Adult Talk During Book Reading for Preschoolers With ASD: Links With Literacy Outcomes

Abstract

Calls for interventions in early childhood that best prepare children with autism spectrum disorder (ASD) to have inclusive school experiences have been numerous. Although there is a developing literature on the efficacy of interactive book reading (IBR) for this population, it is still unclear which aspects of IBR benefit these children and in what way. To address this gap, we reanalyzed data from a recently completed experiment comparing IBR with other early literacy instruction for preschoolers with ASD. Our results showed that frequency of tutors’ use of completion prompts used during IBR was uniquely predictive of better gains in phonological awareness. In addition, tutors’ frequency using open-ended prompts was associated with increased gains in print knowledge. Furthermore, both findings held true after controlling for language gains. Results have implications for the types of structure and support that adults might provide young children with ASD during IBR.

Keywords

autism early intervention exceptionalities reading

Autism spectrum disorder (ASD) is a pervasive neurological disorder with a wide range of impacts, outcomes, and varying levels of severity (American Psychiatric Association, 2013). Children with ASD show relative strengths and equal performance compared with their typically developing (TD) peers in discrete skills such as letter naming and relative weaknesses in meaning-related skills, phonological awareness (PA), and vocabulary development (Dynia et al., 2014; Lanter et al., 2012). Development of oral language in early childhood is critical for emergent and subsequent literacy skills for both TD children and children with developmental disabilities such as ASD (Boudreau & Hedberg, 1999; Lonigan et al., 2007; Scarborough & Dobrich, 1994; Tager-Flusberg et al., 2001; Whitehurst & Lonigan, 1998).

For all children, the development of oral language begins early in life through guided practice with caregivers and interaction with other speakers (Bishop & Adams, 1990; Scarborough, 1989). The term “guided practice” has been used to describe an activity that children engage in with a partner of varying skill who provides guidance and support (Rogoff, 1993). Oral language development is strongly related to important emergent literacy skills that all children need to be successful readers and writers (Storch & Whitehurst, 2002). Extant studies have demonstrated a longitudinal relationship between oral language ability and reading proficiency in children who are TD, children with reading delays, and those with language delays (Bishop & Adams, 1990; Scarborough, 1989). Some aspects of oral language are more closely related to later literacy outcomes than others. The National Early Literacy Project (NELP) report (2008) found that aspects of oral language and listening comprehension have substantial predictive relations with later conventional reading skills. The report also found that literacy skills and oral language are strongly linked and have bidirectional ties. Children who have language and can map complex sentence structures onto corresponding meanings have an advantage over children with poorer language skills learning to read (Shanahan & Lonigan, 2010).

Intervening in oral language development holds promise for enhancing long-term reading development and future school outcomes for children with developmental disabilities such as ASD (Cain & Oakhill, 2008). Furthermore, supported acts of intentional communication where joint attention is maintained and language use is aligned with children’s area of focus also effectively support language development (Perryman et al., 2013; Siller & Sigman, 2002, 2008; Toth et al., 2006). Shared picture book reading during early childhood serves as an ideal forum for intentional communication that can support language development.

Interactive Book Reading

Shared picture book reading is a common way children are introduced to many emergent literacy skills such as new vocabulary words (Rosenquest, 2002). Unfortunately, children with disabilities have historically had little exposure to shared book reading for many reasons, including caregivers’ lack of knowledge of what is effective, belief that shared book reading is not a worthwhile activity for this population, and active resistance by children themselves (Koppenhaver et al., 2001; Light et al., 1994; Light & Smith, 1993; Marvin & Mirenda, 1993; Scarborough & Dobrich, 1994). Taken together, these issues can lead to lower levels of engagement and interaction, and may explain some of the variance in children’s emergent literacy outcomes in early childhood (Tsybina & Eriks-Brophy, 2010; Warren & Yoder, 1996).

Interactive Book Reading (IBR) was designed as a shared picture book reading intervention that teachers, parents, and other school personnel could use to promote oral language development (Whitehurst et al., 1988). In Dialogic Reading (DR), a form of IBR, adult caregivers employ a set of language prompts to engage children in oral language practice during picture book reading experiences. Research shows that DR can foster vocabulary and oral language development in TD children as well as populations such as low-income children and children of immigrants, thereby improving emergent literacy skills (Chow & McBride-Chang, 2003; Storch & Whitehurst, 2002).

IBR for Children With ASD

Research on IBR for children with ASD is a relatively new endeavor. Nevertheless, recent work has shown that IBR is effective for increased verbal participation and lengthened engagement with book reading (Fleury et al., 2014), increased book-specific vocabulary (Fleury & Schwartz, 2017; Towson et al., 2016), expressive vocabulary and listening comprehension (Hudson et al., 2017), story comprehension and engagement (Mucchetti, 2013), vocabulary (Rahn et al., 2016), and correct spontaneous responding during reading (Whalon et al., 2015).

The aim of IBR for young children with ASD follows the same logic as applications with TD children (Lonigan et al., 1999; Whitehurst et al., 1994; Whitehurst et al., 1988) to engage children in literacy activities that demand the kind of joint attention supportive of oral language and vocabulary development. Data from Hudson et al. (2017) provide the basis for the current inquiry. In that study, preschool children with ASD were randomly assigned to three conditions: IBR, phonological awareness (PA), and business as usual (BAU) control. Hudson et al. (2017) found that, immediately after intervention, children in the IBR group had greater growth in expressive vocabulary and listening comprehension in comparison with the PA condition or BAU. However, 1 year after intervention, 82% of the original sample was tested (N = 109), and the follow-up results showed that the PA intervention group had significantly better vocabulary and listening comprehension than the other interventions (average approximate d = 0.63) (Hudson et al., 2019). These conflicting findings showing better outcomes for the IBR treatment in the short-term and better outcomes for PA in the long-term warrant further scrutiny. In particular, the present study aims to identify which elements of the IBR adult–child language interaction might better facilitate language and emergent literacy skill development for children with ASD.

Adult Prompting During IBR

Previous scholarship has demonstrated a link between adult behavior and active engagement during IBR for both TD children and those with a language disorder (Kang et al., 2009; Luo & Tamis-LeMonda, 2017; McGinty et al., 2012). In addition, parents’ use of meaning-related prompts has been found to be associated with growth in children’s vocabulary (both expressive and receptive) as well as other print-related behavior and alphabet knowledge (Han & Neuharth-Pritchett, 2015; Hindman et al., 2008). Tipton et al. (2017) performed a factor analysis of parent behaviors during shared book reading and found that parents of children with ASD employ a variety of strategies during shared book reading. Specifically, the parents in their study were not given explicit instruction in specific prompt use during intervention, which provided a natural opportunity to observe the frequency of specific prompt types. Results showed that parents used Wh- questions (e.g., questions that begin with Wh- such as “What happened?”) more than open-ended or simple yes/no questions.

In its design, IBR employs two sets of prompts during sessions, called Prompt, Evaluate, Expand, and Repeat (PEER) and Completion, Recall, Open-ended, Wh, and Distancing (CROWD); the first is a guideline for the adult caregiver/tutor, and the second refers to types of prompting to be used with the children (Whitehurst et al., 1988). The caregiver/tutor prompts the child (using one of the CROWD prompts), evaluates the response from the child, asks the child to expand their response, or repeats the prompt if there has been no response after appropriate wait time.

These prompts ask the child to engage with the reading and are expanded on further in the discussion section. Below we define each with an example:

Completion prompts aim to increase the child’s attention during reading. An example would be an adult providing the prompt, “He ran through the_____?” and the expected response from the child would be, “woods.”

Recall prompts facilitate text to self-connections. An example would be the adult asking a question such as “Do you remember the boy’s name?” The expected response from the child would be the name of the male character being referred to, which is not available specifically on the page being read but requires that the child “Recall” a detail from the story previously read.

Open-ended prompts indicate a broad inquiry raised by the adult reading partner referring to events or story action currently in the dyads realm of focus. An example would be an adult asking, “Tell me what’s happening in this picture?” immediately after she has turned the page of the book.

Wh- prompts elicit details from the page being viewed and provide the opportunity for oral language practice and vocabulary building. The adult may ask, “What is the boy doing?”

Distancing prompts are opportunities for a text to life or text-to-self connection. In a book that has action that takes place in a zoo, the adult may ask, “Have you ever been to the zoo?”

Present Study

The purpose of the present study is to describe and evaluate how prompts used by tutors during IBR sessions predict language development for a cohort of young children with ASD. In school settings, children are likely to complete book reading sessions with the same tutor or tutors over the course of an academic year. A closer look at the types and amounts of prompts tutors use can help determine whether some prompts in IBR serve as active ingredients in the intervention. Active ingredients are “specific, measurable actions” a tutor performs that affect the mechanism of the intervention (Hart & Ehde, 2015). Characterizing IBR’s active ingredients for young children with ASD could lead to the development of stronger early reading interventions, thereby creating a clearer path to word reading and comprehension. Our research questions were as follows.

Research Question 1 (RQ1): Does the frequency and type of tutor prompting change from the beginning to the end of an IBR intervention for preschoolers with ASD?

Research Question 2 (RQ2): How do rates of tutor usage of different types of prompts (during an IBR intervention for preschoolers with ASD) correlate with each other?

Research Question 3 (RQ3): To what extent does tutor prompting used during IBR predict pretest–posttest change in language and literacy skills for preschoolers with ASD?

Method

Participants

Original study

Child participants for this study were drawn from extant data of one arm of an early literacy treatment comparison study for preschoolers with ASD (Hudson et al., 2017). Most of the children were male (77% of the sample), and ages ranged from 49 to 69 months (M = 56.22, SD = 3.78). In the original study, children with ASD were recruited each year for three consecutive years from 32 schools across eight school districts based on the following inclusion criteria: (a) had either a medical diagnosis of ASD or an educational identification; (b) had an active Individualized Education Program (IEP) and receive services for ASD; (c) were enrolled in their last year of preschool; (d) had no known co-occurring neurological or genetic disorders (e.g., Down syndrome, Fragile X); (e) exhibited functional communication such as the ability to make requests, used two- to three-word phrases, and were able to follow simple directions; and (f) a minimum standard score of 55 on the Expressive One Word Picture Vocabulary Test (Brownell, 2010).

After recruitment, children were randomized into treatment that was implemented individually by a trained tutor within the child’s classroom for 7 to 15 min per day, 3 to 4 days per week, from November to May. In the first trial, children were randomly assigned to IBR (treatment) or BAU (control) preschool literacy environment. In the second trial, a second cohort of preschoolers were randomly assigned to PA (treatment) or control conditions. In the third trial, children were randomly assigned to IBR or PA intervention. Results for the combined samples showed that IBR (n = 47) demonstrated greater gains in vocabulary and listening skills (approximate effect sizes of d* = 0.29 and 0.30), whereas PA (n = 42) had better PA gains (approximate effect size of d* = 0.39).

Present study

Data for the present study are based on N = 41 of the 47 children in the original study’s IBR arm; unfortunately, a cloud-based technology failure deleted one tutor’s videotapes before we could code it. However, when we compared the analytic sample available with the full sample from the original study, we found no significant differences between missing and non-missing children’s demographic characteristics, pretests, or pretest–posttest gains. Moreover, the one tutor who is missing from the present study was similar to the original sample of tutors (the missing tutor was White, female, had at least a 4-year degree, and had experience working with children with ASD). Descriptive statistics on child measures for our present study’s analytic sample are presented in Table 1.

Table 1.

Descriptive Statistics for Child Measures.

Child measures	Pretest	Posttest	Pre–post gain
Child measures	M (SD)	M (SD)	M (SD)
Autism Index Score	81.38 (18.77)	—	—
Receptive vocabulary	81.32 (17.93)	89.76 (17.93)	8.44 (11.94)
Expressive vocabulary	88.20 (17.51)	93.12 (21.40)	4.93 (10.47)
Listening comprehension	79.49 (16.38)	85.49 (18.09)	6.00 (10.69)
Print knowledge	102.12 (15.56)	105.54 (11.70)	3.41 (12.76)
Phonological awareness	73.78 (16.52)	84.85 (18.59)	11.07 (13.93)

Note. N = 41 children and 9 tutors, except N = 37 for Autism index due to missing data. Autism Index = Gilliam Autism Rating Scale–2 total index percentile (measured at pretest only); Receptive Vocabulary = Peabody Picture Vocabulary Test–IV standard score; Expressive Vocabulary = Expressive One-Word Picture Vocabulary Test–IV standard score; Listening Comprehension = Woodcock-Johnson III Listening Comprehension cluster standard score; Print Knowledge = Test of Preschool Early Literacy Print Knowledge standard score; Phonological Awareness = Test of Preschool Early Literacy Phonological Awareness standard score.

IBR Intervention

The IBR intervention and prompting sequence were adapted from DR procedures (Whitehurst et al., 1988). The 20 books used in the study included both expository and narrative text structures. All children engaged with the books in the same order, which was determined using random selection. To facilitate comprehension of the books and provide a familiar structure, each book was read 3 to 4 times. During each book reading session, the tutor and child sat together in a space such as the corner of a classroom, the hallway, or a separate small room.

Tutoring sessions occurred 3 to 4 days per week, for approximately 7 to 15 min per session. Sessions were provided by a total of 11 tutors hired by the project from the local school community. All were women with at least a bachelor’s degree. Children received an average of 63 IBR sessions from fall to spring. There was no tutor attrition. The tutor provided a CROWD prompt of their choice (recall that CROWD stands for completion, recall, open-ended, wh-, and distancing; see our previous description for details), waited for a response from the child, evaluated the statement, expanded the statement by adding more words or otherwise making the language more complex, and either repeated the prompt or asked the child to repeat new vocabulary or syntax. The CROWD prompts were freely chosen by the tutors to use as they deemed appropriate.

In addition to the CROWD prompts, the tutors used supports to provide more structure to the children with ASD. The two most frequent additional supports were: (a) providing extra wait time for children to process the information and respond, and (b) repeating the prompt. Other, less frequently used supports moved from less to more support and reduced language demands: providing the initial sound of the correct completion answer (e.g., This is a k—), providing choices (e.g., Is it a key or a banana?), asking yes/no questions (e.g., Is that a key?), and modeling the right answer (e.g., Say key).

Intervention Fidelity

Fidelity and engagement were measured by coding videotaped book reading sessions. Interrater reliability was established for all fidelity coding, and no data were collected until agreement reached 80%.

Procedural fidelity was measured using a checklist of expected tutor behaviors appropriate to the intervention and a 5-point rating scale for each behavior—for example, “Adult prompts the child to respond at least one time every 2 to 3 pages” (1 = 0%–10%, 5 = 90%–100%). Graduate research assistants and the first author individually observed a videotaped session for each child 2 to 6 times during treatment (average of five per child). The number of points were added up and divided by the total possible to determine a percentage of earned points. Across observations, procedural fidelity averaged 98% (SD = 5%); for the IBR intervention-specific procedures, the average fidelity was 98% (SD = 7%).

In addition to fidelity, on-task engagement was measured to quantify the extent to which children were on-task to the intervention using a 15-s partial interval time sampling procedure on the same videotaped session used for fidelity. The five CROWD prompts and on-task engagement were coded by a member of the research team (the first, third, or fourth author), observed the first 5 min of each time sample, and, within each 15-s period, marked whether the student was actively engaged or not.

Active “on-task” engagement was defined by behaviors such as looking at the object of importance (book, blocks, or picture), pointing, speaking about the topic, looking at the tutor, or leaning toward the activity. Non-engagement (off-task) was defined by behaviors such as looking away from the activity or tutor, moving away from the activity or tutor, speaking about an unrelated topic, pushing the object away, or playing with an object unrelated to the activity at hand. Across sessions, the IBR condition averaged 86% on-task (SD = 19%).

Measures

Tutor prompting types

Video from two IBR sessions for 41 of the original study’s 47 children (in the IBR condition) was available for coding and analysis (due to a cloud storage technology failure, videos from a 10th tutor who tutored six children are missing). Because time-per-session varied, we standardized our definition of beginning- and end-of-session timing. The beginning of the session was defined as when the tutor presented the book or began asking questions about the book; the end of session was defined as when all book-related talk had concluded. Average tutoring session length at the beginning of the intervention was M = 10.28 min (SD = 2.92), and the average length at the end of intervention was M = 11.17 min (SD = 2.37); there was no significant difference between beginning and end of intervention session lengths (t test p = .372).

Interrater reliability

Prior to coding the study IBR sessions, two coders used the tutor prompting and child engagement coding definitions previously described to code non-study videotapes with a checklist adapted from Whitehurst et al. (1988) and Fleury et al. (2014). Once the coders reached 80% agreement on three out of five consecutive videos, a randomly selected 30% of IBR study sessions were coded a second time to establish interrater reliability. Reliability was calculated by dividing the number of agreements by the number of agreements plus the number of disagreements multiplied by 100. Interrater agreement was calculated for each with averages ranging from 95% to 97% (range = 80%–100%) for responses, 95% to 98% (range = 75%–100%) for verbal initiations, and 89% to 96% (range = 75%–100%) for nonverbal initiations.

Prompt per minute rate

To obtain the prompt per minute (PPM) rate used per session, the number of prompts for each CROWD category was divided by the number of minutes in the session. This was necessary because session lengths varied and a PPM rate allowed for more accurate comparison. Because we did not wish to assume that individual tutor–child pair observations were a valid measure of each tutor’s general approach to prompting, we aggregated individual prompt rates to the tutor level for use in subsequent analyses as a more robust measure of tutor prompting.

Child language skills

We used all three pretest–posttest measures of language skills from the original study, including receptive vocabulary, expressive vocabulary, and listening comprehension. Receptive vocabulary was measured using the Peabody Picture Vocabulary Test–IV (PPVT-IV; Dunn & Dunn, 1997). In this test, children point to a picture that best illustrates the meaning of an orally presented stimulus word. Internal consistencies (Cronbach’s alpha) from the original study were .98 at pretest and .96 at posttest. Expressive vocabulary was assessed using the Expressive One-Word Picture Vocabulary Test–IV (Martin & Brownell, 2011). For this test, children are presented with pictures and are asked to orally provide the names of the pictures. Sample-based internal consistencies from the original study were .95 for pretest and .96 for posttest. Listening comprehension was measured using the Oral Communication and Following Directions subtests of the Woodcock-Johnson III Tests of Achievement (WJII; Woodcock, 1997; Woodcock et al., 2007). Because the children had great difficulty following directions from an audio recording, all items were read aloud verbatim by assessors. Items were read one time and were not repeated. The Oral Communication measure consists of oral cloze statements. Sample-based internal consistencies using Cohort 1 were .86 for pretest and .89 for posttest. The Following Directions measure consists of items that ask participants to point to things on a picture after being given directions. Sample internal consistencies from the original study were .93 for pretest and .92 for posttest. Across all measures, norm-referenced standard scores (adjusted for age) were used.

Child early literacy skills

For the present study, we focus on two measures of early literacy from the original study: print knowledge and PA. Print knowledge was measured using the Test of Preschool Early Literacy (TOPEL; Lonigan et al., 2007) Print Knowledge subtest, which measures early knowledge about written language conventions and alphabet knowledge and sounds. This 36-item subtest measures early knowledge about written language conventions and alphabet knowledge and sounds. The raw score was converted to a norm-referenced standard score. Sample-based internal consistencies from the original study were .94 for pretest and .91 for posttest. Phonological awareness was measured using the TOPEL Phonological Awareness subtest, a standardized assessment used for identifying preschoolers at risk of literacy problems. Sample-based internal consistencies from the original study were .93 and .90 for pretest and posttest, respectively. Across all measures, norm-referenced standard scores (adjusted for age) were used.

Data Analysis Plan

Missing data

For brevity, participants who were missing data for the IBR intervention, or who were missing the majority of data across IBR sessions, were treated as “missing at random” and subsequently excluded from analyses. Recall that the final data used in analysis included nine tutors and their 41 children; all children received a uniform IBR intervention, but tutors were free to employ whatever tutor prompts they chose in interacting with the children.

RQ1 and RQ2: What is the nature of adult (tutor) prompting use during IBR? To evaluate our questions pertaining to characterizing tutors’ prompting types, we aggregated tutor–child dyad prompting use data to the tutor level (n = 9) and tested change in prompting from early- to end-intervention sessions using one-sample t-tests and tested the relationship among prompting type rates using Pearson’s zero-order correlation.

RQ3: What are the linkages among adult (tutor) prompting use on children’s pretest–posttest changes? Due to the nesting of children (Level 1, n = 41) within tutors (Level 2, n = 9) and our focus on testing Level 2 predictors (i.e., each of the types of prompting) on child-level outcomes (language and literacy pretest–posttest gains), we modeled pretest–posttest gains using multilevel modeling (e.g., Snijders & Bosker, 2012). We specifically tested each of the prompting types separately, controlling for pretest, as follows:

\begin{matrix} P r e - P o s t G a i n_{i j} = γ_{00} + γ_{10} \times Z P r e t e s t_{i j} + \\ γ_{01} \times Z R a t e P r o m p t T y p e_{j} + U_{0 j} + r_{i j} . \end{matrix}

In the model above, the pretest–posttest gain for the ith child in the jth tutor on a given measure was estimated as a function of the conditional grand mean gain across all children and tutors (γ₀₀), plus the effect of pretest (γ₁₀: predicted change in gain for each standard deviation increase in pretest) and prompting use rate (γ₀₁: predicted change in gain for each standard deviation increase in prompting use rate), and the random effects, including the deviation between the child’s tutor’s mean gain and the predicted grand mean gain (U_0j) and the residual error between the child’s gain and their tutor’s gain (r_ij). Models were implemented in R lme4 (Bates et al., 2015) with maximum likelihood and Satterthwaite df using R lmerTest (Kuznetsova et al., 2017).

Results

RQ1 and RQ2: Did Frequency and Type of Tutor Prompting Change From Beginning to End of the IBR Intervention? How Did Rates of Tutor Usage of Different Types of Prompts Correlate With Each Other?

Our first two RQs were concerned with the nature and frequency of adult prompting during IBR with children with ASD. Recall that tutors were not told which specific CROWD prompts to use with children during sessions, and as such, each prompt type (within the CROWD framework; recall that the acronym stands for completion, recall, open-ended, wh-, and distancing prompt types) was naturally occurring. A series of one-group t tests on tutor change between beginning- and end-session prompting rates (in PPM) showed no significant differences for any category (all ps > .05), although open-ended prompting showed a trend for an increase (p = .087). Given that we found no substantive differences between sessions, we computed the mean prompting rates across sessions (see Table 2) and tabulated the correlations among mean prompting rates (see Table 3). As can be seen, most prompting rates were positively related (i.e., tutors who spent more time prompting in one way tended to spend more time prompting in other ways too). That said, it appeared that wh- prompting was the least related to the other types of prompting use, which is likely related to the fact that wh- prompting was the most common type of adult talk observed during IBR.

Table 2.

Descriptive Statistics for Tutor Prompting Rates by CROWD Prompt Type across Sessions.

Prompt type	Beg session	End session	Mean
Prompt type	M (SD)	M (SD)	M (SD)
Completion PPM	0.62 (0.43)	0.60 (0.61)	0.61 (0.48)
Recall PPM	0.16 (0.11)	0.09 (0.07)	0.13 (0.07)
Open-ended PPM	0.09 (0.10)	0.16 (0.14)	0.12 (0.11)
Wh- PPM	1.51 (0.71)	1.18 (0.50)	1.34 (0.43)
Distancing PPM	0.20 (0.10)	0.20 (0.13)	0.20 (0.09)

Note. N = 9 tutors. CROWD = Completion, Recall, Open-ended, Wh, and Distancing; PPM = prompts per minute, aggregated for each tutor.

Table 3.

Correlations Among Tutor Prompting Rates by CROWD Prompt Type.

Prompt type	1.	2.	3.	4.	5.
1. Completion PPM	—
2. Recall PPM	.14	—
3. Open-ended PPM	.73	−.08	—
4. Wh-PPM	.31	.01	−.28	—
5. Distancing PPM	.38	.55	.35	−.09	—

Note. N = 9 tutors. CROWD = Completion, Recall, Open-ended, Wh, and Distancing; PPM = prompts per minute, aggregated for each tutor, across beginning and end of intervention sessions.

RQ2: To What Extent Did Tutor Prompting During IBR Predict Pretest–Posttest Change in Language and Literacy Skills?

Recall that we first tested the linkage among each of the prompting rates separately, controlling for pretest, on children’s language and literacy gains (we did this to understand the direct effects of the prompt type rates because they are naturally correlated with each other and could overlap to a great extent when combined into one model together). As can be seen in Table 4, tutor use of open-ended prompting was positively related to children’s gains in print knowledge (each standard deviation increase in prompting predicted a 6.16-point increased gain from pretest to posttest, controlling for pretest), and tutor use of completion prompting was positively related to children’s gains in PA (each standard deviation increase in prompting predicted a 4.52-point increased gain from pretest to posttest, controlling for pretest). None of the other prompts was significantly predictive of gains.

Table 4.

Multilevel Model Results Predicting Child Language and Literacy Gains using Each Type of CROWD Tutor Prompting (Separately).

Tutor prompt rate	Child oral language pre–post gains									Child early literacy pre–post gains
	Recept Vocab			Express Vocab			Listen Compreh			Print Know			Phono Aware
	Coefficient	SE	p	Coefficient	SE	p	Coefficient	SE	p	Coefficient	SE	p	Coefficient	SE	p
Completion PPM (Z)	3.28	1.67	.057	−1.29	1.59	.420	2.57	1.57	.109	4.14	1.87	.075	6.16	1.83	.002
Recall PPM (Z)	0.79	1.73	.651	−0.08	1.58	.961	0.09	1.62	.954	0.06	2.35	.979	2.17	2.33	.375
Open-ended PPM (Z)	1.10	1.98	.583	0.19	1.80	.917	2.51	1.81	.173	4.52	1.94	.046	4.74	2.29	.056
Wh- PPM (Z)	2.72	1.62	.100	−1.84	1.56	.244	2.06	1.60	.204	1.06	2.40	.674	3.91	2.02	.098
Distancing PPM (Z)	0.01	2.45	.417	−1.00	2.28	.662	3.88	2.23	.089	5.21	2.96	.126	5.34	3.11	.119

Note. N = 41 children and 9 tutors. CROWD = Completion, Recall, Open-ended, Wh, and Distancing; PPM = prompts per minute. Two-level models (children within tutors) estimated in R lme4 package using full maximum likelihood estimation and Satterthwaite-adjusted degrees of freedom. Recept Vocab = Peabody Picture Vocabulary Test–IV standard score; Express Vocab = Expressive One-Word Picture Vocabulary Test–IV standard score; Listen Compreh = Woodcock-Johnson III Listening Comprehension cluster standard score; Print Know = Test of Preschool Early Literacy Print Knowledge standard score; Phono Aware = Test of Preschool Early Literacy Phonological Awareness standard score.

Boldfaced values are significant at the .05 alpha level.

Importantly, because the IBR intervention in the original study showed significant advantages over typical preschool literacy instruction on gains in language skills but not early literacy skills, we followed up these findings to test whether the prompting effects on literacy gains observed in Table 4 would still be present after controlling for pretest–posttest language gains (in addition to pretest). In other words, do the two effects observed for completion and open-ended prompting on literacy have a unique benefit on literacy skills gains that are unrelated to language skills gains that were presumably a result of the IBR intervention?

To test this, we computed a mean composite language gain (averaging across the three language measures, then standardized as a Z-score) and added it to the previous model. As can be seen in Table 5, the prompting effects persisted even after controlling for language gains. For tutors who had relatively high open-ended prompt use, there was a 4.40-point predicted advantage over average amounts of open-ended prompt use on print knowledge gains, holding all else constant. Similarly, tutors who had relatively high completion prompt use were predicted to have children that gained 4.98 points more from pretest to posttest on PA compared with typical completion prompt use, all else being equal.

Table 5.

Multilevel Model Results Predicting Child Early Literacy Gains using Focal CROWD Prompting Types, Adjusted for Language Gains.

Fixed effect	Print Know				Phono Aware
Fixed effect	Coefficient	SE	p	d	Coefficient	SE	p	d
Intercept (mean gain)	3.78	1.83	.079		11.30	1.59	<.001
Child pretest (Z)	−8.63	1.26	<.001	−1.07	−4.87	1.63	.005	−0.47
Child Pre–Post Lang gain (Z)	0.71	1.77	.691	0.06	8.45	2.29	.001	0.58
Tutor open-ended PPM (Z)	4.40	1.91	.048	0.36	—	—	—	—
Tutor completion PPM (Z)	—	—	—	—	4.98	1.62	.004	0.48

Note. N = 41 children and 9 tutors. CROWD = Completion, Recall, Open-ended, Wh, and Distancing; PPM = prompts per minute. Two-level models (children within tutors) estimated in R lme4 package using full maximum likelihood estimation and Satterthwaite-adjusted degrees of freedom. Print Know = Test of Preschool Early Literacy Print Knowledge standard score; Phono Aware = Test of Preschool Early Literacy Phonological Awareness standard score. Pre–Post Lang Gain = pretest–posttest change on the Peabody Picture Vocabulary Test–IV standard score. Effect size computed as an approximate Cohen’s d value = coefficient/(standard error × √N), where N = 41.

Boldfaced values are significant at the .05 alpha level.

Discussion

This study investigated (a) the extent to which different kinds of adult prompting took place during an IBR intervention and how prompting changed over the course of the intervention period, and (b) how prompting types were linked with changes in language and literacy skills for preschool children with ASD during the IBR intervention. Upon examination of the latter results, we also evaluated whether the prompting effects observed on children’s literacy gains were still significant after controlling for language gains. Our key findings include that tutors were fairly consistent in their use of different CROWD prompt types over the course of the intervention (recall that CROWD prompt types), with most reliance on wh- prompting, and also that variation in tutor’s use of completion and open-ended prompts showed a significant benefit on children’s emergent literacy gains, above and beyond language gains made during the intervention. Both of these findings uniquely contribute to literature around IBR intervention implementation in general, and for children with ASD in particular. This latter finding in particular is consistent with previous research on language development for children with ASD (e.g., Perryman et al., 2013).

Adult Prompting and Emergent Literacy Skills

Our results are consistent with previous research with TD children that has shown adult prompting during shared book reading is linked with targeted emergent literacy skills (Justice & Ezell, 2002; Lonigan et al., 2013; Zucker et al., 2012). In our study with children with ASD, two of the five IBR intervention CROWD prompts were significantly predictive of emergent literacy skills: specifically, completion prompts were related to positive PA skills gains, and open-ended prompts were positively related to alphabet knowledge gains. In other words, although wh- prompts were most frequently used by tutors, completion and open-ended prompts were most predictive of better literacy outcomes.

Our findings also fit into the larger story emphasizing the importance of the development of language during early childhood. Children need to develop a large enough language lexicon before they can develop sensitivity to the smaller sound units of words. The Lexical Restructuring Hypothesis (Metsala, 1997a, 1997b; Walley, 1993) posits that, before a child can detect units of speech at the phoneme level, their lexicon must reach a certain threshold. In addition, existing studies challenge the notion that differential language and emergent literacy skill development can be attributed to ASD symptomatology. For example, Lucas and Norbury (2015) found that language impairment was a better predictor of text inferencing skill than ASD symptomatology for a sample of 10-year-olds with ASD. Taken together, language development is important for literacy development, irrespective of ASD status.

Interactive shared book reading is of course an ideal way to increase a child’s exposure to oral language and receptive vocabulary. Flynn (2011) even recommends implementing IBR in a tiered approach, where tutors present the same text over multiple sessions presenting questions with increasing demand as children indicated mastery of the vocabulary and concepts in the target book. Although the nature of the IBR intervention itself (one-on-one with a tutor who provides predictable routines and repeated readings of the same storybook) could explain the significant growth we observed in children’s PA skills, the fact that completion prompts were uniquely related to PA growth in the present study could be because completion prompts initiate a “episode” of joint attention, resulting in attending to a single word.

Regarding our finding that open-ended prompts were uniquely related to print knowledge gains, it may be the case that, given no specific direction within the shared book reading task, children with ASD may default to their strengths—which include discrete, concrete capacities such as orthographic- and code-related skills (Dynia et al., 2014; Lanter et al., 2012). Furthermore, as evidenced in recent studies with similar participant groups (Dynia et al., 2019), PA skills are related to oral language skills, and broadly speaking, the IBR intervention in the original study had a positive effect on oral language growth (Hudson et al., 2019), which in turn could potentially predict PA growth.

Limitations and Future Directions

There are several limitations to this study. Due to the small tutor and child sample sizes, the effects we found should be regarded as preliminary and future research should investigate whether the results can be replicated with children with ASD as well as other populations. Second, the children included in the study had a baseline level of language at the start of intervention. This was an inclusion criterion made in the original study based on the logic that children need a minimum level of language to benefit from a reading intervention. We note this as a limitation, however, because many children with ASD present with lower or little language in early childhood, and further investigation would be needed to generalize our results to that population. Third, the present study utilized extant data from a previous study in which tutors were free to choose what type of prompt to use and when to use them—in other words, we have a convenience sample of prompt use; without randomly assigning tutors to prompt use types or use frequencies, we cannot draw causal conclusions around the linkage between rates of different prompt usage and child outcomes. Future research might take up this question using an experimental design. It is also highly possible that further refinement of the IBR intervention to promote greater child-specific prompt use and greater engagement (recall that the average on-task engagement in the present study was 86%) could potentially produce greater effects on emergent literacy outcomes.

Conclusion

Consistent with previous work with TD children (Whitehurst et al., 1988; Whitehurst & Lonigan, 1998), the current study shows that naturally occurring adult prompt use during shared book reading is positively linked with emergent literacy growth in preschoolers with ASD. More specifically, greater use of completion and open-ended prompts was found to predict greater literacy skills growth, even after controlling for language growth. We, therefore, propose that selection and timing of prompting types might be more carefully considered in connection with language-focused shared book reading interventions such as Flynn’s (2011) approach, which involves having tutors repeat readings from familiar texts while continually assessing children’s mastery of vocabulary. Using our findings and the work of others, future research can and should take up the important work of investigating how we can further strengthen IBR interventions as a universal tool in combating risk for reading failure for all children.

Footnotes

Acknowledgements

We are grateful to the intervention teachers and data collectors for their work on this project. We especially thank the children, parents, teachers, and principals in Bellevue Public Schools, Edmonds Public Schools, Issaquah Public Schools, Lake Washington Public Schools, Renton Public Schools, Seattle Public Schools, Shoreline Public Schools, and the Haring Center who made this research possible.

Associate Editor: Emily Solari

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

Research reported in this publication was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health under award number 1R01HD072143-01. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). https://doi.org/10.1176/appi.books.9780890425596

Bates

Mächler

Bolker

Walker

(2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1)1–48. https://doi.org/10.18637/jss.v067.i01

Bishop

D. V.

Adams

(1990). A prospective study of the relationship between specific language impairment, phonological disorders and reading retardation. Journal of Child Psychology and Psychiatry, 31(7), 1027–1050. https://doi.org/10.1111/j.1469-7610.1990.tb00844.x

Boudreau

D. M.

Hedberg

N. L.

(1999). A comparison of early literacy skills in children with specific language impairment and their typically developing peers. American Journal of Speech-Language Pathology, 8(3), 249–260. https://doi.org/10.1044/1058-0360.0803.249

Cain

Oakhill

(Eds.). (2008). Children’s comprehension problems in oral and written language: A cognitive perspective. Guilford Press.

Chow

B. W. Y.

McBride-Chang

(2003). Promoting language and literacy development through parent–child reading in Hong Kong preschoolers. Early Education and Development, 14, 233–248. https://doi.org/10.1207/s15566935eed1402_6

Dunn

L. M.

Dunn

L. M.

(1997). Peabody picture vocabulary test (3rd ed.). American Guidance Service.

Dynia

J. M.

Bean

Justice

L. M.

Kaderavek

J. N.

(2019). Phonological awareness emergence in preschool children with autism spectrum disorder. Autism & Developmental Language Impairments. Advance online publication. https://doi.org/10.1177/2396941518822453

Dynia

J. M.

Lawton

Logan

J. A. R.

Justice

L. M.

(2014). Comparing emergent-literacy skills and home-literacy environment of children with Autism and their peers. Topics in Early Childhood Special Education, 34(3), 142–153. https://doi.org/10.1177/0271121414536784

10.

Fleury

V. P.

Miramontez

S. H.

Hudson

R. F.

Schwartz

I. S.

(2014). Promoting active participation in book reading for preschoolers with Autism Spectrum Disorder: A preliminary study. Child Language Teaching and Therapy, 30, 273–288. https://doi.org/10.1177/0265659013514069

11.

Fleury

V. P.

Schwartz

I. S.

(2017). A modified dialogic reading intervention for preschool children with autism spectrum disorder. Topics in Early Childhood Special Education, 37(1), 16–28. https://doi.org/10.1177/0265659013514069

12.

Flynn

K. S.

(2011). Developing children’s oral language skills through dialogic reading: Guidelines for implementation. Teaching Exceptional Children, 44(2), 8–16. https://doi.org/10.1177/004005991104400201

13.

Han

Neuharth-Pritchett

(2015). Meaning-related and print-related interactions between preschoolers and parents during shared book reading and their associations with emergent literacy skills. Journal of Research in Childhood Education, 29(4), 528–550. https://doi.org/10.1080/02568543.2015.1073819

14.

Hart

Ehde

D. M.

(2015). Defining the treatment targets and active ingredients of rehabilitation: Implications for rehabilitation psychology. Rehabilitation Psychology, 60(2), 126–135. https://doi.org/10.1037/rep0000031

15.

Hindman

A. H.

Connor

C. M.

Jewkes

A. M.

Morrison

F. J.

(2008). Untangling the effects of shared book reading: Multiple factors and their associations with preschool literacy outcomes. Early Childhood Research Quarterly, 23(3), 330–350. https://doi.org/10.1016/j.ecresq.2008.01.005

16.

Hudson

R. F.

Gasamis

Schwartz

I. S.

(2019, February). Effects of storybook reading and phonological awareness intervention for preschoolers with ASD [Paper presentation]. Council for Exceptional Children Annual Conference, St. Louis, MO, United States.

17.

Hudson

R. F.

Sanders

E. A.

Greenway

Xie

Smith

Gasamis

Martini

Schwartz

Hackett

(2017). Effects of emergent literacy interventions for preschoolers with autism spectrum disorder. Exceptional Children, 84(1), 55–75. https://doi.org/10.1177/0014402917705855

18.

Justice

L. M.

Ezell

H. K.

(2002). Use of storybook reading to increase print awareness in at-risk children. American Journal of Speech-Language Pathology, 11(1), 17–29. https://doi.org/10.1044/1058-0360(2002/003)

19.

Kang

J. Y.

Kim

Y. S.

Pan

B. A.

(2009). Five-year-olds’ book talk and story retelling: Contributions of mother—child joint book reading. First Language, 29(3), 243–265. https://doi.org/10.1177/0142723708101680

20.

Koppenhaver

D. A.

Erickson

K. A.

Skotko

B. G.

(2001). Supporting communication of girls with Rett syndrome and their mothers in storybook reading. International Journal of Disability, Development and Education, 48(4), 395–410. https://doi.org/10.1080/10349120120094284

21.

Kuznetsova

Brockhoff

P. B.

Christensen

R. H. B.

(2017). LmerTest package: Tests in linear mixed effects models. Journal of Statistical Software, 82(13), 1–26. https://doi.org/10.18637/jss.v082.i13

22.

Lanter

Freeman

Dove

(2012). Procedural and conceptual print-related achievements in young children with Autism Spectrum Disorders. Focus on Autism and Other Developmental Disabilities, 28(1), 14–25. https://doi.org/10.1177/1088357612459270

23.

Light

Binger

Smith

A. K.

(1994). Story reading interactions between preschoolers who use AAC and their mothers. Augmentative and Alternative Communication, 10(4), 255–268. https://doi.org/10.1080/07434619412331276960

24.

Light

Smith

A. K.

(1993). Home literacy experiences of preschoolers who use AAC systems and of their nondisabled peers. Augmentative and Alternative Communication, 9(1), 10–25. https://doi.org/10.1080/07434619312331276371

25.

Lonigan

C. J.

Anthony

J. L.

Bloomfield

B. G.

Dyer

S. M.

Samwel

C. S.

(1999). Effects of two shared-reading interventions on emergent literacy skills of at-risk preschoolers. Journal of Early Intervention, 22, 306–322. https://doi.org/10.1177/105381519902200406

26.

Lonigan

C. J.

Purpura

D. J.

Wilson

S. B.

Walker

P. M.

Clancy-Menchetti

(2013). Evaluating the components of an emergent literacy intervention for preschool children at risk for reading difficulties. Journal of Experimental Child Psychology, 114(1), 111–130. https://doi.org/10.1016/j.jecp.2012.08.010

27.

Lonigan

C. J.

Wagner

R. K.

Torgesen

J. K.

(2007). Test of preschool early literacy. Pro-Ed.

28.

Lucas

Norbury

C. F.

(2015). Making inferences from text: It’s vocabulary that matters. Journal of Speech, Language, and Hearing Research, 58(4), 1224–1232. https://doi.org/10.1044/2015_JSLHR-L-14-0330

29.

Luo

Tamis-LeMonda

C. S.

(2017). Preschool book-sharing and oral storytelling experiences in ethnically diverse, low-income families. Early Child Development and Care, 189, 1602–1619. https://doi.org/10.1080/03004430.2017.1400542

30.

Martin

Brownell

(2011). Expressive one-word vocabulary test. Academic Therapy.

31.

Marvin

Mirenda

(1993). Home literacy experiences of preschoolers enrolled in Head Start and special education programs. Journal of Early Intervention, 17(4), 351–367. https://doi.org/10.1177/105381519301700402

32.

McGinty

A. S.

Justice

L. M.

Zucker

T. A.

Gosse

Skibbe

L. E.

(2012). Shared-reading dynamics: Mothers’ question use and the verbal participation of children with specific language impairment. Journal of Speech, Language, and Hearing Research, 55(4), 1039–1052. https://doi.org/10.1044/1092-4388(2011/10-0298)

33.

Metsala

J. L.

(1997a). An examination of word frequency and neighborhood density in the development of spoken-word recognition. Memory & Cognition, 25(1), 47–56. https://doi.org/10.3758/bf03197284

34.

Metsala

J. L.

(1997b). Spoken word recognition in reading disabled children. Journal of. Educational Psychology, 89(1), 159–169. https://doi.org/10.1037/0022-0663.89.1.159

35.

Mucchetti

C. A.

(2013). Adapted shared reading at school for minimally verbal students with autism. Autism, 17(3), 358–372. https://doi.org/10.1177/1362361312470495

36.

Perryman

T. Y.

Carter

A. S.

Messinger

D. S.

Stone

W. L.

Ivanescu

A. E.

Yoder

P. J.

(2013). Brief report: Parental child-directed speech as a predictor of receptive language in children with autism symptomatology. Journal of Autism and Developmental Disorders, 43(8), 1983–1987. https://doi.org/10.1007/s10803-012-1725-3

37.

Rahn

N. L.

Coogle

C. G.

Storie

(2016). Preschool children’s use of thematic vocabulary during dialogic reading and activity-based intervention. The Journal of Special Education, 50(2), 98–108. https://doi.org/10.1177/0022466915622202

38.

Rogoff

(1993). Children’s guided participation and participatory appropriation in sociocultural activity. In Wozniak

R. H.

Fischer

K. W.

(Eds.), Development in context: Acting and thinking in specific environments (pp. 121–153). Lawrence Erlbaum. https://doi.org/10.1017/cbo9781139174299.008

39.

Rosenquest

B. B.

(2002). Literacy-based planning and pedagogy that supports toddler language development. Early Childhood Education Journal, 29(4), 241–249.

40.

Scarborough

H. S.

(1989). Prediction of reading disability from familial and individual differences. Journal of Educational Psychology, 81(1), 101–108. https://doi.org/10.1037/0022-0663.81.1.101

41.

Scarborough

H. S.

Dobrich

(1994). On the efficacy of reading to preschoolers. Developmental Review, 14(3), 245–302. https://doi.org/10.1006/drev.1994.1010

42.

Shanahan

Lonigan

C. J.

(2010). The National Early Literacy Panel: A summary of the process and the report. Educational Researcher, 39(4), 279–285. https://doi.org/10.3102/0013189x10369172

43.

Siller

Sigman

(2002). The behaviors of parents of children with autism predict the subsequent development of their children’s communication. Journal of Autism and Developmental Disorders, 32, 77–89. https://doi.org/10.1037/t15140-000

44.

Siller

Sigman

(2008). Modeling longitudinal change in the language abilities of children with autism: Parent behaviors and child characteristics as predictors of change. Developmental Psychology, 44, 1691–1704. https://doi.org/10.1037/a0013771

45.

Snijders

T. A. B.

Bosker

R. J.

(2012). Multilevel analysis: An introduction to basic and advanced multilevel modeling (2nd ed.). SAGE.

46.

Storch

S. A.

Whitehurst

G. J.

(2002). Oral language and code-related precursors to reading: Evidence from a longitudinal structural model. Developmental Psychology, 38(6), 934–947. https://doi.org/10.1037/0012-1649.38.6.934

47.

Tager-Flusberg

Joseph

Folstein

(2001). Current directions in research on autism. Developmental Disabilities Research Reviews, 7(1), 21–29. https://doi.org/10.1002/1098-2779(200102)7:1

48.

Tipton

L. A.

Blacher

J. B.

Eisenhower

A. S.

(2017). Young children with ASD: Parent strategies for interaction during adapted book reading activity. Remedial and Special Education, 38(3), 171–180. https://doi.org/10.1177/0741932516677831

49.

Toth

Munson

Meltzoff

A. N.

Dawson

(2006). Early predictors of communication development in young children with autism spectrum disorder: Joint attention, imitation, and toy play. Journal of Autism and Developmental Disorders, 36(8), 993–1005. https://doi.org/10.1007/s10803-006-0137-7

50.

Towson

J. A.

Gallagher

P. A.

Bingham

G. E.

(2016). Dialogic reading: Language and preliteracy outcomes for young children with disabilities. Journal of Early Intervention, 38(4), 230–246. https://doi.org/10.1177/1053815116668643

51.

Tsybina

Eriks-Brophy

(2010). Bilingual dialogic book-reading intervention for preschoolers with slow expressive vocabulary development. Journal of Communication Disorders, 43(6), 538–556. https://doi.org/10.1016/j.jcomdis.2010.05.006

52.

Walley

A. C.

(1993). The role of vocabulary development in children’s spoken word recognition and segmentation ability. Developmental Review, 13(3), 286–350. https://doi.org/10.1006/drev.1993.1015

53.

Warren

S. F.

Yoder

P. J.

(1996). Enhancing communication and language development in young children with developmental delays and disorders. Peabody Journal of Education, 71(4), 118–132. https://doi.org/10.1080/01619569609595132

54.

Whalon

Martinez

J. R.

Shannon

Butcher

Hanline

M. F.

(2015). The impact of reading to engage children with autism in language and learning (RECALL). Topics in Early Childhood Special Education, 35, 102–115. https://doi.org/10.1177/0271121414565515

55.

Whitehurst

G. J.

Arnold

D. H.

Epstein

J. N.

Angell

A. L.

Smith

Fischel

J. E.

(1994). A picture book reading intervention in daycare and home for children from low-income families. Developmental Psychology, 30, 679–689. https://doi.org/10.1037/0012-1649.30.5.679

56.

Whitehurst

G. J.

Falco

F. L.

Lonigan

C. J.

Fischel

J. E.

DeBaryshe

B. D.

Valdez-Menchaca

M. C.

Caulfield

(1988). Accelerating language development through picture book reading. Developmental Psychology, 24(4), 552–559. https://doi.org/10.1037/0012-1649.24.4.552

57.

Whitehurst

G. J.

Lonigan

C. J.

(1998). Child development and emergent literacy. Child Development, 69, 848–872. https://doi.org/10.1111/j.1467-8624.1998.tb06247

58.

Woodcock

R. W.

(1997). The Woodcock-Johnson tests of cognitive ability–Revised.

59.

Woodcock

R. W.

Schrank

F.A.

Mather

McGrew

K. S.

(2007). Woodcock-Johnson III tests of achievement. Riverside.

60.

Zucker

T. A.

Cabell

S. Q.

Justice

L. M.

Pentimonti

J. M.

Kaderavek

J. N.

(2012, October 15). The role of frequent, interactive prekindergarten shared reading in the longitudinal development of language and literacy skills. Developmental Psychology, 49(8), 1425–1439. https://doi.org/10.1037/a0030347