Content uploaded by Richard G. Lomax
Author content
All content in this area was uploaded by Richard G. Lomax on Jul 01, 2015
Content may be subject to copyright.
151
Reading Research Quarterly, 50(2)
pp. 151–169 | doi: 10.1002/rrq.99
© 2015 International Literacy Association
ABSTRACT
The simple view of reading describes reading comprehension as the product
of decoding and listening comprehension and the relative contribution of each
to reading comprehension across development. We present a cross- sectional
analysis of first, second, and third graders ( N = 123–125 in each grade) to
assess the adequacy of the basic model. Participants completed multiple
measures to inform latent constructs of word reading accuracy, word reading
fluency, listening comprehension, reading comprehension, and vocabulary.
In line with previous research, structural equation models confirmed that
the influence of decoding skill decreased with increasing grade and that the
influence of listening comprehension increased. However, several additional
findings indicate that reading development is not that simple and support an
elaboration of the basic model: A strong influence of listening comprehen-
sion on reading comprehension was apparent by grade 2, decoding skill was
best measured by word and nonword reading accuracy in the early grades
and word reading fluency in grade 3, and vocabulary skills indirectly affected
reading comprehension through both decoding skill and listening comprehen-
sion. This new elaborated model, which provides a more comprehensive view
of critical influences on reading in the early grades, has diagnostic and in-
structional ramifications for improving reading pedagogy.
R eading comprehension is determined by a wide range of com-
ponent skills and processes (Kendeou, van den Broek, White,
& Lynch, 2009 ; Oakhill & Cain, 2012 ; Vellutino, Tunmer,
Jaccard, & Chen, 2007 ) making the specification of such models a
challenge. In this article, we evaluate a relatively simple but influen-
tial framework of reading comprehension, the simple view of read-
ing (Gough, Hoover, & Peterson, 1996 ; Hoover & Gough, 1990 ;
Snow, 2002 ). The original article (Hoover & Gough, 1990 ) has made
a substantial impact on the field of reading research, with more than
1,000 citations to date. However, as we outline herein, the simplicity
o f t h i s f r a m ew o r k ha s b e e n q u e s t io n e d i n r e c e nt y e a rs . U nd e r s t a nd i n g
the validity and adequacy of the original model compared with
more complex models is important for two reasons: In relation to
theory, the simple view has influenced the examination and expla-
nation of reading development and disability across languages; in
relation to practice, the simple view has informed both the diagnosis
of reading disability and educational practice.
The Simple View of Reading: An Overview
According to the simple view, the knowledge, skills, and processes that
determine reading comprehension are captured by two broad skill do-
mains: decoding and listening comprehension. Thus, according to this
view, reading comprehension is the product of an individual ’ s ability to
Language and Reading
Research Consortium
Learning to Read:
Should We Keep Things Simple?
152 | Reading Research Quarterly, 50(2)
read words and to understand texts that are presented
aurally. Gough and Tunmer ( 1986 ) used the label decod-
ing , rather than word recognition , to emphasize the im-
portance of letter–sound correspondence rules in the
reading of English. We prefer the term word recognition
to include the reading of words automatically through
sight because of repeated exposures, in addition to pho-
nological decoding skills (Ehri, 2014 ).
Gough and Tunmer ( 1986 ) defined decoding (here-
after word recognition ) as the ability to read isolated
single words “quickly, accurately and silently” (p. 7) and
listening comprehension as the ability to take lexical
information and derive sentence- and discourse- level
interpretations. There is broad agreement of the
importance of both skill sets to the prediction of reading
comprehension (de Jong & van der Leij, 2002 ; Megherbi,
Seigneuric, & Ehrlich, 2006 ; Muter, Hulme, Snowling,
& Stevenson, 2004 ; Oakhill, Cain, & Bryant, 2003 ;
Pazzaglia, Cornoldi, & Tressoldi, 1993 ). In addition,
twin studies demonstrate distinct genetic influences for
each (Olson, Keenan, Byrne, & Samuelsson, 2014 ).
The simple view makes two influential predictions
about reading development and difficulties. First, the rel-
ative weightings of these two components—word recog-
nition and listening comprehension—will change across
development, as the developing reader acquires faster and
more automatic word recognition skills (Gough etal.,
1996 ). Second, poor reading comprehension may arise be-
cause of difficulties in the domain of word recognition,
listening comprehension, or both (Gough & Tunmer,
1986 ). There is broad support for these two central claims.
Diachronic change in the influence of word recognition
and listening comprehension is evident in cross- sectional
studies, which show that the relation between word recog-
nition and reading comprehension lessens with chrono-
logical age and that the association between listening and
reading comprehension increases (Catts, Hogan, & Adlof,
2005 ; Gough et al., 1996 ). Struggling readers with core
deficits in word recognition, comprehension, or both have
been identified across a range of alphabetic orthographies
and educational settings (Cain & Oakhill, 2006b ; Catts,
Adlof, & Weismer, 2006 ; Catts, Hogan, & Fey, 2003 ;
Megherbi & Ehrlich, 2005 ; Protopapas, Simos, Sideridis,
& Mouzaki, 2012 ; Torppa, Tolvanen, Poikkeus, Leskinene,
& Lyytinen, 2007 ). Our focus here is on the validity of the
simple view for describing reading during the first few
years of formal instruction.
In recent years, a series of papers have proposed that
the simple view lacks complexity and, therefore, validity
in relation to both components: word recognition and
listening comprehension (Adlof, Catts, & Little, 2006 ;
Joshi & Aaron, 2000 ; Kirby & Savage, 2008 ; Tunmer &
Chapman, 2012 ). As such, researchers have argued that
additional components should be included. First, we de-
tail the reasons to expect developmental change between
grades in the prediction of reading comprehension and
our hypothesis testing approach to determine the point
of change. Second, we consider the alternative models
that have been proposed for both word recognition and
listening comprehension and outline our approach to
testing these competing theories.
Does the Simple View
Capture Changes in Reading
Development Over Time?
Whereas numerous studies support the basic premise of
the simple view that a combination of word recognition
and listening comprehension explains large amounts of
variance in reading comprehension, it is likely that the
relative contributions of these constructs change across
grades. Individual differences in word recognition
should be the primary contributor to reading compre-
hension in the first years of formal education because
the focus of instruction is on teaching students to read
words. In the later grades, however, the vocabulary,
grammar, and discourse demands of reading materials
increase, and word recognition skills become more
automatic. As a result, the simple view predicts that
students’ listening comprehension should account for
more unique variance in reading comprehension than
does word recognition in older age groups.
This pattern is confirmed by reviews of the literature
and a recent meta- analysis (Garcia & Cain, 2014 ; Gough
etal., 1996 ), but few studies have empirically confirmed
these observations. One empirical study that addressed
this developmental pattern used a series of regression
analyes to determine the unique influence of word recog-
nition and listening comprehension to reading compre-
hension in a longitudinal sample of students tested in
grades 2, 4, and 8 (Catts et al., 2005 ). The researchers
found that the unique contribution of word recognition
to reading comprehension decreased over time from 27%
in second grade, 13% in fourth grade, and finally 2% in
eighth grade. The reverse pattern was revealed for listen-
ing comprehension and reading comprehension; listen-
ing comprehension contributed 9% in second grade, 21%
in fourth grade, and 36% in eighth grade.
Catts et al. ’ s ( 2005 ) study, together with the reviews,
provides compelling evidence for the broad developmental
changes proposed by the simple view. However, the
method for measuring these changes was regression anal-
yses, which do not account for measurement error among
measures. Furthermore, in recent years, a series of studies
have demonstrated that the characteristics of assessments
can greatly influence the extent to which reading compre-
hension is predicted by word recognition (Cain & Oakhill,
2006a ; Cutting & Scarborough, 2006 ; Garcia & Cain, 2014 ;
Keenan & Betjemann, 2006 ; Keenan, Betjemann, & Olson,
Learning to Read: Should We Keep Things Simple? | 153
2008 ; Nation & Snowling, 1997 ). For example, some mea-
sures of reading comprehension may be more heavily de-
pendent on word recognition if a cloze format is used
where the selection of the appropriate word completion
requires discrimination among similarly spelled distrac-
tors (Cain & Oakhill, 2006a ; Keenan etal., 2008 ; Nation &
Snowling, 1997 ). Finally, none of the developmental tests
of the simple view to date have pinpointed the grade at
which the shift from word recognition to listening com-
prehension as the predominant predictor of reading com-
prehension occurs, because measures were not given in
concurrent grades but rather tested at grade intervals.
A key objective in our analyses was to determine
changes in relative contributions of word recognition
and listening comprehension to reading comprehension
in the early years of formal education, grades 1–3, using
structural equation modeling, which takes into account
measurement error.
Should Word Reading Fluency
Be Added to the Simple View?
Good word recognition is essential for accessing the
meaning of written text. As noted previously, Gough and
Tunmer ( 1986 ) defined word recognition as the ability to
read isolated single words “quickly, accurately and si-
lently” (p. 7). Clearly, that definition cannot be operation-
alized easily for research purposes: If words are not read
aloud, how does the researcher know whether they are
read accurately? Assessments of silent reading fluency in-
clude tests of comprehension, which are therefore not
pure measures of word recognition (Wagner, Torgesen,
Rashotte, & Pearson, 2010 ). The majority of studies have,
therefore, assessed word recognition using tasks that re-
quire participants to read words aloud. These research
studies differ in terms of the statistical procedures used to
assess the contribution of word recognition to reading
comprehension, but they also differ markedly in the tasks
used to measure word recognition. Some studies include
measures of word recognition that have real word stimuli,
whereas others use nonwords; some have used tasks that
involve reading single words, whereas others measure
reading words within a text; and some studies use tasks
that measure accuracy, others measure how quickly items
are read, and still others measure both.
A particular focus in recent years has been on the is-
sue of whether accuracy alone is sufficient or whether a
measure that also taps fluency is required. This is because
when word recognition is efficient—that is, when an indi-
vidual can rapidly retrieve accurate phonological and
meaning- based representations of written words—greater
cognitive resources are available for comprehension pro-
cesses (Perfetti, 2007 ). However, the nature of the best
measure of the word recognition component in the simple
view might change over the course of development and by
orthography. In the early stages of reading development,
word recognition will be slow and more error prone, thus
a measure of accuracy alone might be sufficiently sensi-
tive to capture variance in a sample; later in development,
a fluency measure may be a more sensitive indicator of
word recognition skills when accuracy is easily achieved
(Garcia & Cain, 2014 ). In orthographies that have a very
transparent relation between graphemes and phonemes
(the letters and the sounds in the spoken language that
they represent), accuracy is achieved quickly, and mea-
sures that assess fluency have a greater inf luence on read-
ing outcomes early on (Florit & Cain, 2011 ).
Fluency is a complex construct that has been defined
as “a level of accuracy and rate where decoding is rela-
tively effortless; where oral reading is smooth and accu-
rate with correct prosody; and where attention can be
allocated to comprehension” (Wolf & Katzir- Cohen,
2001 , p. 219). In practice, many standardized measures
of fluency comprise word lists (rather than connected
prose) and/or do not assess prosody. For that reason, the
definition of fluency (or efficiency ) that we follow in this
research “is the oral translation of text with speed and
accuracy” (Fuchs, Fuchs, Hosp, & Jenkins, 2001 , p. 239),
although we note that reading with accurate expression
or prosody is considered an essential part of fluency
(Arcand etal., 2014 ; Kuhn, Schwanenf lugel, & Meisinger,
2010 ; Young & Bowers, 1995 ).
Fluency of word reading has been measured in differ-
ent ways. Some measures can be considered proxy mea-
sures of word reading fluency because the stimuli that are
processed are not words, such as recording how long it
takes to name a given array of (drawn) objects (Johnston &
Kirby, 2006 ) or letters (Joshi & Aaron, 2000 ; Silverman,
Speece, Harring, & Ritchey, 2013 ). Other measures directly
assess word reading f luency but differ as to whether the
stimuli are presented out of context (words or nonwords in
a list; Adlof etal., 2006 ; Kershaw & Schatschneider, 2012 ;
Protopapas etal., 2012 ; Silverman et al., 2013 ) or in con-
nected prose (Adlof etal., 2006 ; Høien- Tengesdal & Høien,
2012 ; Kershaw & Schatschneider, 2012 ; Silverman et al.,
2013 ). The fluency index can either be calculated as read-
ing speed (words per minute) or reading time (seconds per
correct word or passage), and some measures take into
account word recognition accuracy as well.
The extant literature does not provide a consistent
picture of the role of word reading fluency in the predic-
tion of reading comprehension. When naming speed for
objects is used as the index, it explains a small, but signifi-
cant, proportion of variance in unselected samples of
fourth and fifth graders but does not predict variance ad-
ditional to accuracy of word reading for third graders
(Johnston & Kirby, 2006 ). In contrast, letter naming speed
explains significant additional variance in third graders’
reading comprehension outcomes (Joshi & Aaron, 2000 ).
Learning to Read: Should We Keep Things Simple? | 159
all three models and across all three grades. For reading
comprehension, construct reliability ranged from .81 to
.89. Construct reliability ranged from .89 to .98 for word
recognition, from .76 to .79 for listening comprehen-
sion, and from .86 to .91 for vocabulary.
The first step in evaluating a model is to assess the
fit between the theoretical model and the sample data.
Because no single global fit index has been deemed accept-
able (Schumacker & Lomax, 2010 ), we used four measures
of model fit to make a more informed judgment than with
a single index. The standardized root mean square residual
index considers the residuals between the observed and the
model- implied covariance matrices. Values below .08 are
deemed as indicating acceptable model fit. The compara-
tive fit index, normed fit index, and nonnormed fit index
are each scaled from 0 to 1, with values over .90 being con-
sidered acceptable. Finally, a vote count was taken across
the four indexes to determine overall model fit.
For each theoretical model, estimates from the stan-
dardized solution are presented. Estimates of the factor
loadings (relating the observed to latent variables in the
measurement model) and the structure coefficients (relat-
ing the latent variables to one another from the structural
model; these are analogous to standardized β weights in
regression analysis) are shown at the top of Tables 3–5 and
the global fit indexes at the bottom.
Does the Simple View ’ s Basic Model
Provide a Good Estimation of Reading
Comprehension in Grades 1
–
3?
As shown in Figure 1 , the first theoretical model hy-
pothesizes that word recognition and listening compre-
hension together influence reading comprehension. As
shown in Table 3 , all of the factor loadings and struc-
ture coefficients were significantly different from zero
( p <.05) and in the expected direction (i.e., positive) for
every grade. Model fit was also deemed to be acceptable
for each grade. The R 2
statistic is additionally presented
for each grade to indicate that around 90% of the vari-
ance in reading comprehension was explained by the
model. These data demonstrate that the basic model of
the simple view provides a good estimation of reading
comprehension in grades 1–3.
Does the Influence of Word Recognition
and Listening Comprehension on Reading
Comprehension Change Across Grades?
To address this question, we need to examine the structure
coefficients across the grades (see Figure 1 and Table 3 ). In
grade 1, word recognition had a much stronger influence
on reading comprehension than did listening comprehen-
sion. This consecutive- grade cross- sectional approach
helps identify when the shift from word recognition to lis-
tening comprehension occurs. There was a shift beginning
in grade 2, such that listening comprehension had a much
stronger influence on reading comprehension than word
recognition, with the same pattern apparent in grade 3.
Th is is consistent with prev ious resea rch (Catts etal ., 2005 )
indicating the increasing automaticity of word recognition
and the emergence of listening comprehension as the pre-
dominant predictors of reading comprehension as literacy
develops.
Do Accuracy and Fluency of Word
Recognition Make Separable
Contributions to the Determination
of Reading Comprehension?
Figure 2 displays a second theoretical model in which
word reading fluency and word reading accuracy latent
variables were split out from the word recognition latent
variable described previously. This theoretical model pos-
its that word recognition accuracy, word recognition
fluency, and listening comprehension each independently
influence reading comprehension. As shown in Table 4 , all
of the factor loadings and most of the structure coefficients
were significantly different from zero ( p <.05) and in the
expected direction (i.e., positive) across grades. Global
model fit was acceptable for each grade. The R 2
statistic is
presented for each grade, where again around 90% of the
variance in reading comprehension was explained by the
model. Most notable are the differential structure coeffi-
cients across the grades. First, as in the single- construct
word recognition model, the strength of the inf luence of
listening comprehension increased after grade 1. Next, the
influence of word reading accuracy decreased from grade
1, becoming nonsignificant by grade 3. Finally, the influ-
ence of word reading fluency was significant only in grade
3. Once again, there is a shift as students’ literacy skills de-
velop. That is, first graders are more reliant on accuracy
because their word reading fluency is still developing (and
continues to do so for several more years).
FIGURE 1
Simple View of Reading Models Across Grades 1
–
3
Note . Standardized estimates from the structural model, where
“.22 (.57) .60” represents the results for the first, second, and
third grades, respectively. All paths are significantly different from
zero ( p < .05).
Learning to Read: Should We Keep Things Simple? | 155
Current Study
In the current study, we used structural equation mod-
eling, within the framework of the simple view of read-
ing, to examine the relation between word recognition
(both accuracy and f luency) and vocabulary, listening
comprehension, and reading comprehension in a large,
cross- sectional sample of students in grades 1–3. The
following research questions guided our study:
1 . Does the basic model of the simple view of read-
ing, including only word recognition and listen-
ing comprehension, provide a good estimation of
reading comprehension in grades 1–3?
2 . Does the inf luence of word recognition and
listening comprehension on reading comprehen-
sion change across grades?
3 . Do accuracy and fluency of word recognition
make separable contributions to the determina-
tion of reading comprehension?
4 . Does vocabulary improve the prediction of read-
ing comprehension, and if so, is vocabulary ’ s
influence through word recognition, listening
comprehension, or both?
Method
Participants
The partic ipants in t his study were part of a larger compre-
hensive longitudinal investigation of reading and listening
comprehension in preschool to third- grade students. The
current sample included all participants in grades 1–3
during the initial year of that study. There were 125 first
graders, 123 second graders, and 123 third graders. Table 1
shows the mean age, income status, gender, ethnicity, per-
centage receiving free or reduced- price lunch, and special
education status of students at each grade and whether
English was the home language. These data indicate that
our sample was racially and ethnically diverse, and
included children with Individualized Education Plans
and from families living below the poverty level.
Students were selected from four research sites in
different regions of the United States, with each site
responsible for approximately the same number of stu-
dent s at e ach gra de. Acro ss res ear ch sites , scho ol dis tri cts
were selected based on size and diversity of the student
populations, as well as willingness to participate in the
project. Once districts (and principals) agreed to partici-
pate, cooperating teachers in the relevant grades re-
ceived recruitment packets to send home for all students
in their class. Among those children whose parents con-
sented to participation, we randomly selected approxi-
mately 32 students per site per grade to receive our
assessment battery.
Measures
Our assessment battery included multiple measures of
reading comprehension, word recognition, listening com-
prehension, and vocabulary. Each of the measures of these
constructs is described next. The assessments were admin-
istered in the latter half of the school year (January–May).
Reading Comprehension
Three measures of reading comprehension were
administered. The Gates–MacGinitie Reading Tests
(MacGinitie, MacGinitie, Maria, & Dreyer, 2000 ) have
different levels of the reading comprehension subtest
for our three different age groups. For first graders, the
level 1 passage comprehension task was administered,
and second graders received level 2. The written pas-
sage is presented in units of one or more sentences, and
from four corresponding pictures, students select the
one that matches the meaning of the sentences. The
grade 3 students completed the level 3 materials in
which the passage is presented as a whole, and are
required to answer questions (with multiple- choice re-
sponses) after each one. Students were given 35 minutes
to complete the task, and the score was the number of
items correctly selected. The internal consistency
(Cronbach ’ s α) for our sample was good across grades
1–3:.89, .82, and .91, respectively.
The passage comprehension subtest from the
Woodcock Reading Mastery Tests–Revised – Normative
Update (WRMT–R/NU; Woodcock, 1997 ) was also ad-
ministered to assess students’ reading comprehension.
This measure employed a cloze procedure in which
TABLE 1
Selected Baseline Student Characteristics
Characteristic Grade 1 Grade 2 Grade 3
N 125 123 123
Age (baseline 2010) 6.56 (0.34) 7.53 (0.35) 8.58 (0.38)
Family income (categorical)
● % <$40K 17.6 26.0 13.8
● % $41K to <$80K 25.6 22.8 29.3
● % >$80K 45.6 43.9 49.6
% female 57 48 54
% white 81 86 75
% receiving free or
reduced- price lunch
16 26 17
% with Individualized
Education Plans
7 6 7
% English home
language
78 86 77
Note . For age, standard deviations are provided in parentheses.
156 | Reading Research Quarterly, 50(2)
students read a short passage with one or more words
missing and were required to provide the missing
word(s). The internal consistency (Cronbach ’ s α) for
our sample from grades 1–3 was good:.91, .87, and .89,
respectively.
We also administered an experimental measure,
the Reading Comprehension Measure (RCM), which
was adapted in part from the fifth edition of the
Qualitative Reading Inventory (QRI–5; Leslie &
Caldwell, 2010 ). The RCM assessed students’ abilities
to read, comprehend, and answer inferential and non-
inferential questions about narrative and expository
passages. Students read the passages silently and noti-
fied the examiner when each passage had been read.
The examiner then asked sets of open- ended questions
for each passage. First graders read one expository and
two narrative passages, whereas second and third
graders read two expository and two narrative pas-
sages. None of the passages overlapped. Five passages
came from the QRI–5, and the remainder was created
specifically for this project. These passages matched
the grade- appropriate passages from the QRI–5 in
terms of length and Lexile levels. Students’ responses
to administered questions were audiotaped. Trained
examiners scored each audiotaped response based on a
rubric of acceptable answers. The total number of cor-
rect responses served as the raw score. Approximately
10% of the sample from each grade was scored by a
second examiner, and the inter- rater reliability was
.93. The internal consistency (Cronbach ’ s α) for our
sample from grades 1–3 was adequate:.77, .77, and .80,
respectively.
Word Recognition Accuracy
Two measures of word recognition accuracy were ad-
ministered: two subtests from the WRMT–R/NU. The
word identification subtest measured students’ ability
to accurately pronounce printed English words rang-
ing from high to low frequency of occurrence. The in-
ternal consistency (Cronbach ’ s α) for our grades 1–3
sample was high: .96, .93, and .93, respectively. The
word attack subtest assessed students’ ability to read
pronounceable nonwords that increased in complexity
(a greater number of syllables). The internal consis-
tency (Cronbach ’ s α) for our sample was high:.92, .91,
and .92, respectively.
Word Recognition Fluency
Two subtests of the Test of Word Reading Efficiency–
Second Edition (TOWRE–2; Torgesen, Wagner, &
Rashotte, 2012 ) were administered to measure word
reading fluency. The sight word subtest measured how
many printed English words, which ranged from high to
low frequency of occurrence, students could accurately
pronounce in 45 seconds. The phonemic decoding sub-
test assessed how many pronounceable nonwords, which
varied in complexity, students could accurately pro-
nounce in 45 seconds. We did not repeat the administra-
tion of the assessment to our sample, so we report
reliabilities from the test manuals. The average test–
retest reliability for the sight word efficiency subtest
reported in the manual is .93 for grades 1–3. The sample
reliability for the phonemic decoding subtest is .91.
We also administered a third measure of word read-
ing fluency in context, adapted from the Florida
Assessment for Instruction in Reading: Oral Reading
Fluency (ORF; State of Florida, 2009 ). Students read two
passages aloud for up to 60 seconds (when the assessor
stopped the reading if not completed). Students were
forewarned that they would be asked a comprehension
question after each story. This instruction was to en-
courage reading for meaning. Words read accurately per
minute was calculated for each passage, and a fluency
score for each student was obtained from the tables pro-
vided online by the Florida Center for Reading Research
( www.fcrr.org/lookup ). As with the TOWRE–2, we did
not repeat the administration of this assessment, so we
could not compute test–retest reliability for our sample.
The published item response theory precision estimates
(using a scale similar to that used for α coefficients) are
consistently above .85, which is good.
Listening Comprehension
Three measures were used to assess listening compre-
hension. The Test of Narrative Language–Receptive
(TNL–R; Gillam & Pearson, 2004 ) assessed students’
ability to listen to three passages read aloud and answer
open- ended questions pertaining to the passages.
Students also completed the expressive components of
this measure, but the data from these components were
not used in this study. The measure was administered
according to test procedures with one exception: Prior to
answering questions for the second expressive passage,
students were required to retell the passage. This retell
was used for other studies within the larger project.
Students’ responses to test items were audiotaped, tran-
scribed, and scored as correct or incorrect. The total
number of correct responses served as the raw score. The
internal consistency (Cronbach ’ s α) for our sample from
grades 1–3 was adequate:.69, .73, and .58, respectively.
We also administered a modified version of the un-
derstanding paragraphs subtest of the Clinical Evaluation
of Language Fundamentals–Fourth Edition (CELF–4;
Semel & Wiig, 2006 ). This measure assessed students’
ability to listen to spoken paragraphs of increasing length
and complexity, understand oral narrative, and answer
questions that tap a range of different skills, including in-
ference making, story- relevant general knowledge, and
Learning to Read: Should We Keep Things Simple? | 157
accurate memory of the information presented. Similar
to the CELF–4 ’ s original version, students listened to
paragraphs read by the assessor and responded to sets of
open- ended questions. Adaptations for our project in-
cluded using two test paragraphs for each grade instead
of administering three paragraphs based on students’ age
in the CELF–4. Also, passages in the CELF–4 were the
same for grades 1–3, but in our version, one test passage
at each grade overlapped with a passage at the preceding/
proceeding grade level. These modifications allowed this
subtest to be administered to a wider age range of stu-
dents and also decreased the amount of administration
time per student. Regardless of grade, all students
answered a total of 10 questions; their responses were au-
diotaped, transcribed, and scored as correct or incorrect.
The total number of correct responses was tallied as the
raw score. Intra- class correlations for our modified
measure of the understanding paragraphs subtest of the
CELF–4 range from .98 to 1.00. Reliability (Cronbach ’ s α)
for our sample was poor (.01–.54 for each story).
This was an experimental measure, and it was the
first time being used as such. Exploratory factor analyses
resulted in a four- to five- factor structure (depending on
the grade), which could explain the low reliability. Because
we were using this (and other measures) to inform latent
constructs, we included additional reliability checks per
construct. For the constructs where understanding spo-
ken paragraphs was included, the reliability for the con-
struct including this measure was above benchmark
values (see the Structural Equation Models subsection in
the Results section for further details), so we decided to
retain this measure in the analyses.
We also administered an experimental measure, the
Listening Comprehension Measure (LCM), which was
adapted in part from the QRI–5. This measure was similar
to the RCM in general format but assessed listening rather
than reading comprehension. Specifically, it assessed stu-
dents’ abilities to listen, comprehend, and answer inferen-
tial and noninferential questions about spoken narrative
and expository passages. Students listened to paragraphs
read aloud by the assessor and responded to sets of open-
ended questions for each. First graders heard one exposi-
tory and two narrative passages, whereas second and third
graders were presented with two expository and two nar-
rative passages. None of the paragraphs overlapped. Seven
passages came directly from the QRI–5, and the remain-
der was created specifically for this project. These passages
matched the grade- appropriate passages from the QRI–5
in terms of length and Lexile levels. Students’ responses to
administered questions were audiotaped. Trained examin-
ers scored each audiotaped response based on a rubric of
acceptable answers. The total number of correct responses
served as the raw score. Approximately 10% of the sample
from each grade was scored by a second examiner, and the
inter- rater reliability was .91. The internal consistency
(Cronbach ’ s α) for our sample from grades 1–3 was ade-
quate:.65, .75, and .83, respectively.
Vocabulary
Three measures of vocabulary were administered. The
fourth edition of the Peabody Picture Vocabulary Test
(PPVT–4; Dunn & Dunn, 2007 ) assessed students’ rec-
ognition of the meanings of spoken words. The exam-
iner read a list of target words aloud, and the students
selected one of four pictures that corresponded to the
meaning of the target word. Test procedures for estab-
lishing a basal and ceiling were followed. The internal
consistency of the PPVT–4 for our sample was high:.95
for all three grades.
We also administered the second edition of the
Expressive Vocabulary Test (EVT–2; Williams, 2007 ).
For this measure, students were required to provide a
single word or synonym for the target word when shown
a picture. Procedures for establishing a ceiling and basal
were followed. The internal consistency for our grades
1–3 sample was high:.94, .93, and .95, respectively.
We also administered the word classes 1 and 2 subtest
from the CELF–4. This subtest assessed students’ abilities
to understand relationships between words that are related
by semantic class features and to verbally express the simi-
larities and differences between those relationships. This
subtest contained receptive and expressive components.
For the receptive component, students listened to three or
four words and chose two that were related. For the ex-
pressive component, students described the relationship
between the two words they chose. We administered the
word classes 1 subtest to first and second graders and word
classes 2 to third graders. Students’ responses for the ex-
pressive component were audiotaped and postscored. The
total numbers of correct responses were tallied for the re-
ceptive component, the expressive component, and the
two components combined. The internal consistency
(combined across receptive and expressive) for our sample
from grades 1–3 was good:.91, .94, and .84, respectively.
Procedures
All measures were administered by trained research
staff in a quiet room within the student ’ s school, local
university site, community center, or home. Assessors
underwent comprehensive training, which included the
completion of online training modules (including quiz-
zes), and in- lab observations by supervising assessors to
ensure consistent measurement administration and
fidelity across sites. The full assessment battery took five
to six hours to complete, with measures administered in
prescribed blocks, each lasting 15–40 minutes. At two
testing sites, measures were administered during these
testing blocks in students’ schools. At the other two sites,
assessments were administered across one or two
158 | Reading Research Quarterly, 50(2)
weekend days. In the latter case, frequent breaks were
taken to ensure that students were attentive during test
administration. All measures were administered indi-
vidually except for the Gates–MacGinitie, which was
administered in small groups or individually, where
necessary.
Results
Descriptive Statistics
Table 2 reports descriptive statistics for each of our
measures. In general, the assessments measure a range
of abilities and increase in difficulty across grades. The
correlations by grade between our variables are pro-
vided in the Appendix online.
Structural Equation Models
We used structural equation modeling (LISREL 9.1;
Joreskog & Sorbom, 2012 ) to assess the relations among
variables in three different theoretical models. The two
typical submodels in structural equation modeling were
used: the measurement model and the structural model.
The measurement model, actually a confirmatory factor
analysis model, specifies the relationships between the
observed (or measured) variables and their underlying
unmeasured latent variables. The use of multiple mea-
sures of each latent variable takes measurement error into
account, thereby resulting in better assessment of each
latent variable. This is a strength of structural equation
modeling over other methods that are reliant on single
measures and cannot take measurement error into ac-
count, such as regression analysis. The structural model
specifies the directed relations among the latent variables,
as shown in Figures 1–3 , for each theoretical model.
Reliabilities for latent variables as used in structural
equation modeling were calculated using Hancock ’ s co-
efficient H (Hancock & Mueller, 2011 ), which captures
the reliability for latent constructs. The recommended
cutoff value for Hancock ’ s coefficient H is .70.
Coefficient H was calculated for each construct across
TABLE 2
Means (and standard deviations) by Grade for Observed Variables
Observed variable Grade 1 Grade 2 Grade 3
Reading comprehension
Gates
–
MacGinitie 30.19 (6.81) 30.73 (5.40) 32.73 (9.48)
WRMT
–
R/NU passage comprehension 25.26 (7.72) 31.64 (6.45) 36.49 (6.52)
Reading Comprehension Measure 10.24 (3.13) 20.58 (4.65) 19.13 (4.68)
Word reading accuracy
WRMT
–
R/NU word identification 49.85 (12.45) 59.70 (9.17) 68.53 (9.91)
WRMT
–
R/NU word attack 21.14 (8.45) 25.42 (8.38) 29.97 (7.86)
Word reading fluency
TOWRE
–
2 sight word 45.63 (14.18) 56.85 (10.18) 63.50 (10.96)
TOWRE
–
2 phonemic decoding 20.23 (10.47) 25.07 (9.43) 31.20 (11.70)
Florida Assessment for Instruction in Reading:
Oral Reading Fluency
79.87 (34.33) 107.55 (37.17) 136.36 (37.70)
Listening comprehension
Test of Narrative Language
–
Receptive 26.73 (4.29) 28.90 (4.71) 30.76 (3.51)
CELF
–
4 understanding spoken paragraphs 6.35 (1.12) 6.22 (1.77) 6.51 (2.03)
Listening Comprehension Measure 11.25 (2.37) 19.53 (4.44) 20.74 (5.47)
Vocabulary
Peabody Picture Vocabulary Test, fourth edition 129.23 (17.01) 137.56 (16.59) 151.12 (17.03)
Expressive Vocabulary Test, second edition 96.98 (14.05) 105.29 (13.57) 113.80 (14.37)
CELF
–
4 word classes 34.07 (4.21) 36.05 (3.53) 17.49 (5.50)
Note. CELF
–
4 = Clinical Evaluation of Language Fundamentals
–
Fourth Edition; TOWRE
–
2: Test of Word Reading Efficiency
–
Second Edition;
WRMT
–
R/NU: Woodcock Reading Mastery Tests
–
Revised
–
Normative Update.
Learning to Read: Should We Keep Things Simple? | 159
all three models and across all three grades. For reading
comprehension, construct reliability ranged from .81 to
.89. Construct reliability ranged from .89 to .98 for word
recognition, from .76 to .79 for listening comprehen-
sion, and from .86 to .91 for vocabulary.
The first step in evaluating a model is to assess the
fit between the theoretical model and the sample data.
Because no single global fit index has been deemed accept-
able (Schumacker & Lomax, 2010 ), we used four measures
of model fit to make a more informed judgment than with
a single index. The standardized root mean square residual
index considers the residuals between the observed and the
model- implied covariance matrices. Values below .08 are
deemed as indicating acceptable model fit. The compara-
tive fit index, normed fit index, and nonnormed fit index
are each scaled from 0 to 1, with values over .90 being con-
sidered acceptable. Finally, a vote count was taken across
the four indexes to determine overall model fit.
For each theoretical model, estimates from the stan-
dardized solution are presented. Estimates of the factor
loadings (relating the observed to latent variables in the
measurement model) and the structure coefficients (relat-
ing the latent variables to one another from the structural
model; these are analogous to standardized β weights in
regression analysis) are shown at the top of Tables 3–5 and
the global fit indexes at the bottom.
Does the Simple View ’ s Basic Model
Provide a Good Estimation of Reading
Comprehension in Grades 1
–
3?
As shown in Figure 1 , the first theoretical model hy-
pothesizes that word recognition and listening compre-
hension together influence reading comprehension. As
shown in Table 3 , all of the factor loadings and struc-
ture coefficients were significantly different from zero
( p <.05) and in the expected direction (i.e., positive) for
every grade. Model fit was also deemed to be acceptable
for each grade. The R 2
statistic is additionally presented
for each grade to indicate that around 90% of the vari-
ance in reading comprehension was explained by the
model. These data demonstrate that the basic model of
the simple view provides a good estimation of reading
comprehension in grades 1–3.
Does the Influence of Word Recognition
and Listening Comprehension on Reading
Comprehension Change Across Grades?
To address this question, we need to examine the structure
coefficients across the grades (see Figure 1 and Table 3 ). In
grade 1, word recognition had a much stronger influence
on reading comprehension than did listening comprehen-
sion. This consecutive- grade cross- sectional approach
helps identify when the shift from word recognition to lis-
tening comprehension occurs. There was a shift beginning
in grade 2, such that listening comprehension had a much
stronger influence on reading comprehension than word
recognition, with the same pattern apparent in grade 3.
Th is is consistent with prev ious resea rch (Catts etal ., 2005 )
indicating the increasing automaticity of word recognition
and the emergence of listening comprehension as the pre-
dominant predictors of reading comprehension as literacy
develops.
Do Accuracy and Fluency of Word
Recognition Make Separable
Contributions to the Determination
of Reading Comprehension?
Figure 2 displays a second theoretical model in which
word reading fluency and word reading accuracy latent
variables were split out from the word recognition latent
variable described previously. This theoretical model pos-
its that word recognition accuracy, word recognition
fluency, and listening comprehension each independently
influence reading comprehension. As shown in Table 4 , all
of the factor loadings and most of the structure coefficients
were significantly different from zero ( p <.05) and in the
expected direction (i.e., positive) across grades. Global
model fit was acceptable for each grade. The R 2
statistic is
presented for each grade, where again around 90% of the
variance in reading comprehension was explained by the
model. Most notable are the differential structure coeffi-
cients across the grades. First, as in the single- construct
word recognition model, the strength of the inf luence of
listening comprehension increased after grade 1. Next, the
influence of word reading accuracy decreased from grade
1, becoming nonsignificant by grade 3. Finally, the influ-
ence of word reading fluency was significant only in grade
3. Once again, there is a shift as students’ literacy skills de-
velop. That is, first graders are more reliant on accuracy
because their word reading fluency is still developing (and
continues to do so for several more years).
FIGURE 1
Simple View of Reading Models Across Grades 1
–
3
Note . Standardized estimates from the structural model, where
“.22 (.57) .60” represents the results for the first, second, and
third grades, respectively. All paths are significantly different from
zero ( p < .05).
160 | Reading Research Quarterly, 50(2)
Does Vocabulary Influence Both
Word Recognition and Listening
Comprehension in a Model Predicting
Reading Comprehension?
As shown in Figure 3 , the third theoretical model
hypothesizes that (a) vocabulary influences both
word recognition and listening comprehension,
(b) which in turn each influences reading compre-
hension. The purposes of testing this theoretical
model were to determine (a) what the indirect effects
of vocabulary are on reading comprehension and,
(b)in a subsequent model, whether there is a direct
effect of vocabulary on reading comprehension. In
the indirect effects model (see Table 5 ), all of the
factor loadings and structure coefficients were signi-
ficantly different from zero ( p < .05) and in the ex-
pected direction (i.e., positive) for every grade. Model
fit was also acceptable for each grade. Of particular
interest are the structure coefficients across the
grades. As previously shown in the word recogni-
tionmodel, in grade 1, word recognition had a much
stronger influence on reading comprehension than
did listening comprehension. As before, in grades 2
and 3, listening comprehension had a much stronger
influence on reading comprehension than word rec-
ognition. Vocabulary had a strong effect on both word
recognition and listening comprehension, which was
somewhat stronger for listening comprehension than
for word recognition at every grade.
TABLE 3
Word Recognition Model: Standardized Solution
a
and Global Fit Indexes
b
Latent variable Observed variable Grade 1 Grade 2 Grade 3
Factor loadings
Reading comprehension Gates
–
MacGinitie .82 .81 .92
WRMT
–
R/NU passage comprehension .91 .82 .76
Reading Comprehension Measure .63 .70 .74
Word recognition WRMT
–
R/NU word identification .95 .88 .88
WRMT
–
R/NU word attack .72 .71 .69
TOWRE
–
2 sight word .93 .81 .85
TOWRE
–
2 phonemic decoding .82 .84 .86
Florida Assessment for Instruction in Reading:
Oral Reading Fluency
.93 .90 .89
Listening comprehension Test of Narrative Language
–
Receptive .85 .73 .60
CELF
–
4 understanding spoken paragraphs .34 .65 .68
Listening Comprehension Measure .68 .76 .83
Structure coefficients
Word recognition → reading comprehension .81 .48 .48
Listening comprehension → reading comprehension .22 .57 .60
R
2
.93 .88 .86
Global fit indexes
CFI .97 .97 .97
SRMR .05 .07 .06
NFI .96 .95 .95
NNFI .96 .95 .95
Note . CELF
–
4 = Clinical Evaluation of Language Fundamentals
–
Fourth Edition; TOWRE
–
2: Test of Word Reading Efficiency
–
Second Edition; WRMT
–
R/NU:
Woodcock Reading Mastery Tests
–
Revised
–
Normative Update.
a
All parameter estimates are statistically different from zero ( p < .05).
b
Comparative fit index (CFI), standardized root mean square residual (SRMR), normed fit index (NFI), and nonnormed fit index (NNFI). Acceptable
model fit defined as root mean square error of approximation and SRMR < .08, and CFI, NFI, and NNFI > .90.
Learning to Read: Should We Keep Things Simple? | 161
In the direct effect model, although the specific
results are not shown here, the direct effect of vocabulary
on reading comprehension was not significant for any
grade (i.e., the direct effect was approxi matelyzero), and
the fit of this model significantly deteriorated from the
indirect effects model. Thus, wesee evidence that the in-
fluence of vocabulary onreading comprehension is only
an indirect effect through both word recognition and
listening comprehension.
Discussion
The simple view of reading describes reading compre-
hension as the product of word recognition and listening
comprehension. Over the past three decades, consider-
able evidence has supported this framework for the study
of reading development and the diagnosis of reading dif-
ficulties (e.g., Aaron, Joshi, & Williams, 1999 ; Catts etal.,
2003 ; Hoover & Gough, 1990 ). In this study, we asked,
generally, whether the simple view is too simple to ex-
plain the complexities of reading comprehension in the
early years of formal education. Our response to this
question is yes. Our study has many notable strengths,
including the comparison of consecutive age groups, the
use of multiple measures of each construct in our age-
appropriate assessment battery, and our hypothesis-
driven approach to model fitting. Using a cross- sectional
sample of students in g rades 1, 2 , and 3, we conf irmed t he
basic premise of the simple view using a thorough, devel-
opmentally appropriate test battery: Individual differ-
ences in reading comprehension were explained by a
student ’ s ability to read words and comprehend language.
Critically, our use of multiple measures provided a thor-
ough and comprehensive assessment of each construct,
and together, word recognition and listening compre-
hension explained approximately 90% of the variance in
reading comprehension at each grade. Clearly, our data
demonstrate that the basic model of the simple view of
reading provides a good estimation of reading ability in
these grades.
However, our findings also revealed important
developmental nuances that build on the basic model:
(a) Listening comprehension influences reading com-
prehension during the earliest stages of reading devel-
opment; (b) the transition point at which listening
comprehension becomes more prominent occurs early,
around grade 2; (c) word recognition is best measured
by word reading accuracy in the early grades and word
reading fluency in the later grades; and (d) vocabulary
skills affect reading comprehension indirectly through
both word recognition and listening comprehension.
Thus, our analyses provide a more comprehensive view
of reading development, the implications of which we
subsequently discuss.
The Influence of Listening
Comprehension on Reading
Comprehension Is Evident Early
and Increases Over Time
Our finding of a key role for listening comprehension
in the earliest stages of reading comprehension is
supported by independent evidence of distinct factors
representing the foundations of word recognition and
listening comprehension in 4–6- year- olds (Kendeou,
Savage, & van den Broek, 2009 ). A few key studies have
shown that word recognition explains a greater propor-
tion of variation in reading comprehension in the early
grades compared with listening comprehension, which
contributes more to reading comprehension in the later
grades (Catts etal., 2005 ; Garcia & Cain, 2014 ; Oakhill
& Cain, 2012 ). Our study adds to these findings by pin-
pointing the grade at which this shift occurs, thus ex-
tending our understanding of the diachronic change
predicted by the simple view.
Using structural equation modeling, we deter-
mined that word recognition began to contribute less
variance to reading comprehension as early as grade 2,
when listening comprehension began to account more
for individual differences in reading comprehension.
There are several likely reasons for this shift. First, as
subsequently discussed in more detail, word recogni-
tion skills are more automatic in older readers, thus
enabling listening comprehension to play a greater role
in the prediction of reading comprehension. Future
research with more transparent orthographies will
determine how early this shift is seen for readers of
these languages who typically acquire fluent word
recognition more easily.
FIGURE 2
Word Reading Fluency and Word Reading Accuracy Model
Note . Standardized estimates from the structural model, where
“.24 (.57) .61” represents the results for the first, second, and
third grades, respectively. All paths are significantly different from
zero ( p < .05), unless otherwise indicated by *.
162 | Reading Research Quarterly, 50(2)
Second, we need to consider change in instructional
focus and text complexity. At the onset of formal literacy
instruction, a key aim is to teach students to read words.
With this goal in mind, early texts include basic vocabu-
lary, grammar, and discourse structures that provide
practice in reading with a core vocabulary of easily
decodable words. Our data show that in grade 2, a shift
occurs in which individual differences in reading com-
prehension are more strongly related to a student ’ s lis-
tening comprehension than his or her word recognition.
It may be that the student ’ s basic word reading abilities
bootstrap him or her into more complex texts that, to
comprehend, rely on robust listening comprehension
skills. Thus, the simple view of reading could be ex-
panded to describe important developmental changes in
early reading comprehension.
These findings have ramifications for the diagnosis
of and intervention for poor readers. Catts etal. ( 2003 )
used the simple view to categorize poor readers in a lon-
gitudinal sample of students tested in grades K, 2, and
4. The researchers found that students’ word recogni-
tion and listening comprehension skills were relatively
stable from second to fourth grades but that listening
comprehension difficulties, as a direct associate of read-
ing comprehension, increased between grades 2 and 4.
The increasing influence of listening comprehension to
TABLE 4
Fluency and Accuracy Model: Standardized Solution
a
and Global Fit Indexes
b
Latent variable Observed variable Grade 1 Grade 2 Grade 3
Factor loadings
Reading comprehension Gates
–
MacGinitie .82 .80 .91
WRMT
–
R/NU passage comprehension .91 .82 .77
Reading Comprehension Measure .64 .73 .73
Word reading accuracy WRMT
–
R/NU word identification .99 .98 .99
WRMT
–
R/NU word attack .82 .81 .76
Word reading fluency TOWRE
–
2 sight word .93 .83 .85
TOWRE
–
2 phonemic decoding .84 .86 .85
Florida Assessment for Instruction in Reading:
Oral Reading Fluency
.93 .89 .89
Listening comprehension Test of Narrative Language
–
Receptive .85 .73 .61
CELF
–
4 understanding spoken paragraphs .34 .64 .68
Listening Comprehension Measure .68 .81 .83
Structure coefficients
Word reading accuracy → reading comprehension .64 .47 .21
*
Word reading fluency → reading comprehension .17
*
.05
*
.30
Listening comprehension → reading comprehension .24 .57 .61
R
2
.94 .92 .88
Global fit indexes
CFI .98 .97 .97
SRMR .05 .06 .06
NFI .96 .95 .94
NNFI .97 .95 .95
Note . CELF
–
4 = Clinical Evaluation of Language Fundamentals
–
Fourth Edition; TOWRE
–
2: Test of Word Reading Efficiency
–
Second Edition; WRMT
–
R/NU:
Woodcock Reading Mastery Tests
–
Revised
–
Normative Update.
a
All parameter estimates are statistically different from zero ( p < .05) except those denoted by *.
b
Comparative fit index (CFI), standardized root mean square residual (SRMR), normed fit index (NFI), and nonnormed fit index (NNFI). Acceptable
model fit defined as root mean square error of approximation and SRMR < .08, and CFI, NFI, and NNFI > .90.
Learning to Read: Should We Keep Things Simple? | 163
reading comprehension means that the impact of listen-
ing comprehension deficits on poor reading compre-
hension will result in late- emerging, but long- standing,
poor comprehenders (Catts, Compton, Tomblin, &
Bridges, 2012 ). We did not set out to test the prediction
of different forms of reading disability, specifically
students whose problems lie at either the word level
(dyslexia), comprehension level (hyperlexia), or both
(Gough & Tunmer, 1986 ). However, we anticipate that
these reader types exist in our sample. The findings of
Catts et al. ( 2012 ) and these current analyses suggest
that for diagnostic purposes, listening comprehension
should be included in the assessment for reading diffi-
culties. Such practice will enable intervention to target
the specific skill weakness: word reading, text compre-
hension, or both.
These findings also have implications for instruc-
tion and, in particular, recent changes in the literacy tar-
gets in both the United States and the United Kingdom.
Although a primary task of early formal education is to
teach students to read words, time would be well spent
also increasing listening comprehension skills to
improve reading comprehension long term as these
become an increasingly dominant force in reading for
meaning. The Common Core State Standards (National
Governors Association Center for Best Practices &
Council of Chief State School Officers, 2010 ) and the re-
vised U.K. national curriculum for English (Department
for Education, 2014 ) emphasize the need for students to
develop comprehension through experiences of a wide
variety of genres for a range of purposes. This enables
students to build knowledge through reading and to
acquire and develop the skills needed to succeed not
only in school but also in further education and employ-
ment. Our findings are in line with views that listening
comprehension, as a valid and strong predictor of read-
ing comprehension, can be the vehicle, for both young
beginner readers and those who continue to struggle, to
acquire age- appropriate word recognition skills to access
and learn how to process these more challenging materi-
als to develop key comprehension and critical thinking
skills.
Word Recognition Is Best Measured
by Word Reading Fluency, not Word
Reading Accuracy, in Later Grades
Studies of the simple view of reading have measured word
recognition in varied ways. Some have measured word
recognition with single- word reading accuracy (Kershaw
& Schatschneider, 2012 ; Protopapas etal., 2012 ), and oth-
ers have quantified word recognition through word read-
ing in connected text (Adlof etal., 2006 ; Høien- Tengesdal
& Høien, 2012 ; Kershaw & Schatschneider, 2012 ), whereas
another approach has been to include the rate or ease with
which single words or connected prose is read aloud
(Adlof et al., 2006 ; Høien- Tengesdal & Høien, 2012 ;
Kershaw & Schatschneider, 2012 ), Our objective was to
determine whether word recognition was best character-
ized in the simple view by word recognition accuracy,
FIGURE 3
Vocabulary Model
Word
Recognition
Reading
Comprehension
Listening
Comprehension
Vocabulary
.68 (.51) .67 .81 (.50) .46
.83 (.93) .78 .26 (.61) .61
Note . Standardized estimates from the structural model, where “.83 (.93) .78” represents the results for the first, second, and third grades,
respectively. All paths are significantly different from zero ( p < .05).
164 | Reading Research Quarterly, 50(2)
word recognition fluency, or both. We found that both
were separable constructs in our sample of students learn-
ing English orthography in grades 1–3. This finding is in
line with Protopapas etal. ’ s study of young Greek readers
(although in contrast with Adlof et al. ’ s study of young
U.S. readers). Further, we determined that the nature of
the relation between word recognition and reading com-
prehension changes over time. In grades 1 and 2, individ-
ual differences in word recognition were best quantified
by accuracy measures. In contrast, for third graders, word
TABLE 5
Vocabulary Model: Standardized Solution
a
and Global Fit Indexes
b
Latent variable Observed variable Grade 1 Grade 2 Grade 3
Factor loadings
Reading comprehension Gates
–
MacGinitie .81 .81 .92
WRMT
–
R/NU passage comprehension .84 .76 .77
Reading Comprehension Measure .62 .71 .75
Word recognition WRMT
–
R/NU word identification .95 .88 .89
WRMT
–
R/NU word attack .72 .72 .69
TOWRE
–
2 sight word .93 .81 .84
TOWRE
–
2 phonemic decoding .82 .84 .85
Florida Assessment for Instruction in Reading:
Oral Reading Fluency
.93 .89 .88
Listening comprehension Test of Narrative Language
–
Receptive .82 .73 .63
CELF
–
4 understanding spoken paragraphs .37 .63 .67
Listening Comprehension Measure .72 .76 .84
Vocabulary Peabody Picture Vocabulary Test, fourth edition .84 .89 .86
Expressive Vocabulary Test, second edition .94 .82 .82
CELF
–
4 word classes .48 .52 .82
Structure coefficients
Word recognition → reading comprehension .81 .50 .46
Listening comprehension → reading comprehension .26 .61 .61
Vocabulary → word recognition .68 .51 .67
Vocabulary → listening comprehension .83 .93 .78
R
2
for reading comprehension .96 .91 .88
R
2
for word recognition .47 .27 .45
R
2
for listening comprehension .70 .87 .61
Global fit indexes
CFI .97 .96 .97
SRMR .06 .08 .05
NFI .95 .93 .95
NNFI .96 .94 .97
Note . CELF
–
4 = Clinical Evaluation of Language Fundamentals
–
Fourth Edition; TOWRE
–
2: Test of Word Reading Efficiency
–
Second Edition; WRMT
–
R/NU:
Woodcock Reading Mastery Tests
–
Revised
–
Normative Update.
a
All parameter estimates are statistically different from zero ( p < .05).
b
Comparative fit index (CFI), standardized root mean square residual (SRMR), normed fit index (NFI), and nonnormed fit index (NFI). Acceptable model
fit defined as root mean square error of approximation and SRMR < .08, and CFI, NFI, and NNFI > .90.
Learning to Read: Should We Keep Things Simple? | 165
reading fluency explained individual differences in com-
prehension, not word reading accuracy.
The convergence between our findings with an
English- speaking sample and those of Protopapas etal.
( 2012 ) with a Greek- speaking sample (as well as conver-
gence with the slightly older sample studied by
Silverman et al., 2013 ) suggest that the separability of
accuracy and fluency is not language dependent.
However, our finding that accuracy and fluency were
separable constructs is at odds with the study by Adlof
etal. ( 2006 ) that included a similar battery of word rec-
ognition measures but did not identify fluency as sepa-
rable from word recognition accuracy. One possible
reason for this discrepancy is that Adlof et al. ’ s mea-
sures of reading fluency for connected text and word
reading accuracy involved the same written materials,
which will overestimate the association of the two mea-
sures. In our study, fluency and reading comprehension
were assessed with separate stimuli. In addition, there
was some indication that in Adlof etal. ’ s sample,
f luency was tapping an additional construct, such as
processing speed, because a substantial proportion of
readers with poor fluency also had poor listening
comprehension.
Our finding that a different measure of word recog-
nition was the best predictor of reading comprehension
at different grades is not wholly consistent with the sim-
ple view of reading and clearly suggests that a more nu-
anced model of reading development is required. There
are several reasons for the change in inf luence of word
reading accuracy and f luency across development. First,
this may be explained by considering the changing na-
ture of word recognition in the early grades. When stu-
dents are first learning to read words, word reading is
slow and more error prone. Thus, a measure of accuracy
alone would be sufficient to capture variance. Later in
development, fluency would become a more sensitive
indicator of word recognition skills, when accuracy is
easily achieved by most students and words can be read
from memory (e.g., Ehri, 2014 ). This reasoning is sup-
ported by research on word recognition in languages
with transparent orthographies, which have a direct re-
lation between graphemes and phonemes (the letters
and the sounds in the spoken language that they repre-
sent). For such languages, accuracy is achieved quickly,
and measures that assess f luency have a greater influ-
ence on reading outcomes early on (Florit & Cain, 2011 ).
Thus, the relative influence of these two aspects of word
recognition may be language dependent, particularly in
young readers.
Another possibility for the shift to reading f luency
in predicting reading comprehension in grade 3 may be
related to our inclusion of a measure of fluency for con-
nected text. We note that performance on this measure
was above average (Hasbrouck & Tindal, 2006 ), but it
was in line with the slightly above- average standard-
ized scores for vocabulary (PPVT–4) and decoding
(TOWRE–2; means = 98–109). Words in context are
typically read faster than words in isolation because
word recognition can be facilitated through semantic
activation when in context (Jenkins, Fuchs, van den
Broek, Espin, & Deno, 2003 ). For this reason, f luency
may be a better metric of reading comprehension in
grade 3. However, contextual facilitation for word read-
ing is typically stronger for poorer readers than better
or older readers, who have superior decoding ability
and faster and more automatic sight word reading skills
(Nation & Snowling, 1998 ; Stanovich & West, 1979 ).
Critically, we are concerned with automaticity or effi-
ciency of retrieval of phonological and semantic pre-
sentations of words, not simply the ability to read text
faster. As noted in our introduction, broader defini-
tions of fluency include prosody or the expression with
which text is read aloud. Such definitions provide a
theoretical link between reading fluency and reading
comprehension (Kuhn et al., 2010 ) and should be
adopted in future research to test the role of fluency in
the simple view further.
Our results demonstrate the need to better repre-
sent developmental changes within the construct of
word recognition as related to reading comprehension
in the simple view of reading. Clearly, there is a complex
relation between fluency of word reading in context and
reading comprehension skill. Our findings are limited
because we were not able to identify the precise causal
mechanism for the fluency–comprehension relation-
ship. A priority of future research should be to directly
test whether the nature of the assessment of reading flu-
ency inf luences the strength of the relation between
word recognition and reading comprehension and
whether this differs between good and poor readers.
Vocabulary Skills Indirectly Affect
Reading Comprehension Through
Both Word Recognition and
Listening Comprehension
It is well established that individual differences in
vocabulary predict listening comprehension skills
(Nation & Snowling, 2004 ; Ouellette & Beers, 2010 ).
Studies have also shown a link between vocabulary
abilities and individual differences in word reading
(Mitchell & Brady, 2013 ; Nation & Snowling, 2004 ;
Ouellette, 2006 ). Using a latent construct approach,
Tunmer and Chapman ( 2012 ) found that the construct
of listening comprehension fed into reading compre-
hension directly, and also indirectly through its influ-
ence on word recognition. Although they did not test
whether vocabulary was the specific basis for this
mediating link, our results support that hypothesis. We
166 | Reading Research Quarterly, 50(2)
found that vocabulary indirectly predicted reading
comprehension: It influenced listening comprehension
and word recognition, which in turn predicted reading
comprehension. Moreover, our results showed that the
influence of vocabulary was stronger for listening com-
prehension than word recognition.
Students’ vocabulary knowledge differs widely
upon beginning formal literacy instruction (Hart &
Risley, 1995 ). Our results demonstrate how vocabulary
knowledge influences reading development in begin-
ner readers and, therefore, how it fits into the simple
view. Vocabulary knowledge is related to word recogni-
tion in at least two ways, because it reflects consoli-
dated knowledge about familiar individual word forms
and because a wide vocabulary supports the proces-
sing of unfamiliar words through strategies such
as eading by analogy (Ehri, 2014 ). In addition, vocabu-
lary knowledge might be related to reading ability
because it reflects students’ general language compe-
tence, which will inf luence reading development
(Nation & Snowling, 2004 ). Our data suggest a stronger
influence of vocabulary in the prediction of listening
comprehension than for word recognition (see also
Nation & Snowling, 2004 ) but does not identify a direct
relation to reading comprehension. This model con-
firms the important role of word knowledge for
sentence- and text- level processing, which is supported
by other work that demonstrates a critical role for
higher level language skills in early listening and read-
ing comprehension development (Florit, Roch, &
Levorato, 2011 ; Kendeou, van den Broek, et al., 2009 ;
Lepola, Lynch, Laakkonen, Silvén, & Niemi, 2012 ;
Oakhill & Cain, 2012 ).
Vocabulary may be important for comprehension
because words form the basis of sentences and longer
units of text and also because it specifically enables in-
tegration and inference making (Cain & Oakhill, in
press ; Kintsch & Rawson, 2005 ; Perfetti & Stafura,
2014 ). However, other forms of knowledge enable suc-
cessful comprehension as well. For example, knowledge
about text structure can influence comprehension by
providing a framework (Cain, 1996 ), topic knowledge
supports better comprehension of text (Compton,
Miller, Elleman, & Steacy, 2014 ), and instruction in
reading strategies improves reading comprehension
(Rosenshine & Meister, 1994 ). A broader range of
knowledge and its inf luence on reading comprehension
is a target for future research.
Strengths and Limitations
A strength of our study was the use of multiple mea-
sures to provide a comprehensive assessment of each
construct. This is one reason why our models typically
explained around 90% of the variance in reading
comprehension. With a single measure of a given con-
struct, there is measurement error in the system due to
reliability and/or validity issues. Using multiple mea-
sures of a construct in a latent variable isolates and takes
measurement error into account. As a result, the rela-
tions among the latent variables are likely to have more
explained variance. However, we note that there re-
mains around 10% of unexplained variance in the sys-
tem, which is likely due to factors such as instruction,
individual differences among the participants, and
other literacy constructs not included in our model.
We used a range of measures of the decoding con-
struct in the simple view and found that accuracy and
f luency measures were separable. A recent meta- analysis
supports this finding and further demonstrates that other
test characteristics inf luence the strength of the relation
between word recognition and reading comprehension
(see Garcia & Cain, 2014 , for a summary of different as-
sessments). Thus, based on our own work and the work of
our colleagues in the field, we do not believe that a single
measure of word recognition, listening comprehension, or
reading comprehension is best. Reading is complex, and
different assessments tap different things. This is illus-
trated in a recent study that examined the identification
of students with comprehension difficulties across a range
of standardized tests (Keenan & Meenan, 2014 ): Notably,
not all assessments identified the same students each
time. Together, these findings lead us to conclude that re-
searchers and practitioners should strive to take multiple
measures of target constructs.
In relation to the complexity of reading, we note
that we did not include other assessments of the broader
language and cognitive skills that might underpin both
word recognition and reading comprehension. Other
researchers have found that broader language skills
(e.g., semantic, morphological, grammatical; Catts
etal., 2006 ; Nation & Snowling, 2004 ; Tong, Deacon, &
Cain, 2014 ) and also cognitive skills (e.g., working
memory, executive functions; Cain, Oakhill, & Bryant,
2004 ; Locascio, Mahone, Eason, & Cutting, 2010 ) influ-
ence reading comprehension outcomes. We did not in-
clude independent assessments of these skills in our
analysis for two reasons. First, we focused on the exam-
ination of the inf luence of two additional components—
fluency and vocabulary—about which there has been
significant recent debate. Second, it is likely that all of
these language and cognitive processes contribute to
both reading and listening comprehension. As a result,
we believe that they all, in part, underpin the listening
comprehension construct in our model.
Finally, it is worth noting that our results should not
be taken to suggest that word reading is not a continu-
ing source of difficulty for some readers past grade 2:
Aproportion of poor readers have either specific diffi-
culties with word reading (i.e., dyslexia) or difficulties
Learning to Read: Should We Keep Things Simple? | 167
with both decoding and language comprehension that
continue across the lifespan. These individuals will
benefit from support beyond the early grades to develop
their word reading skills.
Conclusions
The simple view of reading explains the complex phe-
nomenon of reading comprehension by decomposing it
into word recognition and listening comprehension.
Our results support this broad framework, but our an-
swer to the question posed in our title is no. In line with
other research, our findings show that reading develop-
ment is not so simple and provide us with a more
comprehensive view of early reading development. The
simple view should be elaborated to include develop-
mental changes in its subcomponents, a more nuanced
view of word recognition, and indirect effects of vocab-
ulary. This more fine- grained view of reading develop-
ment has diagnostic and instructional ramifications for
improving reading pedagogy.
N O T E S
This paper was prepared by a task force of the Language and
Reading Research Consortium (LARRC) consisting of Kate Cain
(convener), Hugh Catts, Tiffany Hogan, and Richard Lomax.
LARRC project sites and investigators are as follows:
• Ohio State University (Columbus , OH): Laura M. Justice (site prin-
cipal investigator; SPI), Richard Lomax, Ann O ’ Connell, Jill
Pentimonti, Stephen A. Petrill,
1
and Shayne B. Piasta
• Arizona State University (Tempe, AZ): Shelley Gray (SPI) and
Maria Adelaida Restrepo
• Lancaster University (Lancaster, UK): Kate Cain (SPI)
• University of Kansas (Lawrence, KS): Hugh Catts 2
(SPI), Mi ndy
Bridges, and Diane Nielsen
• University of Nebraska–Lincoln (Lincoln, NE): Tiffany Hogan 3
(SPI), Jim Bovaird, and J. Ron Nelson
4
This work was supported by grant R305F100002 of the Institute
of Education Sciences ’ s Reading for Understanding Initiative. The
views presented in this work do not represent those of the federal
government, nor do they endorse any products or findings pre-
sente d her ein. We are deep ly gr atef ul t o the nume rous st aff , resea rch
associates, school administrators, teachers, students, and families
who participated. Key personnel at study sites include Garey Berry,
Beau Bevens, Jennifer Bostic, Shara Brink ley, L ori Chleborad, Dawn
Davis, Michel Eltschinger, Tamarine Foreman, Rashaun Geter, Sara
Gilliam, Miki Herman, Trudy Kuo, Gustavo Lujan, Carol Mesa,
Denise Meyer, Maria Moratto, Marcie Mutters, Trevor Rey, and
Stephanie Williams. We would like to thank Gloria Yeomans-
Maldonado and Jill Pentimonti in particular for their help with the
reliability analyses.
Correspondence concerning this work should be sent to Kate
Cain at k.cain@lancaster.ac.uk.
1 Petrill was a LARRC coinvestigator from 2010 to 2013.
2 Catts is now at Florida State University (Tallahassee, FL).
3 Hogan is now at the MGH Institute of Health Professions (Boston,
MA).
4 Nelson was a LARRC coinvestigator from 2010 to 2012.
REFERENCES
A a r o n , P . G . , J o s h i , M . , & W i l l i a m s , K . A . ( 1 9 9 9 ) . N o t a l l r e a d i n g
disabilities are alike . Journal of Learning Disabilities , 32 ( 2 ),
120 – 137 .
A d l o f , S . M . , C a t t s , H . W . , & L i t t l e , T . D . ( 2 0 0 6 ) . S h o u l d t h e s i m p l e
view of reading include a f luency component? Reading and
Wri ting , 19 ( 9 ), 933 – 958 . doi: 10.1007/s11145- 006- 9024- z
A r c a n d , M . S . , D i o n , E . , L e m i r e - T h é b e r g e , L . , G u a y , M . H . , B a r r e t t e ,
A . , G a g n o n , V. , … F u c h s , D . ( 2 0 1 4 ) . S e g m e n t i n g t e x t s i n t o m e a n -
ingful word groups: Beginning readers’ prosody and comprehen-
sion . Scientific Studies of Reading , 18 ( 3 ), 20 8 – 223 . doi: 10.108 0/108
88438.2013.864658
B r a z e , D . , T a b o r , W . , S h a n k w e i l e r , D . P . , & M e n c l , W . E . ( 2 0 0 7 ) .
Speaking up for vocabulary: Reading skill differences in young
adults . Journal of Learning Disabilities , 40 ( 3 ), 2 26 – 243 . do i: 10.117
7/00222194070400030401
Cain , K. ( 1996 ). Story knowledge and comprehension skill . In
C . C o r n o l d i & J . O a k h i l l ( E d s . ) , Reading comprehension diffi-
culties: Processes and remediation (pp. 167 – 192 ). Mahwah, NJ :
Erlbaum .
C a i n , K . , & O a k h i l l , J . ( 2 0 0 6 a ) . A s s e s s m e n t m a t t e r s : I s s u e s i n t h e m e a -
surement of reading comprehension . Br itish Journal ofEducational
Psychology , 76 ( 4 ), 697 – 708 . doi: 10.1348/000709905X69807
C a i n , K . , & O a k h i l l , J . ( 2 0 0 6 b ) . P r o f i l e s o f c h i l d r e n w i t h s p e c i f i c
reading c omprehension diff iculties . Br itish Journal of E ducational
Psychology , 76 ( 4 ), 683 – 696 . doi: 10.1348/00 0709905X67610
C a i n , K . , & O a k h i l l , J . ( i n p r e s s ) . R e a d i n g c o m p r e h e n s i o n a n d
vocabulary: Is vocabulary more important for some aspects of
comprehension? L ’ A n n é e Ps y c h o l ogi q u e .
C a i n , K . , O a k h i l l , J . V . , & B r y a n t , P . E . ( 2 0 0 4 ) . C h i l d r e n ’ s r e a d i n g
comprehension ability: Concurrent prediction by working mem-
ory, verbal ability, and component skills . Journal of Educational
Psychology , 96 ( 1 ) , 3 1 – 4 2 . d o i : 1 0 . 1 0 3 7 / 0 0 2 2 - 0 6 6 3 . 9 6 . 1 . 3 1
C a t t s , H . W . , A d l o f , S . M . , & W e i s m e r , S . E . ( 2 0 0 6 ) . L a n g u a g e d e f i c i t s
in poor comprehenders: A case for the simple view of reading .
Journal of Speech, Language, and Hearing Research , 49 ( 2 ), 278 –
293 . doi: 10.104 4/1092- 4388(2006/023
C a t t s , H . W . , C o m p t o n , D . L . , T o m b l i n , J . B . , & B r i d g e s , M . S .
( 2012 ). Prevalence and nature of late- emerging poor readers .
Journal of Educational Psychology , 10 4 ( 1 ), 16 6 – 181 . d oi: 10.1037/
a0025323
C a t t s , H . W . , H o g a n , T . P. , & A d l o f , S . M . ( 2 0 0 5 ) . D e v e l o p m e n t a l
changes in reading and reading disabilities . In H.W. Catts & A.G.
Kamhi (Eds.), The connections between language and reading dis-
abilities (pp. 25 – 40 ). Mahwah, NJ : Erlbaum .
C a t t s , H . W . , H o g a n , T . P . , & F e y , M . E . ( 2 0 0 3 ) . S u b g r o u p i n g p o o r
readers on the basis of individual differences in reading- related
abilities . Journal of Learning Disabilities , 36 ( 2 ), 151 – 16 4 .
doi: 10.1177/002221940303600208
C o m p t o n , D . L . , M i l l e r , A . C . , E l l e m a n , A . M . , & S t e a c y , L . M . ( 2 0 1 4 ) .
Have we forsaken reading theor y in the name of “quick fix” inter-
ventions for children with reading disability? Scientific Studies of
Reading , 18 ( 1 ) , 5 5 – 7 3 . d o i : 1 0 . 1 0 8 0 / 1 0 8 8 8 4 3 8 . 2 0 1 3 . 8 3 6 2 0 0
C u n n i n g h a m , A . E . , S t a n o v i c h , K . E . , & W i l s o n , M . R . ( 1 9 9 0 ) . C o g n i t i v e
variat ion in adult college st udents differi ng in reading abil ity . In T.H.
C a r r & B . A . L e v y ( E d s . ) , Reading and its development: Component
skills approaches (pp. 129 – 159 ). San Diego, CA : Academic .
C u t t i n g , L . E . , & S c a r b o r o u g h , H . S . ( 2 0 0 6 ) . P r e d i c t i o n o f r e a d i n g
comprehension: Relative contributions of word recognition, lan-
guage proficiency, and other cognitive skills can depend on how
comprehension is measured . Scientific Studies of Reading ,
10 ( 3 ),
277 – 299 . doi: 10.1207/s1532799xssr1003_5
d e J o n g , P . F . , & v a n d e r L e i j , A . ( 2 0 0 2 ) . E f f e c t s o f p h o n o l o g i c a l
abilities and linguistic comprehension on the development of
reading . Scientific Studies of Reading , 6 ( 1 ) , 5 1 – 7 7 . d o i : 1 0 . 1 2 0 7 /
S1532799XSSR0601_03
168 | Reading Research Quarterly, 50(2)
Department for Education . ( 2014 ). National curriculum in England:
English programmes of study . Retrieved from https://ww w.gov.uk/
government/publications/national-curriculum-in-england- english-
programmes-of-study/national-curriculum-in-england-english-
programmes-of -s tud y
D u n n , L . M . , & D u n n , D . M . ( 2 0 0 7 ) . Peabody Picture Vocabulary Test
(4th ed .). San Antonio, TX : Pearson .
Ehri , L.C. ( 2014 ). Orthographic mapping in the acquisition of
sight word reading, spelling memory, and vocabulary learn-
ing . Scientific Studies of Reading , 18 ( 1 ) , 5 – 2 1 . d o i : 1 0 . 1 0 8 0 /
10888438.2013.819356
F l o r i t , E . , & C a i n , K . ( 2 0 1 1 ) . T h e s i m p l e v i e w o f r e a d i n g : I s i t v a l i d
for different types of alphabetic orthographies? Educational
Psychology Review , 23 ( 4 ), 553 – 576 . doi: 10.1007/s10648-011-
9175- 6
F l o r i t , E . , R o c h , M . , & L e v o r a t o , M . C . ( 2 0 1 1 ) . L i s t e n i n g t e x t c o m p r e -
hension of explicit and implicit information in preschoolers: The
role of verbal and inferential skills . Discourse Processes , 48 ( 2 ),
119 – 138 . doi: 10.1080/0163853X.2010.494 24 4
F u c h s , L . S . , F u c h s , D . , H o s p , K . K . , & J e n k i n s , J . R . ( 2 0 0 1 ) . O r a l r e a d -
ing fluency as an indicator of reading competence: A theoretical,
empirical, and historical analysis . Scientific Studies of Reading ,
5 ( 3 ), 239 – 256 . doi: 10.1207/S1532799XSSR0503_3
G a r c i a , J . R . , & C a i n , K . ( 2 0 1 4 ) . D e c o d i n g a n d r e a d i n g c o m p r e h e n s i o n :
A meta- analysis to identify which reader and assessment charac-
teristics inf luence the strength of the relationship in English .
Review of Educational Research , 84 ( 1 ) , 7 4 – 1 1 1 . d o i : 1 0 . 3 1 0 2 /
00346 5 4 313499616
G i l l a m , R . B . , & P e a r s o n , N . A . ( 2 0 0 4 ) . Test of Narrative Language–
Receptive . Austin, TX : Pro-Ed .
G o u g h , P . B . , H o o v e r , W . A . , & P e t e r s o n , C . L . ( 1 9 9 6 ) . S o m e o b s e r v a -
t i o n s o n a s i m p l e v i e w o f r e a d i n g . I n C . C o r n o l d i & J . O a k h i l l
(Eds.), Reading comprehension difficulties: Processes and inter-
ventions (pp. 1 – 13 ). Mahwah, NJ : Erlbaum .
G o u g h , P . B . , & Tu n m e r , W . E . ( 1 9 8 6 ) . D e c o d i n g , r e a d i n g a n d r e a d -
ing disability . Remedial and Special Education , 7 ( 1 ), 6 – 10 .
doi: 10.1177/074193258600700104
H a n c o c k , G . R . , & M u e l l e r , R . O . ( 2 0 1 1 ) . T h e r e l i a b i l i t y p a r a d o x i n
assessing structural relations within covariance structure mod-
els . Educational and Psychological Measurement , 71 ( 2 ), 306 – 324 .
doi: 10.1177/0013164410384856
H a r t , B . , & R i s l e y , T . ( 1 9 9 5 ) . Meaningful differences in the everyday
experience of young American children . Baltimore, MD : Paul H.
Brookes .
H a s b r o u c k , J . , & T i n d a l , G . A . ( 2 0 0 6 ) . O r a l r e a d i n g f l u e n c y n o r m s : A
valuable assessment tool for reading teachers . The Reading
Tea c he r , 59 ( 7 ), 636 – 644 . doi: 10.1598/RT.59.7.3
H ø i e n - T e n g e s d a l , I . , & H ø i e n , T . ( 2 0 1 2 ) . T h e r e a d i n g e f f i c i e n c y
model: An extension of the componential model of reading .
Journal of Learning Disabilities , 45 ( 5 ), 4 67 – 479 . doi: 10.117 7/
0022219411432688
H o o v e r , W . A . , & G o u g h , P . B . ( 1 9 9 0 ) . T h e s i m p l e v i e w o f r e a d i n g .
Reading and Writing , 2 ( 2 ), 127 – 160 . doi: 10.1007/BF00401799
J e n k i n s , J . R . , F u c h s , L . S . , v a n d e n B r o e k , P . , E s p i n , C . , & D e n o , S . L .
( 2003 ). Sources of individual differences in reading comprehen-
sion and reading fluency . Journal of Educational Psychology ,
95 ( 4 ), 719 – 729 . doi: 10.1037/0022- 0663.95.4.719
J o h n s t o n , T . C . , & K i r b y , J . R . ( 2 0 0 6 ) . T h e c o n t r i b u t i o n o f n a m i n g
speed to the simple view of reading . Reading and Writing , 19 ( 4 ),
339 – 361 . doi: 10.1007/s11145- 005- 4644- 2
J o r e s k o g , K . G . , & S o r b o m , D . ( 2 0 1 2 ) . LISREL 9.1 . Lincolnwood, IL :
Scientific Software International .
J o s h i , R . M . , & A a r o n , P . G . ( 2 0 0 0 ) . T h e c o m p o n e n t m o d e l
of reading: Simple view of reading made a little more
complex . Reading Psychology , 21 ( 2 ) , 8 5 – 9 7 . d o i : 1 0 . 1 0 8 0 /
02702 710 05 0 084428
K e e n a n , J . M . , & B e t j e m a n n , R . S . ( 2 0 0 6 ) . C o m p r e h e n d i n g t h e G r a y
Oral Reading Test without reading it: Why comprehension tests
should not include passage- independent items . Scientific Studies
of Reading , 10 ( 4 ), 363 – 380 . doi: 10.1207/s1532799xssr1004_ 2
K e e n a n , J . M . , B e t j e m a n n , R . S . , & O l s o n , R . K . ( 2 0 0 8 ) . R e a d i n g c o m -
prehension tests vary in the sk ills they assess: Differential depen-
dence on decoding and oral comprehension . Scientific Studies of
Reading , 12 ( 3 ), 281 – 30 0 . doi: 10.1080/10888430802132 279
K e e n a n , J . M . , & M e e n a n , C . E . ( 2 0 1 4 ) . T e s t d i f f e r e n c e s i n d i a g n o s i n g
reading comprehension deficits . Journal of Learning Disabilities ,
47 ( 2 ), 125 – 135 . doi: 10.1177/0022219412439326
K e n d e o u , P. , S a v a g e , R . , & v a n d e n B r o e k , P. ( 2 0 0 9 ) . R e v i s i t i n g t h e
simple view of reading . British Journal of Educational Psychology ,
79 ( 2 ), 353 – 370 . doi: 10.1348/978185408X369020
K e n d e o u , P . , v a n d e n B r o e k , P . , W h i t e , M . , & L y n c h , J . S . ( 2 0 0 9 ) .
Predicting reading comprehension in early elementary school:
The independent contributions of oral language and decoding
skills . Journal of Educational Psychology , 101 ( 4 ), 765 – 778 .
doi: 10.1037/a0015956
K e r s h a w , S . , & S c h a t s c h n e i d e r , C . ( 2 0 1 2 ) . A l a t e n t v a r i a b l e a p p r o a c h
to the simple view of reading . Reading and Writing , 25 ( 2 ), 433 –
4 6 4 . d o i : 1 0 . 1 0 0 7 / s 1 1 1 4 5 - 0 1 0 - 9 2 7 8 - 3
K i n t s c h , W . , & R a w s o n , K . A . ( 2 0 0 5 ) . C o m p r e h e n s i o n . I n M . J .
S n o w l i n g & C . H u l m e ( E d s . ) , The science of reading: A handbook
(pp. 211 – 226 ). Malden, MA : Blackwell . doi: 10.1002/9780470
757642.ch12
K i r b y , J . , & S a v a g e , R . ( 2 0 0 8 ) . C a n t h e s i m p l e v i e w d e a l w i t h t h e
complexities of reading? Literacy , 42 ( 2 ), 75 – 82 . doi : 10.1111/
j.1741- 4369.2008.00487.x
K u h n , M . R . , S c h w a n e n f l u g e l , P . J . , & M e i s i n g e r , E . B . ( 2 0 1 0 ) . A l i g n i n g
theory and assessment of reading fluency: Automaticity, prosody,
and definitions of fluency . Reading Research Quarterly , 45 ( 2 ),
230 – 251 . doi: 10.1598/RRQ.45.2.4
L e p o l a , J . , L y n c h , J . S . , L a a k k o n e n , E . , S i l v é n , M . , & N i e m i , P . ( 2 0 1 2 ) . T h e
role of inference making and other language skills in the develop-
ment of narrat ive listening comprehension i n 4–6- year- old children .
Reading Research Quarterly , 47 ( 3 ), 259 – 282 . doi: 10.1002/RRQ.020
L e s l i e , L . , & C a l d w e l l , J . S . ( 2 0 1 0 ) . Qualitative Reading Inventory (5th
ed .). Boston, MA : Pearson .
L o c a s c i o , G . , M a h o n e , E . M . , E a s o n , S . , & C u t t i n g , L . E . ( 2 0 1 0 ) .
Executive dysfunction among children with reading comprehen-
sion deficits . Journal of Learning Disabilities , 43 ( 5 ), 441 – 45 4 .
doi: 10.1177/0022219409355476
M a c G i n i t i e , W . H . , M a c G i n i t i e , R . K . , M a r i a , K . , & D r e y e r , L . G .
( 2000 ). Gates–MacGinitie Reading Tests ( 4th ed. ). Itasca, IL :
Riverside .
M e g h e r b i , H . , & E h r l i c h , M . F . ( 2 0 0 5 ) . L a n g u a g e i m p a i r m e n t i n l e s s
skilled comprehenders: The on- line processing of anaphoric pro-
nouns in a listening situation . Reading and Writing , 18 ( 7–9 ), 715 –
753 . doi: 10.1007/s11145- 005- 8131- 6
M e g h e r b i , H . , S e i g n e u r i c , A . , & E h r l i c h , M . - F . ( 2 0 0 6 ) . R e a d i n g c o m -
prehension in French 1st and 2nd grade children: Contribution of
decoding and language comprehension . European Journal of
Psychology of Education , 21 ( 2 ), 135 – 147 . doi: 10.1007/BF03173573
M i t c h e l l , A . M . , & B r a d y , S . A . ( 2 0 1 3 ) . T h e e f f e c t o f v o c a b u l a r y
knowledge on novel word identification . Annals of Dyslexia ,
63 ( 3/4 ), 201 – 216 . doi: 10.1007/s11881- 013- 0080- 1
M u t e r , V . , H u l m e , C . , S n o w l i n g , M . , & S t e v e n s o n , J . ( 2 0 0 4 ) .
Phonemes, rimes, vocabulary, and grammatical skills as founda-
tions of early reading development: Evidence from a longitudinal
study . Developmental Psychology , 40 ( 5 ), 665 – 6 81 . doi: 10.10 37/
0012- 1649.40.5.665
N a t i o n , K . , & S n o w l i n g , M . ( 1 9 9 7 ) . A s s e s s i n g r e a d i n g d i f f i c u l t i e s :
The validity and utility of current measures of reading skill .
British Journal of Educational Psychology , 67 ( 3 ), 359 – 370 .
doi : 10 .1111/j. 204 4- 8279.19 97.t b01250.x
Learning to Read: Should We Keep Things Simple? | 169
N a t i o n , K . , & S n o w l i n g , M . J . ( 1 9 9 8 ) . I n d i v i d u a l d i f f e r e n c e s i n
contextual facilitation: Evidence from dyslexia and poor reading
comprehension . Child Development , 69 ( 4 ), 996 – 1011 . doi: 10.1111/
j.1467- 8624.1998.tb06157.x
N a t i o n , K . , & S n o w l i n g , M . J . ( 2 0 0 4 ) . B e y o n d p h o n o l o g i c a l s k i l l s :
Broader language skills contribute to the development of reading .
Journal of Research in Reading , 27 ( 4 ), 342 – 35 6 . doi: 10.1111/
j.1467- 9817.2004.00238.x
Nationa l Governors Associat ion Center for Best Pract ices & Council
of Chief State School Officers . ( 2010 ). Common Core State
Standards for English language arts and literacy in history/social
studies, science, and technical subjects . Washington, DC : Authors .
O a k h i l l , J . , & C a i n , K . ( 2 0 1 2 ) . T h e p r e c u r s o r s o f r e a d i n g c o m p r e -
hension and word reading in young readers: Evidence from a
four- year longitudinal study . Scientific Studies of Reading , 16 ( 2 ),
91 – 121 . doi: 10.108 0/10 88 84 38.2 010.529219
O a k h i l l , J . , C a i n , K . , & B r y a n t , P . E . ( 2 0 0 3 ) . T h e d i s s o c i a t i o n o f w o r d
reading and text comprehension: Evidence from component
skills . Language and Cognitive Processes , 18 ( 4 ), 443 – 4 68 .
doi: 10.1080/01690960344000008
O l s o n , R . K . , K e e n a n , J . M . , B y r n e , B . , & S a m u e l s s o n , S . ( 2 0 1 4 ) . W h y
do children differ in their development of reading and related
skills? Scientific Studies of Reading , 18 ( 1 ) , 3 8 – 5 4 . d o i : 1 0 . 1 0 8 0 / 1 0 8 8
8438.2013.800521
Ouellette , G.P. ( 2006 ). What ’ s meaning got to do with it: The role
of vocabulary in word reading and reading comprehension .
Journal of Educational Psychology , 98 ( 3 ), 554 – 566 . doi: 10.1037/
0022- 0663.98.3.554
O u e l l e t t e , G . P . , & B e e r s , A . ( 2 0 1 0 ) . A n o t - s o - s i m p l e v i e w o f r e a d i n g :
How oral vocabular y and visual- word recognition complicate the
story . Reading and Writing , 23 ( 2 ), 189 – 208 . doi: 10.100 7/s11145-
008- 9159- 1
P a z z a g l i a , F . , C o r n o l d i , C . , & T r e s s o l d i , P . E . ( 1 9 9 3 ) . L e a r n i n g t o
read: Evidence on the distinction between decoding and compre-
hension skills . European Journal of Psychology of Education , 8 ( 3 ),
2 4 7 – 2 5 8 . d o i : 1 0 . 1 0 0 7 / B F 0 3 1 7 4 0 8 0
P e r f e t t i , C . A . ( 2 0 0 7 ) . R e a d i n g a b i l i t y : L e x i c a l q u a l i t y t o c o m p r e h e n -
sion . Scientific Studies of Reading , 11 ( 4 ), 357 – 383 . doi: 10.1080/
10888430701530730
P e r f e t t i , C . A . , & S t a f u r a , J . ( 2 0 1 4 ) . R e a d i n g c o m p r e h e n s i o n : I n c l u d i n g
word knowledge in a theoretical framework . Scientific Studies of
Reading , 18 ( 1 ) , 2 2 – 3 7 . d o i : 1 0 . 1 0 8 0 / 1 0 8 8 8 4 3 8 . 2 0 1 3 . 8 2 7 6 8 7
P r o t o p a p a s , A . , S i m o s , P . G . , S i d e r i d i s , G . D . , & M o u z a k i , A . ( 2 0 1 2 ) .
The components of the simple view of reading: A confirmatory
factor analysis . Reading Psychology , 33 ( 3 ), 217 – 240 . d oi : 10.1080/
02702711.2010.507626
R a n s b y , M . J . , & S w a n s o n , H . L . ( 2 0 0 3 ) . R e a d i n g c o m p r e h e n s i o n
skills of young adults with childhood diagnoses of dyslexia .
Journal of Learning Disabilities , 36 ( 6 ), 538 – 555 . doi: 10.1177/00222
194030360060501
R o s e