisition. These psycholinguistic skills were examined in children who learned to read prior to formal instruction in school, in comparison with an age-matched group of nonreaders, and a group of older children reading at the same level. None of the skills assessed appeared to be true prerequisites to beginning reading. However, early readers were advanced in their ability to perform a complex task involving manipulation of sounds within a temporal sequence, and performance on this task was predictive of nonsense word decoding and spelling skills. Issues related to the interpretation of these skills as consequent, facilitative, or independent factors in reading acquisition are addressed./// [French] Le but de cette étude était de rechercher le rôle de la segmentation parole-son, la fusion et la discrimination dans l'acquisition de la lecture. Ces compétences psycholinguistiques ont été éxaminées parmi des enfants qui ont appris à lire avant l'instruction formelle en milieu scolaire, en comparaison avec un groupe de non-lecteurs du même âge et un groupe d'enfants plus âgés lisant au même niveau. Aucune des compétences établies n'a semblé fournir de vraies conditions préalables pour une lecture de début. Cependant, les lecteurs prématurés étaient avancés dans leur capacité d'accomplir une tâche complexe comprenant la manipulation de sons dans une séquence temporelle et l'accomplissement dans cette tâche a affirmé le décodage de mots inintelligibles et des capacités d'épellation. On a adressé des questions liées à l'interprétation de ces compétences comme facteurs conséquents, facilitateurs ou indépendants de l'acquisition de la lecture./// [Spanish] El objetivo de este estudio era la investigación de la función de la segmentación de hablasonido, el amalgamiento, y el discernimiento en el aprendizaje de lectura. Estas destrezas psicolingüísticas fueron examinadas entre alumnos que aprendieron a leer antes de inicar instrucción formal escolar, comparados con un grupo de la misma edad que no había aprendido a leer, y con un grupo de alumnos mayores que leían al mismo nivel. Ninguna de las destrezas evaluadas parecieron ser prerequisitos indispensables para principiantes de lectura. Sin embargo, lectores de temprana edad mostraron estar avanzados en su habilidad de ejecutar una actividad compleja relacionada a manipulación de sonidos dentro de una secuencia temporal, y los resultados de esta actividad sirvieron de predicción de las destrezas de descifre de palabras sin significado y de deletreo. Se discuten temas relacionados a la interpretación de estas destrezas como consecuentes, de ayuda, o como factores independientes en el aprendizaje de lectura.
The purpose of this study was to determine the reliability and validity of tests that have been used to operationalize the concept of phonemic awareness. Ninety-six kindergarten children were given 10 tests of phonemic awareness and a test of the rate at which they learned to decode novel words. The reliability, validity, and relative difficulty of each test were determined. A principal factor analysis with oblique rotation revealed that two highly related factors underlie phonemic awareness tests. A multiple regression analysis indicated that a combination of two tests, one related to each factor, has greater predictive validity for the beginning steps in reading acquisition than does any test alone. /// [French] Cette recherche avait pour but de vérifier la fidélité et la validité de tests utilisés pour opérationnaliser le concept de conscience phonémique. L'auteur a fait subir à 96 enfants de maternelle 10 tests de conscience phonémique et un test pour évaluer leur rapidité de décodage de mots nouveaux. La fidélité, la validité et la difficulté relative de chaque test ont ensuite été évaluées. Une analyse factorielle principale avec rotation oblique a révélé que les tests de conscience phonémique s'appuient sur deux facteurs étroitement reliés. L'analyse de régression multiple a par ailleurs, indiqué, qu'une combinaison deux à deux de deux tests, présente une plus grande validité prédictive des premières étapes de l'apprentissage de la lecture que chaque test pris isolément. /// [Spanish] El propósito de este estudio fue determinar la confiabilidad y validez de los tests que han sido usados para definir operacionalmente el concepto de alerta fonémica (phonemic awareness). A 96 niños de jardín se les administraron 10 tests de alerta fonémica y un test de la velocidad a la que aprendían a decodificar palabras nuevas. Se determinó la confiabilidad, validez y dificultad relativa de cada test. Un análisis factorial principal con rotación oblicua reveló que hay dos factores altamente relacionados que subyacen a los tests de alerta fonémica. Un análisis de regresión múltiple indicó que una combinación de dos tests, uno relacionado con cada factor, tiene mayor validez predictiva para los pasos iniciales en la adquisición de la lectura que cualquier test por sí solo. /// [German] Der zweck dieser Studie war, die Verläßlichkeit und Gültigkeit von Tests festzulegen, welche benutzt wurden, das Konzept phonemischer Erkenntnis einzusetzen. Neunundsechzig Fünfjährige wurden 10 phonemische Erkenntnis-Tests gegeben und ein Test auf die Schnelligkeit hin, mit der sie lernten, neue Wörter zu entschlüsseln. Die Verläßlichkeit, Gültigkeit und verhältnismäßige Schwierigkeit von jedem Test wurden festgelegt. Eine Hauptfaktor-Analyse mit versteckter Rotation zeigte, daß zwei stark verbundene Faktoren phonemischen Erkenntnis-Tests zugrundeliegen. Eine Mehrfach-Regressions-Analyse zeigte auf, daß die Kombination von zwei Tests, jeweils auf einen der beiden Faktoren eingehend, eine bessere Voraussage-Gültigkeit für die Anfangsschritte beim Lesen hat als jeder Test für sich allein.
The effects of varying linguistic manipulations on the difficulty level of phonological awareness tasks were examined. Participants included 32 kindergarten and 35 first-grade students who were administered two alternate forms of five different phonological awareness tasks (Rhyme-Providing, Sound-Providing, Blending, Segmentation, and Initial Deletion). Items within the tasks were selected and ordered using the following linguistic manipulations: continuant vs. noncontinuant sounds, number of phonemes within a word, and number of phonemes within an initial consonant cluster. Results suggested that significant differences in item difficulty varied depending on the type of task. However, four out of five tasks had some combination of manipulations that was easier or more difficult than others. Implications for creating new or evaluating existing measures of phonological awareness are discussed.
This study examined the performances of 171 children in kindergarten through second grade on 11 tasks of phonological awareness. The purpose was to assess phonological awareness skill acquisition across age and type of task. Results provided support for an ordering of tasks by difficulty, or age of mastery, as follows: rhyme, alliteration, blending, segmentation, manipulation. Performance on all of the tasks demonstrated rapid growth in 6-year-old children (first grade), and most tasks were mastered (90% correct) by the age of 7 years. The study also supports previous work indicating certain variables (i.e., age, verbal ability, letter-sound knowledge, reading skill) that are significantly related to performance on the phonological measures. Knowing the age at which students typically master these tasks and the order in which they acquire these skills should enhance our ability to assess both normal and delayed phonological awareness performance and help to inform instructional practices.
In the present study, the nature of Dutch children's phonological awareness was examined throughout the elementary school grades. Phonological awareness was assessed using five different sets of items that measured rhyming, phoneme identification, phoneme blending, phoneme segmentation, and phoneme deletion. A sample of 1405 children from kindergarten through fourth grade participated. Results of modified parallel analysis and analyses within the context of item response theory (IRT) showed phonological awareness to be unidimensional across different tasks and grades. Despite the evidence for a single underlying ability, the cognitive task requirements for the various tasks were found to differ. In addition to some overlap between the item sets, those for rhyming, phoneme identification, and phoneme blending were easier than those for phoneme segmentation and phoneme deletion. The results lend support to the assumption that phonological awareness is a continuum of availability for phonological representations which can range from partial availability (i.e., access) to full availability (i.e., access).