Justice, 2017

AJSLP
Research Article
Longitudinal Impacts of Print-Focused

Read-Alouds for Children With
Language Impairment
Laura M. Justice,a Jessica Logan,a and Joan N. Kaderavekb
Purpose: Preschoolers with language impairment (LI) are Results: Results of hierarchical linear models examining
prime candidates for early-literacy interventions, given their children’s print knowledge at 1-year postintervention
susceptibility for future reading difficulties. To date, most showed that the effect size (d = 0.20) favoring the
studies of early-literacy interventions for this population treatment group was similar to that observed one year
has assessed short-term impacts, with limited attention to prior (d = 0.21) at the end of intervention, suggesting that
whether initial effects are sustained over time. This study results did not fade over time. Results also showed that
was designed to evaluate longitudinal impacts of print- children with LI and comorbid low nonverbal cognition
focused read-alouds implemented by early childhood special benefited the most from the intervention delivered 1 year
education teachers for a clinic sample of children with LI. earlier.
Method: Assessment data available for 172 children with Conclusion: The maintenance of short-term effects
LI were analyzed to examine their print knowledge 1-year to 1-year postintervention supports the value of early
postintervention. Measures examined children’s alphabet childhood teachers using print-focused read-alouds to
knowledge, print concepts, and name-writing skills, which improve the early-literacy skills of children with LI in their
were used to derive a print-knowledge composite. classrooms.
I
n recent years, researchers have shown considerable commercial storybooks to their students each week and
interest in identifying effective avenues for improving integrated systematic discussions about print form and fea-
the literacy skills of children with disabilities across tures into these read-alouds (Justice, Logan, Kaderavek, &
all grades of schooling (McClintock, Pesco, & Martin‐Chang, Dynia, 2015). Results showed that children with LI whose
2014; Wright, Mitchell, O’Donoghue, Cowhey, & Kearney, teachers implemented the program had significantly better
2015). Given the well-documented susceptibility of children print knowledge in the spring of the year (controlling for
with language impairment (LI) for reading difficulties, a fall scores) compared with children with LI whose teachers
significant volume of this work has focused specifically implemented a comparison read-aloud program. However,
on this subset of children with disabilities (Justice, Skibbe, as with many other experimental studies focused on improv-
McGinty, Piasta, & Petrill, 2011; Lovelace & Stewart, 2007; ing the literacy skills of children with LI (e.g., Lovelace &
McClintock et al., 2014; Wright et al., 2015). For instance, Stewart, 2007; McClintock et al., 2014; Wright et al., 2015),
one recent study reported results of a randomized controlled the study did not attend to whether the short-term positive
trial (RCT) involving 291 three- to five-year-old children effects were sustained past the period of intervention. To this
with LI enrolled in early childhood special education (ECSE) end, it is unclear whether the observed short-term improve-
classrooms in the United States. In this study, teachers ments in literacy skills lead to longer term sustained benefits
implemented a 30-week classroom-based intervention featur- for children with LI, or fade out with time.
ing print-focused read-alouds; specifically, teachers read The present study represents a follow-up to the initial
report of the RCT referenced previously, and was designed
to determine if the intervention impacts observed during the
a
The Ohio State University, Columbus intervention year were sustained at 1-year postintervention.
b
University of Toledo, Perrysburg, Ohio An additional focus was to determine whether treatment
Correspondence to Laura M. Justice: justice.57@osu.edu effects at 1-year follow-up were associated with children’s
Editor: Krista Wilkinson nonverbal cognition. The latter focus was of interest given
Associate Editor: Julie Wolter that the children with LI showed significant heterogeneity
Received December 28, 2015
Revision received June 23, 2016
Accepted October 19, 2016 Disclosure: The authors have declared that no competing interests existed at the time
https://doi.org/10.1044/2016_AJSLP-15-0200 of publication.
American Journal of Speech-Language Pathology • 1–14 • Copyright © 2017 American Speech-Language-Hearing Association 1
Downloaded From: http://ajslp.pubs.asha.org/pdfaccess.ashx?url=/data/journals/ajslp/0/ by a Univ Of Newcastle User on 04/07/2017
Terms of Use: http://pubs.asha.org/ss/rights_and_permissions.aspx
in their cognitive skills, with approximately half of the sam- consistently shows that even at very young ages, children
ple exhibiting nonverbal cognition scores more than 1 SD with LI show significant lags in their development of early-
below the mean (i.e., 85 or lower standard score points). literacy skills as compared with typical peers (Boudreau
The nature of the sample allowed us to consider whether & Hedberg, 1999; Cabell, Justice, Zucker, & McGinty,
children with LI with very low nonverbal cognition may 2009; Cabell et al., 2010). Early-literacy skills are considered
have responded differently to the intervention than children foundational antecedents for or precursors to reading
with average or better cognitive performance. Further, achievement, and acquisition of these antecedents is consid-
results of the initial RCT suggested that the intervention ered necessary for formal reading instruction to be successful.
was most influential for children with LI who had low For instance, children with SLI at 4 years of age know,
nonverbal cognition (Justice, Logan, Kaderavek, et al., on average, about seven letters of the alphabet, com-
2015)—namely, children with lower levels of cognition par- pared with age-matched peers who know an average
ticularly benefitted from being exposed to print-focused 20 letters (Cabell et al., 2009). Alphabet knowledge is
read-alouds. considered one of the strongest and most reliable predic-
tors of future reading achievement (Lonigan & Shanahan,
2009); consequently, this lag in letter-naming skills seen
Early Literacy Skills of Children With LI in children with LI during the preschool years seems to
The presence of LI is one of the most common reasons represent a significant risk factor in their overall literacy
for which young children qualify for participation in ECSE development.
programs. ECSE programs are early-education settings To attenuate such early risks, and to potentially
that serve children between 3 and 5 years of age who have improve future reading achievement, researchers have
a diagnosed disability. Oftentimes, these classrooms serve developed and tested targeted approaches to improving
both children with and without disabilities in an effort for the early-literacy skills of children with LI, particularly
children with disabilities to be educated alongside children their print knowledge and phonological awareness (Ezell,
who are typically developing. By far, the majority of children Justice, & Parsons, 2000; Gillon, 2002; Justice, Chow,
with disabilities served in ECSE settings exhibit LI, either Capellini, Flanigan, & Colton, 2003; Lovelace & Stewart,
as a primary condition or comorbid with/secondary to 2007; McNamara, Vervaeke, & Van Lankveld, 2008; van
other conditions, such as autism or intellectual disability Bysterveldt, Gillon, & Moran, 2006). In general, these
(e.g., Down syndrome). The former is typically referred to studies show that targeted interventions for developing
as specific or primary LI, whereas the latter is often referred children’s early-literacy skills can lead to significant improve-
to as nonspecific LI (Tomblin et al., 1997). Children with ments in these skills for children with LI. Targeted inter-
nonspecific LI typically have lower nonverbal cognition ventions are those that explicitly focus on improving
than those with specific LI (SLI), and low nonverbal cogni- children’s early-literacy skills, such as their knowledge of the
tion is usually an exclusionary factor when a diagnosis of letter names or their awareness of rhyming patterns; often,
SLI is made (see Tomblin et al., 1997). In fact, many stud- such interventions follow a scope and sequence delineating
ies of children with LI will exclude children with signifi- specific early-literacy targets to address over an intervention
cant deficits in nonverbal cognition (e.g., L.-Y. Guo & period (e.g., Gillon, 2002; Justice et al., 2003). These can
Schneider, 2016; Justice et al., 2011). be differentiated from the interventions typically delivered to
Of relevance to this study are reports of the heightened children with LI, in which oral-language goals are focused
susceptibility of children with LI for reading disability, to on most intensively so as to improve children’s vocabulary
include both children with specific and nonspecific variations and morphosyntactic skills (Fey, Cleave, Long, & Hughes,
of LI (Catts, Fey, Tomblin, & Zhang, 2002; Morgan, Farkas, 1993; Girolametto, Pearce, & Weitzman, 1996). A com-
& Wu, 2011; Skibbe et al., 2008). Morgan et al. (2011) con- parison of the effects of a 12-week early-literacy intervention
ducted growth-curve modeling to examine the trajectory that explicitly targeted print knowledge and phonological
of reading development for children with LI compared with awareness to 12 weeks of traditional language intervention,
typically developing peers. At kindergarten, children with focused only on oral-language goals, for 3- to 5-year-olds
LI performed about 0.4 SD lower than those without dis- with LI showed that only the former had significant impacts
abilities in reading achievement. Over time, however, the on children’s early-literacy development (McNamara et al.,
gap between the two groups widened: By fifth grade, children 2008). We can interpret such work to show that traditional
with LI performed 0.8 SD lower than those without reading language intervention focused exclusively on facilitating
disabilities. Such work has led to great interest in determining children’s oral language skills (e.g., vocabulary, grammar)
how reading disabilities among children with LI can be does not contribute to growth in early-literacy skills for
prevented, potentially by enhancing children’s early-literacy children with LI; thus, interventions explicitly designed to
skills prior to entrance to formal schooling and the rigors of lead to growth in early-literacy skills should accompany
reading instruction in kindergarten. the provision of traditional language intervention.
To this end, a growing body of work has sought to An important issue within the early-literacy inter-
improve our understanding of the early-literacy skills of vention literature is whether short-term treatment effects
children with LI, prior to the advent of formal schooling are sustained over time. This is an important avenue of
and exposure to formal reading instruction. Such work inquiry, given that the focus of early-literacy intervention
2 American Journal of Speech-Language Pathology • 1–14

is not necessarily to facilitate targeted skills in the short-term, print knowledge and their future reading achievement.
but rather to promote longer term advancements in chil- From a practical standpoint, print-focused read-alouds
dren’s literacy abilities. Some studies have provided evidence appear to be a promising approach for addressing the print-
of sustainability for children who are typically developing knowledge lags seen in children with LI, with potential
(Lundberg, Frost, & Petersen, 1988; Nancollis, Lawrie, & benefits to their reading skills longer term.
Dodd, 2005; Piasta, Justice, McGinty, & Kaderavek, 2012).
For instance, in a seminal study, Lundberg et al. (1988) pro-
vided daily phonological awareness instruction to Danish Print-Focused Read-Alouds
preschoolers for an 8-month period, and examined their Young children’s development of print knowledge
reading and spelling skills in second grade. Second graders can be directly facilitated by repeated read-alouds in which
who had received phonological awareness instruction 4 years adults embed explicit discussions regarding the forms and
prior were significantly better readers and spellers than functions of print into the read-aloud. This approach to
those who had not. reading books with children varies from the conventional,
For children with LI, however, we have considerably as when adults typically read books with children, they
less evidence concerning the sustainability of the effects of tend to focus largely on the illustrations and the storyline,
early-literacy interventions. An exception is Gillon’s (2002) with little explicit attention to mediating the child’s under-
work with New Zealand children, which has provided some standing about the forms and functions of print in the
evidence of the sustainability of the effects of early-literacy book (Zucker, Cabell, Justice, Pentimonti, & Kaderavek,
intervention delivered to children with speech and LIs at 2012). Print-focused read-alouds were designed to provide
5 to 7 years of age. However, the children in this study adults with explicit guidance in how to read storybooks
primarily had speech-sound disorders; therefore, the results with children in a way that increases their knowledge of
do not necessarily generalize to children with LI. Further, the forms of print within the book (e.g., the names of letter,
Gillon’s research focused specifically on delivery of phono- and difference between letters and words) and the functions
logical awareness intervention. Although phonological aware- of print (e.g., the role of the title). More recent studies of
ness is causally related to reading achievement (Wagner print-focused read-alouds (e.g., Piasta et al., 2012) follow
& Torgesen, 1987) and has been targeted in many pub- a treatment manual that provides a scope and sequence of
lished studies of the effects of early-literacy intervention intervention that is delivered through repeated reading of
(Gillon, 2002; Justice, Kaderavek, Bowles, & Grimm, 30 commercially available titles over a 30-week period
2005; Lundberg et al., 1988; van Bysterveldt et al., 2006; (Justice & Sofka, 2013). The scope of instruction is organized
Ziolkowski & Goldstein, 2008), there are other early-literacy to develop children’s knowledge in four domains of print
skills that are areas of concern for children with LI, namely knowledge: book and print organization (e.g., the role of
their development of print knowledge. the title page), print meaning (e.g., the role of print in con-
Print knowledge is a dimension of early-literacy devel- veying information), letters (e.g., the names of the alphabet
opment that represents children’s developing knowledge letters), and words (e.g., the difference between letters and
of the forms and functions of written language (Whitehurst words). Fifteen objectives across these four domains are
& Lonigan, 1998). It includes children’s emerging knowl- cycled over a 30-week period to provide repeated, explicit
edge of the alphabet as well as rules governing print, such instruction towards children’s learning of these objectives
as directionality and organizational units (e.g., letters and during adult–child read-alouds. The Appendix shows these
words; Justice, Bowles, & Skibbe, 2006). As we noted pre- 15 objectives with sample references to print adults may
viously, preschool-aged children with LI show delays in their make to introduce explicit instruction regarding each.
development of print knowledge (Boudreau & Hedberg, Pilot and experimental studies have consistently shown
1999; Cabell et al., 2009), and these seem to contribute to that print-focused read-alouds can improve the print-related
the significant underachievement in reading seen among skills of young children with LI, on the basis of such mea-
children with LI over the primary grades (Justice et al., sures as name-writing skill, alphabet knowledge, and print-
2006). Studies of typically developing children have pro- concept knowledge (Ezell et al., 2000; Justice & Ezell,
vided causal evidence that improvements in children’s print 2002; Justice, Kaderavek, Fan, Sofka, & Hunt, 2009; Justice
knowledge during the preschool years can yield longer term et al., 2011; Lovelace & Stewart, 2007; van Bysterveldt
improvements in reading skills (McGinty, Justice, Piasta, et al., 2006). Effects are seen whether the intervention is
Kaderavek, & Fan, 2012). For example, Piasta et al. (2012) implemented by parents (Ezell et al., 2000; Justice & Ezell,
showed that children whose preschool teachers imple- 2002; van Bysterveldt et al., 2006), speech-language patholo-
mented an early-literacy intervention over an academic gists (SLPs; Lovelace & Stewart, 2007) or teachers (Justice
year (30 weeks of implementation) featuring print-focused et al., 2009; Justice et al., 2011). A Spanish adaptation
read-alouds performed significantly better 2-years post- of the intervention was recently tested with monologual,
intervention, when children were in first grade, on standard- Spanish speakers with LI, as implemented by parents, with
ized measures of word reading, reading comprehension, significant and positive effects on children’s print knowledge
and spelling, compared with children whose teachers were (Pratt, Justice, Perez, & Duran, 2015). However, a short-
assigned to a control condition. Such research theoretically coming within this literature is that no study has involved
provides causally interpretable linkages between children’s long-term follow-up of children following intervention
Justice et al.: Longitudinal Impacts 3

exposure; thus it is unknown whether the short-term effects 64 months at initial study enrollment (M = 55, SD = 4.7),
of print-focused read-alouds are sustained beyond the initial and 59 to 84 months at the 1-year follow-up investigation
period of intervention. (M = 72, SD = 4.8). All of the children spoke English as
To address this question, the present study represents their first language; 83% of children were White and 11%
a follow-up investigation of children with LI who attended were African American (6% were of other race/ethnicities).
ECSE classrooms in which their teachers implemented print- General descriptive information regarding these children
focused read-alouds over a 30-week period, corresponding appears in Table 1. The children were a subset of the larger
generally to an academic year of instruction. The recently sample of the 291 children with LI enrolled in the RCT,
published RCT (Justice, Logan, Kaderavek, et al., 2015) representing those who had the data necessary for the present
reported an effect-size estimate of 0.21 for children with analyses, which included a measure of nonverbal cogni-
LI whose teachers implemented print-focused read-alouds tion administered at entrance into the study (n = 214). The
as compared with children whose teachers read the same set 78 children who were not assessed on this assessment were
of storybooks over an academic year but read using their nor- between 45 and 48 months of age and therefore not age-
mal read-aloud style. It is interesting to note that the effect- eligible for this assessment at study entry. These 214 children
size estimate for children with LI was identical to that seen are the same sample as were reported in the initial inter-
for typically developing children whose preschool teachers vention findings (Justice, Logan, Kaderavek, et al., 2015).
implemented print-focused read-alouds for an academic year The second inclusionary criterion was that students have
(d = 0.21; Justice et al., 2009). Thus, it seems that children the 1-year follow-up assessment (remaining n = 172), repre-
with LI respond similarly to the intervention as children who senting an 80% retention rate for this subset.
are typically developing. Although relatively small in size,
this effect size is consistent with that seen in educational
Table 1. Children’s descriptive characteristics across conditions.
research generally (d = 0.23; Hill, Bloom, Black, & Lipsey,
2008) and may be interpreted to correspond to a clinically Condition
meaningful effect in terms of magnitude (Hill et al., 2008).
Over the last decade, there has been an increasing emphasis Comparison Experimental
Variable (n = 66) (n = 106)
on the use of effect-size estimates to interpret research findings,
rather than relying on the results of significance tests used Age in months
in traditional inferential statistical modeling (Fritz, Morris, M 54.77 55.06
& Richler, 2012). Some scholars have argued that effect-size SD 4.68 4.77
reporting is particularly important for understanding the Female (%) 26 25.
Speaking English as 100 100.
practical significance of a study, and for translating research first language (%)
findings to policy makers and consumers (McCartney & Race–ethnicity (%)
Rosenthal, 2000). For instance, if the average intervention White, non-Hispanic 82 84.
delivered within the primary grades affects students’ achieve- African American 10 12.
Other 8 7.
ment by about 0.2 SD units, this can be informative for Missing 7 14.
interpreting the effects of a newly developed intervention. Maternal education (highest
The present study reports on two aims. The first aim degree attained, %)
was to determine whether exposure to print-focused read- No high school diploma 10 9.
High school diploma 21 18.
alouds resulted in sustained improvements in children’s print With some college 34 31.
knowledge at one-year post-intervention, whereas the second Associate’s degree 5 6.
was to examine whether post-intervention effects were mod- Bachelor’s degree 13 19.
erated by children’s nonverbal cognition. To be specific, the Graduate degree 18 16.
Family income (%)
second aim sought to test the hypothesis that children with
$0–$15,000 26 17.
lower levels of nonverbal cognition (≤70 on a standardized $15,001–$30,000 5 15.
measure) would experience the greatest sustained benefit $30,001–$45,000 21 5.
from the intervention, as suggested in the results reported $45,001–$60,000 9 16.
from the RCT (Justice, Logan, Kaderavek, et al., 2015). >$60,000 40 46.
CELF Core Language
Although we report the results of inferential statistical tests, M 75.80 78.88
we emphasize the interpretation of effect-size estimates for SD 16.81 16.27
understanding the practical significance of this intervention Nonverbal cognition 81.59 84.87
for children’s early-literacy development. M
SD 18.30 17.85
Note. CELF Core Language = Core Language Composite of the

Clinical Evaluation of Language Fundamentals–Preschool (Wiig,
Method Secord, & Semel, 2004). Nonverbal cognition = Matrices subtest of
Participants the Kaufmann Brief Intelligence Test (Kaufman & Kaufman, 1985).
Note that both the CELF and KBIT are based on a distribution such
The present study involved 172 children with LI that M = 100, SD = 15.
(75% boys, 25% girls) who ranged in age from 45 to

In the RCT, children’s participation spanned a nearly of children with LI (Tomblin et al., 1997). Research-based
2-year period, to include three assessment time points (T1, sampling typically employs standardized assessments and
T2, and T3) and exposure to their teachers’ assigned study application of specific cut-points to identify children with
condition between T1 and T2. T1 and T2 corresponded to LI and does not consider whether or not a child has been
the fall and spring of an academic year in which children clinically identified to have LI and is being provided services.
were enrolled in one of 83 ECSE classrooms across a large For instance, only one-fourth of an epidemiological sample
Midwestern state; T3 occurred at 12 months post-T2, in of children with SLI identified by Tomblin et al. (1997) was
the spring of the next academic year when children were in receiving speech-language services. A drawback to the use of
preschool for a second year or in kindergarten. Between research-based sampling procedures is that study results may
T1 and T2 (fall and spring of the academic year), the children not have strong application to clinical practice.
participated in ECSE classrooms designed to serve both Given that the sample was clinically ascertained, it
children with disabilities and typical peers. These classrooms is not surprising that standardized test data for the present
are led by credential teachers who are certified to teach chil- sample, collected at T1, T2, and T3, show that not all of
dren with disabilities. The curriculum in such classrooms is the sample performed below commonly applied cut-points
designed to support the development of all children served, for identification of LI (i.e., −1 or −1.5 SD of the mean).
and not only those with disabilities. However, for the latter To be specific, children were administered the three Clinical
group of children, specialists are available on-site as part Evaluation of Language Fundamentals–Preschool 2 (Wiig,
of typical programming to support any identified needs of Semel, & Secord, 2004) subtests that together comprise a
the children with disabilities, including physical therapy, language composite: Sentence Structure, Word Structure,
occupational therapy, mental health supports, and speech- and Expressive Vocabulary. The subtest scores are based
language therapy. on a normalized mean of 10 and SD of 3, whereas the com-
posite is based on a normalized mean of 100 and SD of 15.
Initial Ascertainment Procedures At T1, the mean composite score was 76.4 (SD = 17.3), with
At T1, the children were purposefully selected for 70% of the sample receiving a composite score less than
enrollment in the study from among all of the children or equal to −1 SD of the mean and 30% of the sample
enrolled in the 83 participating ECSE classrooms. The class- within the average range (≥ −1 SD). At T3, approximately
rooms served an average of 13 children (M = 4; range = 5 18 months after T1 and the point of interest for the present
to 21), about half of whom were typically developing and the study, children received an average score of 7.3 (SD = 3.5),
other half of whom had identified disabilities. The primary 6.9 (SD = 3.5), and 7.8 (SD = 3.3) on the Sentence Struc-
aim of the RCT was to examine effects of the experimental ture, Word Structure, and Expressive Vocabulary subtests,
literacy intervention (print-focused read-alouds) on children respectively. For the composite, children received an average
with LI, who represented a subset of those in the partici- score of 84.6 (SD = 17.4), and one half of the children (49%)
pating ECSE classrooms. had language composites that were ≤ 1 SD of the mean.
To identify these children, a multistep process was It is worth noting that children’s language skills, on the basis
used (Justice, Logan, Kaderavek, et al., 2015). First, teachers of the composite, were significantly better than they were
completed a screening questionnaire for each child in their 1 year prior, on the basis of a comparison to T2 (M = 80.5,
classroom. The screening items were designed to differentiate SD = 19.8) and T3 scores, t(167) = 3.25, p < .001. It may
children with LI from children with other types of disabilities be that a number of children no longer had impaired lan-
(e.g., autism, hearing loss) and those who were typically guage skills subsequent to T1 and T2 due to provision
developing, given the heterogeneity of the students within of therapy between these time points. Other studies have
each ECSE classroom. For instance, teachers identified showed relatively high levels of transient LI among children
whether a child was being treated for language problems at this age (e.g., Dale, Price, Bishop, & Plomin, 2003).
by an SLP. Second, for children who appeared to have LI At T1 the children were also administered a test of
on the basis of the teacher screener, their Individualized nonverbal cognition, the Matrices subtest of the Kaufman
Education Programs were examined to ensure that they had Brief Intelligence Test (KBIT; Kaufman & Kaufman, 1990).
language-specific objectives on their Individualized Education The children had nonverbal-cognition scores at T1 that
Programs and that these were being addressed by an SLP ranged from 53 to 124 (M = 81.8; SD = 18.4); 52% of the
as the designated service provider. Using these procedures, sample had KBIT scores that were ≤ 1 SD of the mean
291 children with LI were identified from the 83 classrooms, (below a standard score of 85), which is often applied as a
99% of whom were receiving school-based services from an cut-point to differentiate specific from nonspecific LI (Catts
SLP and 1% (n = 4) who were in a prereferral state. Twenty- et al., 2002).
four percent of children had specific, known diagnoses,
such as hearing loss, autism spectrum disorder, and Down
syndrome. General Procedures
These procedures for identifying children with LI All study procedures were approved by the institu-
allowed for identification of a clinically based sample tional review board at the primary university research
of children with LI, currently served by SLPs, yet differ site. Children were enrolled in the larger study within one
from procedures used to identify research-based samples of three consecutive cohorts corresponding to academic

years 2008–2009, 2009–2010, and 2010–2011. During the practice/ccec-curricula/star2/materials/) as is a list of the
first year of study participation (between T1 and T2), chil- storybook titles used in the intervention. With respect to
dren’s ECSE teachers were randomly assigned to imple- caregivers, those in Condition 2 read using their typical
ment one of three planned variations of intervention. The reading style, whereas in Condition 3 they implemented
teachers were the lead educators in mostly half-day pro- print-focused read-alouds. To facilitate caregivers’ imple-
grams, most offering inclusive programming to serve chil- mentation of print-focused read-alouds, caregivers re-
dren with disabilities as well as those without, per state ceived one-on-one training in their home, period phone
regulations which advocate for a 1:1 ratio (students with check-ins and home visits, as well as a parent-friendly
disability:typical peers). insert embedded in each book to provide a guide for in-
corporating print-focused discussions within read-alouds
with their children.
Random Assignment and Treatment Conditions Fidelity to the print-focused read-aloud intervention
Random assignment occurred at the level of the class- was carefully monitored in the trial for both teachers and
room and teacher; specifically, each classroom (n = 83) caregivers. Teacher fidelity was monitored using weekly
was assigned to one of three study conditions on the basis logs completed by teachers as well as well as twice-monthly
of the type of read-aloud program to be implemented by videotapes of intervention sessions collected by teachers
the ECSE teacher and the caregivers of children in that and submitted to research staff using stamped, addressed
classroom—that is, both teachers and the caregivers of mailers; these were submitted by both the control (Condi-
children within their classrooms implemented read-alouds tion 1) and treatment teachers (Conditions 2 and 3). Teachers
over a 30-week period. The three study conditions stipulated submitted 15 videos in total over the 30-week intervention,
whether teachers and parents would read to children using and these were coded using a fidelity coding checklist (FCC)
print-focused read-alouds or their typical reading style for developed for this study. The FCC involves event-based
the academic year. Regardless of condition, teachers and coding of teachers’ references to one of four print-related
caregivers were provided an identical set of 30 storybooks categories: book and print organization (e.g., “Where is the
and a schedule to follow for 30 weeks that stipulated which title of this book?”), print meaning (e.g., “What do you
book was read per week. Teachers were asked to read the think this word says?”), letters (e.g., “This is the letter
weekly book four times per week in their classroom, whereas ‘T’.”), and words (e.g., “This word says ‘fragile’.”). A com-
caregivers were asked to read it two times to their children parison of teachers’ references to print across these four
in the home environment. Those who were assigned to use categories, averaged across the 15 submitted videos, showed
print-focused read-alouds received additional materials that teachers in Condition 2 and 3 made significantly more
with each book that specified objectives to target during references to print (all ps < .05) across three of the four cate-
the reading of each book to address the specified objectives gories as compared with those in Condition 1. To be spe-
in the Appendix. cific, treatment teachers made more references to book
Condition 1 was the true comparison condition, in and print organization (M = 6.51, SD = 2.99) than control
which both teachers and children’s caregivers read to their teachers (M = 2.29, SD = 2.14); to letters (M = 11.92,
children using their typical reading style for the academic SD = 4.85) than control teachers (M = 3.71, SD = 4.89);
year. Teachers read each of 30 storybooks four times during and to words (M = 10.36, SD = 4.09) than control teachers
the week it was assigned, whereas caregivers were asked to (M = 2.05, SD = 2.77). For print meaning, treatment
reach the title twice during the week it was assigned. Teachers teachers (M = 1.91, SD = 0.85) made about the same num-
read the book as a whole-group classroom book reading, ber of references as control teachers (M = 1.67, SD = 1.14).
whereas caregivers read the assigned books in their home, Recent work by Guo et al. (2016) showed that teachers’
typically one on one with their children. Teachers and care- references to print served to mediate relations between the
givers were instructed simply to “read the books as they intervention and children’s early-literacy gains.
normally read with their children.” For caregivers, intervention fidelity was assessed
Conditions 2 and 3 both involved teacher implemen- largely by submission of weekly logs on a preprinted post-
tation of the print-focused read-aloud intervention, with card to document implementation of the two required read-
the differentiation between conditions reflecting differ- aloud sessions each week. Within each of the 30 books
ences in caregivers were assigned to do. With respect to provided to caregivers, for both those in the control and
teachers, in both Conditions 2 and 3 teachers read each of the treatment conditions, a stamped, addressed postcard
the 30 storybooks four times during the week it was assigned, was placed on the last page which provided a brief log to
as with teachers in Condition 1, yet within each read-aloud complete and then submit to the project staff. On average,
session they incorporated explicit discussions about print caregivers returned about 45% of the requested postcards:
to address 15 objectives (see Appendix). To facilitate these 48% of those in Condition 1, 46% for those in Condition 2,
discussions and adherence to the targeted objectives, an and 42% for those in Condition 3. Whereas those in Con-
insert was placed within each book that identified the ditions 1 and 2 read with their normal read-aloud style,
objectives aligned to the book and provided guidance in those in Condition 3 read with a print-focused read-aloud
how to address these in discussions. These inserts are avail- style; however, postcard submission rates did not signifi-
able for free download on the web (http://ccec.ehe.osu.edu/ cantly differ across the three conditions, F(2, 694) = 1.535,

p = .216. If caregiver logs provide a valid representation of print organization (e.g., identify the front of the book, iden-
fidelity, these data suggest that implementation fidelity was tify the title, directionality of text on a page), concept of
generally low among the caregivers, consistent with prior letter (e.g., identify a capital letter), and print function (e.g.,
studies (Lonigan & Whitehurst, 1998). identify speech bubbles as characters talking). One point is
awarded for most test items answered correctly, with two
points awarded for select items requiring more complete
Treatment Comparisons in Current Study answers (maximum score = 17). The measure exhibits ade-
Condition 1 in the initial RCT represented a control quate interrater reliability and construct validity (Justice
or comparison condition, with both teachers and caregivers et al., 2006).
reading books to their children using their typical reading To assess children’s knowledge of the alphabet,
style. For Conditions 2 and 3, the primary distinction con- the Upper and Lower Case Letter Knowledge subtests
cerned what children’s caregivers were asked to do with of the Phonological Awareness Literacy Screening–PreK
their children during the intervention year. In Condition 2, (PALS-PreK; Invernizzi, Sullivan, Meier, & Swank, 2004)
caregivers read using their typical reading style, whereas in were administered. Children are asked to provide the name
Condition 3 they implemented print-focused read-alouds of each of 26 letters randomly ordered on an 8.5- × 11-in.
and received training to do so. For the purposes of the page; for the present purposes, children were tested first
present study, Conditions 2 and 3 were collapsed to create on the uppercase letters and then on the lowercase letters.
a single condition that represented whether the teacher One point is awarded for each letter correctly named, for a
implemented print-focused read-alouds in children’s ECSE maximum score of 52 points. The PALS-PreK Teacher’s
classroom (n = 55 teachers) or their typical read-aloud Manual reports an interrater reliability of .99 (Pearson
style (n = 28 teachers), irrespective of the caregivers’ condi- correlation; Invernizzi et al., 2004).
tion. Our rationale for collapsing the second and third treat- To assess children’s name-writing skills, the Name
ment conditions in this study was twofold. First, results of Writing subtest of the PALS-PreK (score range 1–7) was
the initial RCT showed no significant difference between administered. Children are provided a blank sheet of paper
Conditions 2 and 3 with respect to impacts on children’s and are asked to draw a self-portrait and to sign it with
print knowledge (d = 0.10). Second, as reported previously, their name. Children’s signatures are evaluated on a 7-point
caregiver implementation of print-focused read-alouds was scale, with higher scores reflecting increasingly sophisticated
generally poor, as we discuss elsewhere (Justice, Logan, & name-writing abilities. For instance, a child’s name that is
Damschroder, 2015) and have found in previous studies not distinctly separate from the picture receives a score of
(Justice et al., 2011). This stands in direct comparison 0 or 1; a child’s name that contains some correct letters and
to teachers’ implementation of print-focused read-alouds; some random symbols receives a score between 3 and 5; and
overall, teachers showed a high level of adherence to the representations containing all letters in children’s names
schedule of read-aloud sessions and delivery of the scope receive a score of 6 or 7. Interrater reliability is reported
and sequence of print-focused instruction in each session; at .99 (Invernizzi et al., 2004).
details regarding fidelity of implementation are detailed Given that scores on these three measures are highly
in the initial report of the RCT (Justice, Logan, Kaderavek, correlated and reflect a single underlying construct (print-
et al., 2015). Therefore, the focus of this work is to deter- concept knowledge; see McGinty & Justice, 2009), a single
mine 1-year postintervention effects for children with LI print-knowledge composite was derived for each time
whose ECSE teachers, in the year prior, implemented print- point. This composite was created by standardizing each
focused read-alouds or their typical read-aloud style. observed measure (M = 0, SD = 1) and subsequently sum-
ming the three scores together; in doing so, each measure
was equally weighted in deriving the composite. A com-
Measures posite was created for each time point, with standardiza-
Three general measures of print knowledge, measuring tion always on the basis of the pretest mean and standard
children’s print-concept knowledge, alphabet knowledge, deviation; in this way, each measure’s observed mean
and name-writing ability, were administered at T1, T2, and and standard deviation at the first time point were used
T3. The measures are designed to collectively capture chil- to calculate the z-score values for that measure’s scores
dren’s developing knowledge about print and thus are well at T2 and T3.
aligned to the intervention. However, the intervention does The calculated mean scores for each observed mea-
not explicitly teach the skills tested in the three measures. sure and the composite score at each time point are pre-
Further, it should also be pointed out that children’s teachers, sented in Table 2. Because of the way these scores were
who implemented the intervention (or comparison condition), calculated, we can determine that children generally grew
were unaware of the content of the assessments given to in their skills over time. Print knowledge, for example,
their students. has a calculated mean score of 0.99 at T2, suggesting
The Preschool Word and Print Awareness (Justice & children grew, on average, 1 T1 SD from fall to spring of
Ezell, 2001) is an index of print-concept knowledge, includes preschool, and the score of 1.89 suggests that children grew,
14 items embedded in the shared reading of a children’s on average almost 2 full T1 SDs between fall of preschool
storybook to examine children’s knowledge of book and and fall of kindergarten.

Table 2. Print-knowledge indices at all three study time points.
print knowledge collected at three time points, as well as
the composite, whereas Table 3 provides correlational find-
Time Measure n M SD
ings relative to children’s nonverbal cognition, language
Fall T1 Print-Concept Knowledge 171 0.15 1.02 skills, and print knowledge on the basis of the T1, T2, and
Letter Knowledge 163 0.07 1.00 T3 composites. As can be seen, there are strong, positive
Name Writing 172 0.27 1.02 relations among all of these variables: Children with higher
Print-Knowledge Composite 162 0.45 2.32 levels of nonverbal cognition and language have higher
Spring T2 Print-Concept Knowledge 170 0.99 1.33
Letter Knowledge 170 0.56 1.06 print knowledge at all three time points. In addition, chil-
Name Writing 170 0.86 0.93 dren who have higher levels of print knowledge at T1 also
Print-Knowledge Composite 170 2.42 2.66 have higher levels of print knowledge at T2 and T3.
Spring T3 Print-Concept Knowledge 172 1.89 1.37 To address the first aim, we examined the early-
Letter Knowledge 172 1.45 0.85
Name Writing 172 1.50 0.62 literacy composite scores of the two conditions at T3. The
Print-Knowledge Composite 172 4.84 2.41 mean score for children in the comparison group was 4.54
(SD = 2.4), as compared with 5.03 (SD = 2.41) for those
Note. Print-Concept Knowledge = Preschool Word and Print in the experimental group; this corresponds to an effect size
Awareness (Justice & Ezell, 2001); Letter Knowledge = Upper and
Lower Case Letter Knowledge subtests of the Phonological of 0.2 SD units (Cohen’s d). This effect-size contrast is nearly
Awareness Literacy Screening–PreK (Invernizzi, Sullivan, Meier, & identical to that seen when comparing groups immediately
Swank, 2004); Name Writing = Name Writing subtest of the PALS- postintervention (d = 0.21), suggesting that the treatment
PreK; Print-Knowledge Composite = standardized composite index effects held over time. However, these results do not account
of print knowledge on the basis of the three assessments.
for the potential nesting of children within classrooms during
the preschool year, when the intervention conditions were
received. We therefore fit a hierarchical linear model predict-
Results ing children’s scores on the T3 print-knowledge composite,
which showed that 21% of the variance in children’s T3
The primary aims of this study were (a) to examine scores was attributable to the preschool teacher the child
the extent to which children with LI showed longitudinal had ( p = .02). Thus, all further analyses were conducted
benefits of exposure to print-focused read-alouds in their controlling for the child’s ECSE teacher.
ECSE classrooms, as implemented by their ECSE teachers, These further analyses were conducted in two models:
and (b) to determine whether any treatment effects observed Model 1 included only main effects, whereas Model 2 included
at 1-year follow-up were moderated children’s nonverbal a moderating effect. The main effects models considered
cognition. The analyses served to compare print knowledge the impacts of children’s treatment condition between T1
at 1-year postintervention as a function of whether children’s and T3 while controlling for children’s nonverbal cognition
ECSE teachers implemented print-focused read-alouds (as assessed at T1 using the KBIT). For this analysis, we
(n = 55) or conducted read-alouds using their typical reading conducted a hierarchical linear model predicting the T3
style (n = 28). To this end, all children experienced regular, print-knowledge composite from children’s T1 composite,
repeated read-alouds with the same set of books within their KBIT scores, and whether the teacher provided the treat-
ECSE classrooms, yet the way in which their teachers read ment (no = 0, yes = 1). In Model 2, we considered the poten-
the books varied on the basis of the assigned treatment tial moderating role of children’s nonverbal cognition on
condition. the treatment effect, by adding to Model 1 an interaction
Tables 2 and 3 provide descriptive information regard- between KBIT and treatment. For both of these analyses,
ing several measures of interest in this study. To be specific, the KBIT was centered at the lowest observed score for the
Table 2 provides descriptives concerning the measures of sample (standard score of 53). Centering a variable allows
for a more straightforward interpretation of the resulting
regression coefficients than occurs in regression models, in
Table 3. Intercorrelations among main child measures.
which all predictors are equal to 0. Centering a variable is
particularly important in interaction models and interpreting
Measure 1 2 3 4 5
effects of these models (such as we do in Model 2 below;
1. T1 KBIT — see Cohen, Cohen, West, & Aiken, 2003). Without center-
2. T3 CELF .55 — ing, regression coefficients would represent estimated values
3. T1 Print-Knowledge Composite .46 .53 — when children scored a 0 on the KBIT, which is not an
4. T2 Print-Knowledge Composite .50 .64 .83 — observable or possible value. By centering the KBIT at
5. T3 Print-Knowledge Composite .49 .66 .64 .73 —
53, all of the results can be interpreted relative to a score
Note. All r coefficients significant at p < .001. T1, T2, and T3 = Time 1 of 53 on the KBIT.
(fall of preschool), Time 2 (spring of preschool), Time 3 (one year The results of the two analyses are presented in
after T2). KBIT = Matrices subtest of the Kaufmann Brief Intelligence Table 4. For Model 1, we found a main effect of the pre-
Test (Kaufman & Kaufman, 1985). CELF = Core Language Composite
of the Clinical Evaluation of Language Fundamentals–Preschool test composite (estimate = 0.53, p < .01) and of the KBIT
(Wiig et al., 2004). (estimate = 0.03, p < .01), but no significant differences
due to condition (estimate = 0.18, p = .57). For Model 2,

Table 4. Results of multilevel regression model predicting T3
very few studies seeking to study the longitudinal benefits
print-knowledge composite.
of interventions delivered in ECSE settings by classroom
Model Estimate SE df t value Pr > |t|
teachers. Indeed, although there is theoretical appeal in
advancing our understanding of the causal relations between
Model 1: Main effects certain intervention approaches and children’s outcomes,
Intercept 1.71 0.75 68 2.28 0.03 there is significant, practical import as well: ECSE educators
T1 composite 0.53 0.07 90 7.81 <0.001 and allied professionals who seek to improve children’s
KBIT 0.03 0.01 90 3.75 <0.001
Treatment 0.18 0.32 68 0.57 0.57 foundational skills in reading benefit from having access to
Model 2: Interactions empirically supported practices with demonstrated effects
Intercept 0.22 1.12 68 0.20 0.84 on clinically identified samples.
T1 composite 0.53 0.07 89 7.86 <0.001 An important characteristic of the practice studied
KBIT 0.05 0.01 89 3.79 <0.001
Treatment 2.56 1.37 68 1.87 0.06 in this investigation, print-focused read-alouds, is that it is
KBIT × Treatment −0.03 0.02 89 −1.79 0.07 an easy-to-use, low-cost, repeated-reading intervention in
which teachers systematically promote children’s knowledge
Note. KBIT was centered at the lowest observed KBIT score (53, about print during whole-class read-alouds. Prior studies
which was observed for 7 children). T1 = Time 1 (fall of preschool).
KBIT = Matrices subtest of the Kaufmann Brief Intelligence Test show that this practice yields positive short-term effects on
(Kaufman & Kaufman, 1985). the print knowledge of children who are disadvantaged yet
typically developing (e.g., Justice et al., 2009) as well as
those who have LI that is specific (e.g., Justice et al., 2011)
and nonspecific in nature (Justice, Logan, Kaderavek, et al.,
which included the interaction between KBIT and treatment, 2015; van Bysterveldt et al., 2006), with the present work
we found the same main effect for the pretest composite suggesting that results are sustained over time to 1-year
(estimate = 0.53, p < .001), as we would expect. Children’s postintervention. By showing the potential for treatment
KBIT scores also remained a significant predictor of the T3 benefits to last over time, to children’s entry into formal
composite for the control group (estimate = .05, p < .01). reading instruction, ECSE educators and other early inter-
These two results indicate that children with higher print- ventionists can potentially affect the longitudinal reading
knowledge scores and nonverbal cognition at T1 had higher trajectory of children highly vulnerable to developing read-
print-knowledge scores at T3, about 18 months later. In ing difficulties.
Model 2, the treatment showed a nonsignificant trend in Here, we discuss three points of most significance
favor of the experimental condition (2.55, p = .06), suggest- regarding the results presented here. First, why it is that
ing that children in the treatment had slightly higher print- treatment effects continue to be apparent 1-year post-
knowledge scores at T3 than those in the comparison group. intervention? Second, how are we to interpret the size of
The interaction term (estimate = −.03, p = .07) suggests that the effect observed? Third, why does children’s nonverbal
the gap between the control and treatment group narrows cognition serve to moderate the treatment effects?
as KBIT scores increase—that the treatment is less effective Considering the first point, we must consider why
for children who score higher on the KBIT. it is that treatment effects continued to be observed 1-year
A graph depicting the moderating effect of children’s postintervention, albeit as a statistical trend (p = .06) that
KBIT on the treatment effect is presented in Figure 1. In does not meet the conventional alpha level of .05 but that,
Figure 1, separate lines are provided for the control group as an effect-size estimate (d = 0.2), is coherent with the effect
and the treatment group, the x-axis is children’s KBIT score, observed 1 year prior for children who received the inter-
and the y-axis is print knowledge at kindergarten. As can vention (d = 0.21). Treatment effects historically were inter-
be seen, the largest difference between the comparison and preted only in light of the effects of statistical testing, in
experimental groups occurs at the lowest levels of nonverbal which inferential statistics help one to determine whether
cognition (M = 55), the results of which were reported in one’s tested hypothesis should be rejected or accepted.
Table 4. A simple slopes analysis (Preacher, Curran, & Bauer, Experts have argued that the threshold for significance, often-
2006) suggested a similarly nonsignificant trend favoring times at p = .05 to differentiate chance from nonchance,
the experimental group when KBIT = 70 (estimate = −.82, is not only arbitrary but offers little to consumers with
p = .09). The simple slopes analyses also showed that respect to how to interpret an outcomes (see McCartney &
differences between the mean predicted scores of two Rosenthal, 2000). For instance, the hypothesis tested in
groups at the higher levels of nonverbal cognition of 85 this study is that the print-focused read-aloud intervention
and 100 were not statistically significant ( p = .44 and would have positive effects on children’s print knowledge.
.45, respectively). By applying a strict, arbitrary significance test on the basis
of p = .05, we would need to reject this hypothesis and
accept the null hypothesis that the intervention had no dis-
Discussion cernible effect on children. However, experts would argue
This study makes a significant contribution to our that interpretation of the effect-size estimate is as appropriate
understanding of practices that can support the early-literacy as interpreting the inferential test result, especially given
development of children with LI, representing one of only that effect sizes help one to interpret the practical significance

Figure 1. T3 print-knowledge composite as a function of treatment and children’s nonverbal cognition.
of a work. Therefore, for the present purposes, we acknowl- Even within the context of formal schooling and receipt
edge the result of the inferential test statistics but elect to of special-education services, children with LI significantly
not arbitrarily reject the tested hypothesis. lag behind their typical peers in the development of reading
To this end, we interpret the results of this 1-year across the primary grades (Morgan et al., 2011). Thus, at
intervention follow-up favorably, noting that the observed least in part, it may be that treatment effects are sustained
effect size suggests there to be continued impacts of print- over time due to the counterfactual, in that children with LI
focused read-alouds 1 year after the close of the intervention, are likely developing their print knowledge only slowly over
with those in the experimental condition performing one- time, outside of an explicit intervention context targeting
fifth of a standard deviation better than those in the control print-knowledge development.
group in their print knowledge. This modest yet sustained Considering the second point, it is worth considering
effect is coherent with the results of a separate study exam- how best to interpret the size of the effect observed, which
ining 1- and 2-year effects of print-focused read-alouds for corresponded to a 0.2 SD unit on the print-knowledge
children from low-income homes (Piasta et al., 2012). This composite. Interpretation of effect-size magnitude is often
study reported a postintervention effect at 1 year of 0.26 and made relative to Cohen’s benchmarks, such that 0.2 is small,
0.27 at 2 years, finding that preschool children who experi- 0.5 is medium, and 0.8 is large (Cohen, 1988). However,
enced print-focused read-alouds performed about one-fourth these benchmarks are generic and not tied to a given sub-
of a standard deviation better in reading than those in a stantive field. As Hill et al. (2008) pointed out, effect sizes
comparison condition in kindergarten and first grade. are best interpreted empirically and in relation to what is
The observation that treatment effects are maintained normative for a given aspect of development or, in this case,
over a year, resulting in a difference in the print knowledge a particular type of intervention. Hill et al. conducted a
between children who did and did not experience classroom- meta-analysis, in which they aggregated the mean effect sizes
based print-focused read-alouds in the prior year, seems to reported in 76 past meta-analyses of educational interventions
beg the question as why the children in the control condition (kindergarten through twelfth grade) employing the RCT
did not catch up with those in the treatment condition in design, which represents the most “rigorous impact design
their print knowledge. Catch-up by those in the control possible.” The mean effect extracted from 76 studies for
condition may be the culprit behind the fade-out effect children in the early elementary grades (Grades 1–3, the
commonly seen in studies of early-childhood interventions lowest grade range available and most comparable to the
(Barnett, 2011). That the children in the control condition present sample) was 0.23 (SD = 0.21). This benchmark
did not catch up may be indicative of the fact that children suggests that print-focused read-alouds yield effects that
with LI tend to have more difficulties developing print knowl- are comparable to those seen in the random-assignment
edge than typical children (Boudreau & Hedberg, 1999; literature. Therefore, we can interpret the use of print-
Gillam & Johnston, 1985) and that intensive experiences focused read-alouds as having a sustained effect on student
with print (such as that received by children in the treatment achievement that is comparable to that seen in RCTs more
condition in this study) are necessary for achieving catch-up. generally.

A noteworthy point regarding the intervention tested the relations between children’s nonverbal cognition and
here, in line with the size of the effects observed, is the costs their overall language skills. In the sample of children
associated with the intervention and the time required to involved in the present study, the relations between the
implement print-focused read-alouds. The intervention two variables was significant and moderate in size, r = .43
requires the purchase of a set of 30 commercially available ( p < .001). The relations between these two variables also
storybooks (approximately $200) and download of the implies that the intervention was particularly beneficial for
treatment manual, which is available at no charge on the children with lower levels of language skill and with most
internet (http://ccec.ehe.osu.edu/files/2014/07/Engaging- profound LI. This is an exciting finding, as children with
Children-with-Print-Building-Early-Literacy-Skills.pdf). more profound LI may be at greatest risk for future reading
The costs borne for implementation are far less than many difficulties (Catts et al., 2002). Few studies to date have
of the commercially available early-literacy interventions focused specifically on this sector of children with LI, but
that have comparable impacts. For instance, the effect the present results suggest that they may be particularly
size estimates reported for print knowledge (d = 0.37) by responsive to early-literacy interventions. Substantiating this
Debaryshe and Gorecki (2007) involved material costs possibility should be a targeted area of future investigation.
exceeding $3,000 per classroom and intensive coaching of In sum, the present study represents an important
teachers. In addition, the time needed to implement the contribution to the literature, in that it is one of the few
intervention averaged about 20 min per day, four times studies that has conducted follow-up investigation of children
per week. For ECSE teachers who already provide daily with LI who had earlier received an early-literacy inter-
read-alouds to their students, the print-focused intervention vention. The benefits children derived from this intervention
requires no additional time for implementation. Therefore, with respect to their early-literacy skills are not only appar-
the intervention is likely to more readily fit into ECSE ent in the short-term but also appeared to be sustained in
classrooms as compared with other literacy interventions size for at least one additional year. The finding is consis-
that require considerable instructional time or that differ tent with that seen for children who were typically develop-
from the typical daily routine. ing but from low-income households (Piasta et al., 2012),
Considering the third point, we want to elaborate who had better reading and spelling skills 1- and 2-years
on the treatment effects with respect to the moderating role postintervention when they experienced print-focused read-
of children’s nonverbal cognition, a result seen also in the alouds during preschool compared with a control condition.
analysis of short-term effects of print-focused read-alouds Limitations of this work warranting note are as
(Justice et al., 2015). Children with low nonverbal cognition, follows. First, the 1-year follow-up in this study involved
comorbid with LI, seemed to show special benefit from only a subset of those who were initially enrolled in the
being in the treatment classrooms during the preschool RCT, in part due to study attrition over time. We cannot
year, with results largely holding to 1-year postintervention. know if the results would be similar if long-term effects
Why is it that these children seem to derive particular could have been investigated for the entire sample. Second,
benefit? Several studies involving secondary analyses of the the follow-up assessment in this study included only mea-
data set used in this study provide some potential guidance sures of print knowledge. It is not clear if the effects observed
on this issue. First, Dynia, Lawton, Logan, and Justice would generalize to standardized measures of word read-
(2014) examined the preschool home-literacy experiences ing, although there are strong correlations between the
of 30 children with autism spectrum disorder represented measures used in this study and children’s word reading
in this sample (all of whom had low nonverbal cognition skills (Lonigan & Shanahan, 2009). Third, this study included
and LI) as compared with children in their classrooms who only a 1-year post-intervention time point. We cannot
were typically developing. The researchers found that the know if the effects we see are sustained longer term to miti-
children with autism spectrum disorder had significantly gate children’s future risks for reading difficulties. Fourth,
fewer literacy experiences at home, comparatively. Thus, the study methods did not involve examination of chil-
it seems that children with significant disabilities have few dren’s learning experiences in the year following the inter-
experiences to develop print knowledge at home and thus vention (i.e., between T2 and T3). It would be helpful to
derive special benefits from classroom-based print-targeted understand how these experiences may have mitigated or
interventions. Second, Y. Guo, Sawyer, Justice, and Kaderavek enhanced intervention effects. Last, the extent to which
(2013) examined the classroom literacy environment for these results may generalize to other children with LI is
54 of the ECSE classrooms in which children were enrolled unknown. Our sample was ascertained from ECSE settings
during the preschool year, in which they received the inter- serving children with LI as well as other children with dis-
vention or control conditions. The study showed that abilities. Relatively few children spoke languages other
the literacy environment in many classrooms was low-to- than English. It may be that different results would be seen
moderate quality. Thus, it may be that the print-focused with a more diverse sample or for children not enrolled in
read-alouds compensates for a literacy environment that ECSE programs. With these limitations in mind, the results
is insufficiently robust to develop children’s early-literacy presented here can, we hope, serve as an impetus for future
skills. longitudinal work in which the effects of early-literacy
Further interpretation of the result regarding children’s interventions are modeled long into children’s future as
nonverbal cognition as an intervention moderator concerns readers.

Acknowledgments vocabulary delays. Journal of Speech and Hearing Research,
39, 1274–1283.
The research reported herein was funded by the Institute Guo, L.-Y., & Schneider, P. (2016). Differentiating school-aged
of Education Sciences grant #R324A080037. We are grateful for children with and without language impairment using tense
the many families, children, teachers, and research staff who were and grammaticality measures from a narrative task. Journal of
involved in this study. Speech, Language, and Hearing Research, 59, 317–329.
Guo, Y., Dynia, J. M., Logan, J. A., Justice, L. M., Breit-Smith,
A., & Kaderavek, J. N. (2016). Fidelity of implementation for
References an early-literacy intervention: Dimensionality and contribution
to children’s intervention outcomes. Early Childhood Research
Barnett, W. S. (2011, August 19). Effectiveness of early educa-
Quarterly, 37, 165–174.
tional intervention. Science, 333, 975–978.
Guo, Y., Sawyer, B., Justice, L. M., & Kaderavek, J. (2013). Qual-
Boudreau, D. M., & Hedberg, N. L. (1999). A comparison of early
ity of the literacy environment in inclusive early childhood
literacy skills in children with specific language impairment
special education classrooms. Journal of Early Intervention,
and their typically developing peers. American Journal of
35(1), 40–60.
Speech-Language Pathology, 8, 249–260.
Hill, C. J., Bloom, H. S., Black, A. R., & Lipsey, M. W. (2008).
Cabell, S. Q., Justice, L. M., Zucker, T. A., & McGinty, A. S.
Empirical benchmarks for interpreting effect sizes in research.
(2009). Emergent name-writing abilities of preschool-age
Child Development Perspectives, 2, 172–177.
children with language impairment. Language, Speech, and
Hearing Services in Schools, 40, 53–66. Invernizzi, M. A., Sullivan, A., Meier, J. D., & Swank, L. (2004).
Cabell, S. Q., Lomax, R. G., Justice, L. M., Breit-Smith, A., Skibbe, Phonological awareness Literacy Screening for Preschool:
L. E., & McGinty, A. S. (2010). Emergent literacy profiles Teacher’s manual. Charlottesville, VA: University of Virginia.
of preschool-age children with specific language impairment. Justice, L. M., Bowles, R. P., & Skibbe, L. E. (2006). Measuring
International Journal of Speech-Language Pathology, 12, preschool attainment of print-concept knowledge: A study
472–482. of typical at-risk 3- to 5-year-old children using item response
Catts, H. W., Fey, M. E., Tomblin, J. B., & Zhang, X. (2002). theory. Language, Speech, and Hearing Services in Schools,
A longitudinal investigation of reading outcomes in children 37, 224–235.
with language impairments. Journal of Speech, Language, Justice, L. M., Chow, S.-M., Capellini, C., Flanigan, K., & Colton,
and Hearing Research, 45, 1142–1157. S. (2003). Emergent literacy intervention for vulnerable pre-
Cohen, J. (1988). Statistical power analysis for the behavioral schoolers: Relative effects of two approaches. American Journal
sciences. Mahwah, NJ: Erlbaum. of Speech-Language Pathology, 12, 320–332.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied Justice, L. M., & Ezell, H. K. (2001). Word and print awareness
multiple regression/correlation analysis for the behavioral sciences. in 4-year-old children. Child Language Teaching and Therapy,
Mahwah, NJ: Erlbaum. 17, 207–225.
Dale, P. S., Price, T. S., Bishop, D. V., & Plomin, R. (2003). Out- Justice, L. M., & Ezell, H. K. (2002). Use of storybook reading to
comes of early language delay: I. Predicting persistent and increase print awareness in at-risk children. American Journal
transient language difficulties at 3 and 4 years. Journal of of Speech-Language Pathology, 11, 17–29.
Speech, Language, and Hearing Research, 46, 544–560. Justice, L. M., Kaderavek, J., Bowles, R., & Grimm, K. (2005).
DeBaryshe, B. D., & Gorecki, D. M. (2007). An experimental Language impairment, parent-child shared reading, and phono-
validation of a preschool emergent literacy curriculum. Early logical awareness. Topics in Early Childhood Special Education,
Education and Development, 18, 93–110. 25, 143–156. https://doi.org/10.1177/02711214050250030201
Dynia, J. M., Lawton, K., Logan, J. A., & Justice, L. M. (2014). Justice, L. M., Kaderavek, J., Fan, X., Sofka, A., & Hunt, A.
Comparing emergent-literacy skills and home-literacy environ- (2009). Accelerating preschoolers’ early literacy development
ment of children with autism and their peers. Topics in Early through classroom-based teacher–child storybook reading
Childhood Special Education, 34, 142–153. and explicit print referencing. Language, Speech, and Hearing
Ezell, H. K., Justice, L. M., & Parsons, D. (2000). Enhancing the Services in Schools, 40, 67–85.
emergent literacy skills of preschoolers with communication Justice, L. M., Logan, J. A. R., Kaderavek, J. N., & Dynia, J. M.
disorders: A pilot investigation. Child Language Teaching and (2015). Print-focused read-alouds in early childhood special
Therapy, 16, 121–140. education programs. Exceptional Children, 81, 292–311.
Fey, M. E., Cleave, P. L., Long, S. H., & Hughes, D. L. (1993). Justice, L. M., Logan, J. R., & Damschroder, L. (2015). Designing
Two approaches to the facilitation of grammar in children with caregiver-implemented shared-reading interventions to overcome
language impairment: An experimental evaluation. Journal of implementation barriers. Journal of Speech, Language, and
Speech and Hearing Research, 36, 141–157. Hearing Research, 58, S1851–S1870.
Fritz, C. O., Morris, P. E., & Richler, J. J. (2012). Effect size esti- Justice, L. M., Skibbe, L. E., McGinty, A. S., Piasta, S. B., &
mates: Current use, calculations, and interpretation. Journal Petrill, S. (2011). Feasibility, efficacy, and social validity of
of Experimental Psychology: General, 141, 2–18. home-based storybook reading intervention for children with
Gillam, R. B., & Johnston, J. R. (1985). Development of print language impairment. Journal of Speech, Language, and Hear-
awareness in language-disordered preschoolers. Journal of ing Research, 54, 523–538.
Speech and Hearing Disorders, 28, 521–526. Justice, L. M., & Sofka, A. E. (2013). Engaging children with
Gillon, G. T. (2002). Follow-up study investigating the benefits of print: Building early literacy skills through quality read-alouds.
phonological awareness intervention for children with spoken New York, NY: Guilford.
language impairment. International Journal of Language & Kaufman, A. S., & Kaufman, N. L. (1985). Kaufman Test of Educa-
Communication Disorders, 37, 381–400. tional Achievement. Circle Pines, MN: American Guidance Service.
Girolametto, L., Pearce, P. S., & Weitzman, E. (1996). Inter- Lonigan, C. J., & Shanahan, T. (2009). Developing early literacy:
active focused stimulation for toddlers with expressive Report of the National Early Literacy Panel. Executive summary.

A scientific synthesis of early literacy development and impli- shared reading: Longitudinal effects on literacy achievement.
cations for intervention. Jessup, MD: National Institute for Child Development, 83, 810–820. https://doi.org/10.1111/j.1467-
Literacy. 8624.2012.01754.x
Lonigan, C. J., & Whitehurst, G. J. (1998). Relative efficacy of Pratt, A. S., Justice, L. M., Perez, A., & Duran, L. K. (2015).
parent and teacher involvement in a shared-reading inter- Impacts of parent‐implemented early‐literacy intervention for
vention for preschool children from low-income backgrounds. Spanish‐speaking children with language impairment. Inter-
Early Child Research Quarterly, 13, 263–290. national Journal of Language & Communication Disorders,
Lovelace, S., & Stewart, S. R. (2007). Increasing print aware- 50, 569–579.
ness in preschoolers with language impairment using non- Preacher, K. J., Curran, P. J., & Bauer, D. J. (2006). Computational
evocative print referencing. Language, Speech, and Hearing tools for probing interactions in multiple linear regression,
Services in Schools, 38, 16–30. https://doi.org/10.1044/0161- multilevel modeling, and latent curve analysis. Journal of Edu-
1461(2007/003) cational and Behavioral Statistics, 31, 437–448.
Lundberg, I., Frost, J., & Petersen, O.-P. (1988). Effects of Skibbe, L. E., Grimm, K. J., Stanton-Chapman, T. L., Justice,
an extensive program for stimulating phonological aware- L. M., Pence, K. L., & Bowles, R. P. (2008). Reading trajec-
ness in preschool children. Reading Research Quarterly, 23, tories of children with language difficulties from preschool
263–284. through fifth grade. Language, Speech, and Hearing Services
McCartney, K., & Rosenthal, R. (2000). Effect size, practical in Schools, 39, 475–486.
importance, and social policy for children. Child Development, Tomblin, J. B., Records, N. L., Buckwalter, P., Zhang, X., Smith,
71, 173–180. E., & O’Brien, M., (1997). Prevalence of specific language
McClintock, B., Pesco, D., & Martin‐Chang, S. (2014). Think- impairment in kindergarten children. Journal of Speech, Lan-
ing aloud: Effects on text comprehension by children with guage, and Hearing Research, 40, 1245–1260.
specific language impairment and their peers. International van Bysterveldt, A. K., Gillon, G. T., & Moran, C. (2006). Enhanc-
Journal of Language & Communication Disorders, 49, ing phonological awareness and letter knowledge in preschool
637–648. children with Down syndrome. International Journal of Dis-
McGinty, A. S., & Justice, L. M. (2009). Predictors of print knowl- ability, Development and Education, 53, 301–329.
edge in children with specific language impairment: Experiential Wagner, R. K., & Torgesen, J. K. (1987). The nature of phono-
and developmental factors. Journal of Speech, Language, and logical processing and its causal role in the acquisition of
Hearing Research, 52(1), 81–97. reading skills. Psychological Bulletin, 101, 192–212.
McGinty, A. S., Justice, L. M., Piasta, S. B., Kaderavek, J., & Whitehurst, G. J., & Lonigan, C. J. (1998). Child development
Fan, X. (2012). Does context matter? Explicit print instruction and emergent literacy. Child Development, 69, 848–872.
during reading varies in its influence by child and classroom Wiig, E. H., Secord, W. A., & Semel, E. (2004). Clinical Evaluation
factors. Early Child Research Quarterly, 27, 77–89. https://doi. of Language Fundamentals Preschool (2nd ed.). San Antonio,
org/10.1016/j.ecresq.2011.05.002 TX: Harcourt Assessment.
McNamara, J. K., Vervaeke, S.-L., & Van Lankveld, J. (2008). Wright, A., Mitchell, S., O’Donoghue, A., Cowhey, S., & Kearney,
An exploratory study of emergent literacy intervention for M. (2015). Making sense of it: A brief programme to improve
preschool children with language impairments. Exceptionality reading comprehension in adolescents with language impairments
Education International, 18, 9–32. in main stream school. International Journal of Language &
Morgan, P. L., Farkas, G., & Wu, Q. (2011). Kindergarten chil- Communication Disorders, 50, 776–787.
dren’s growth trajectories in reading and mathematics: Who Ziolkowski, R. A., & Goldstein, H. (2008). Effects of an embedded
falls increasingly behind? Journal of Learning Disabilities, 44, phonological awareness intervention during repeated book
472–488. https://doi.org/10.1177/0022219411414010 reading on preschool children with language delays. Journal
Nancollis, A., Lawrie, B.-A., & Dodd, B. (2005). Phonological of Early Intervention, 31, 67–90.
awareness intervention and the acquisition of literacy skills in Zucker, T. A., Cabell, S. Q., Justice, L. M., Pentimonti, J. M., &
children from deprived social backgrounds. Language, Speech, Kaderavek, J. N. (2012). The role of frequent, interactive pre-
and Hearing Services in Schools, 36, 325–335. kindergarten shared reading in the longitudinal development
Piasta, S. B., Justice, L. M., McGinty, A. S., & Kaderavek, J. N. of language and literacy skills. Developmental Psychology, 49,
(2012). Increasing young children’s contact with print during 1425–1439. https://doi.org/10.1037/a0030347

Appendix
Scope and objectives for print-focused read-alouds (with sample references to print).
Objectives Sample print reference
Instructional Domain 1: Book and print organization

1. Page order: Knows the order in which pages are read in a I am going to read this page first and then this page over
book. here next.
2. Author: Knows the role of the author. The author, Eric Carle, wrote all the words in this book.
3. Page organization: Knows that reading occurs from the top This is the top of the page. This is where I begin reading.
of the page to the bottom of the page.
4. Title of book: Knows the role of the title of the book. This is the title of the book. It tells us the name of the book.
5. Print direction: Knows that reading occurs from left to right. I start reading here and I read this way.
Instructional Domain 2: Print meaning
1. Print function: Understands the relationship between meaning Here are the penguin’s words. He says, “thank you.”
and print.
2. Environmental print: Knows the purpose of print embedded This sign says “Stop.”
within the environment.
3. Metalinguistic concept of reading: Understands the meaning We’re going to read these words; what will these words
behind reading and the contexts in which reading occurs. tell us?
Instructional Domain 3: Letters
1. Upper- and lowercase letters: Knows letters come in This M is an uppercase letter. See how it’s bigger than
upper- and lowercase forms. these lowercase letters?
2. Names of letters: Knows the names of the majority of What is this letter?
uppercase letters.
3. Metalinguistic concept of letter: Knows that letters are Show me one letter in this word.
symbols used in written language.
Instructional Domain 4: Words
1. Word identification: Identifies some written words in This word is “the” —this word is in this book all the time,
familiar contexts. can you help me find it?
2. Short versus long words: Knows that words can contain This word is “lollapaloosh.” It is a long word. It has a lot
various numbers of letters. of letters in it! Let’s count all the letters.
3. Letters versus words: Knows that letters make up words. This is the letter K. K is in the words kangaroo and kick.
4. Concept of word in print: Represents the systematic Let’s point to each word as I read it. Ready?
relationship between spoken words and written words.


Justice, 2017

Uploaded by

Copyright:

Available Formats

Justice, 2017

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Justice, 2017

Uploaded by

Copyright:

Available Formats

AJSLP

Longitudinal Impacts of Print-Focused

2 American Journal of Speech-Language Pathology • 1–14

Downloaded From: http://ajslp.pubs.asha.org/pdfaccess.ashx?url=/data/journals/ajslp/0/ by a Univ Of Newcastle User on 04/07/2017

Justice et al.: Longitudinal Impacts 3

Note. CELF Core Language = Core Language Composite of the

4 American Journal of Speech-Language Pathology • 1–14

Downloaded From: http://ajslp.pubs.asha.org/pdfaccess.ashx?url=/data/journals/ajslp/0/ by a Univ Of Newcastle User on 04/07/2017

Justice et al.: Longitudinal Impacts 5

6 American Journal of Speech-Language Pathology • 1–14

Downloaded From: http://ajslp.pubs.asha.org/pdfaccess.ashx?url=/data/journals/ajslp/0/ by a Univ Of Newcastle User on 04/07/2017

Justice et al.: Longitudinal Impacts 7

8 American Journal of Speech-Language Pathology • 1–14

Downloaded From: http://ajslp.pubs.asha.org/pdfaccess.ashx?url=/data/journals/ajslp/0/ by a Univ Of Newcastle User on 04/07/2017

Justice et al.: Longitudinal Impacts 9

10 American Journal of Speech-Language Pathology • 1–14

Downloaded From: http://ajslp.pubs.asha.org/pdfaccess.ashx?url=/data/journals/ajslp/0/ by a Univ Of Newcastle User on 04/07/2017

Justice et al.: Longitudinal Impacts 11

12 American Journal of Speech-Language Pathology • 1–14

Downloaded From: http://ajslp.pubs.asha.org/pdfaccess.ashx?url=/data/journals/ajslp/0/ by a Univ Of Newcastle User on 04/07/2017

Justice et al.: Longitudinal Impacts 13

Objectives Sample print reference

Instructional Domain 1: Book and print organization

14 American Journal of Speech-Language Pathology • 1–14

Downloaded From: http://ajslp.pubs.asha.org/pdfaccess.ashx?url=/data/journals/ajslp/0/ by a Univ Of Newcastle User on 04/07/2017

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.