Instruction in Letter-Sound Correspondences for Children With Autism and Limited Speech

Abstract

This study used a multiple probe across participants’ research design to evaluate the effects of instruction on the acquisition of letter-sound correspondences (LSCs) by three young children with autism spectrum disorder and limited speech. All three children (ages 3–5 years) reached criterion for identifying the LSCs targeted during instruction, and a Nonoverlap of All Pairs (NAP) analysis showed evidence of medium to strong treatment effects. All three children also provided evidence of maintenance and generalization of LSC skills to other tasks. Implications for instruction and future research directions are discussed.

Keywords

autism spectrum disorder (ASD)disability populations literacy intervention strategies reading single-subject designs experimental studies research methodologies

Literacy skills are critical to successful participation in school and society, and often provide access to valued leisure and recreational activities (Lonigan & Shanahan, 2008). Literacy skills are especially important for children with autism spectrum disorder (ASD) who experience difficulty using speech to communicate; for these children, who have both autism and limited speech, literacy not only provides the traditional benefits but may also serve as a powerful method of communication with others. Once an individual can spell, that person is able to communicate a wide variety of concepts with a wide variety of communication partners (Koppenhaver, Evans, & Yoder, 1991; Light & McNaughton, 2014; Mirenda, 2003). Although the ability to read and write is especially critical for this population, it has been estimated that more than 90% of persons with severe speech limitations (e.g., individuals with cerebral palsy, Down syndrome, ASD) will not acquire functional literacy skills (Foley & Wolter, 2010). These low rates of literacy acquisition appear to be the result of both extrinsic (e.g., lack of appropriate instruction, low expectations) and intrinsic (e.g., physical and cognitive disabilities) factors (Light & McNaughton, 2009).

Recent research provides evidence that young children with ASD can acquire and make use of literacy skills (Travers et al., 2011; Whalon, Al Otaiba, & Delano, 2009); however, many of the children in these studies demonstrated spoken language skills that supported participation in a wide variety of literacy assessment and instructional activities. At present, there is only a limited understanding of how to support the acquisition of literacy skills for young children with ASD who experience severe difficulty in using speech to communicate (Dynia, Lawton, Logan, & Justice, 2014).

Literacy Instruction in Early Childhood Special Education (ECSE) Settings

Increased attention to the development of literacy skills has resulted in an examination of the early reading and writing experiences of young children with disabilities (Dynia et al., 2014). Emergent literacy activities during the preschool years often include meaningful and motivating language experiences (e.g., storybook reading, discussion of text in the environment) to assist young children in learning concepts about print and story structure (Lonigan & Shanahan, 2008). Although these early literacy experiences are important, researchers have found that formal instruction is necessary to support the acquisition of conventional skills that are critical to early reading achievement (Adams, 1990; Lonigan & Shanahan, 2008). While a wide variety of skills can and should be addressed in early reading instruction, providing explicit instruction in letter-sound correspondence (LSC) and phonemic awareness (i.e., learning that sounds can be represented by letters, and words are made up of sounds) has been identified as particularly important for younger learners considered at risk for reading difficulties (Lonigan & Shanahan, 2008).

Reading Instruction for Young Children With Autism and Limited Speech

Although a number of literacy instruction programs have been observed to be effective with children with ASD who have speech communication skills (Travers et al., 2011; Whalon et al., 2009), these programs require the use of speech as a response mode by the learner. It has been estimated that as many as 30% of individuals with ASD will not develop the ability to speak in even short phrases (Wodka, Mathy, & Kalb, 2013). These children, who have both ASD and limited speech, often experience difficulty participating in reading instruction, which requires the use of spoken responses (Coleman-Martin, Heller, Cihak, & Irvine, 2005; Light & McNaughton, 2009). Perhaps for this reason, many reading programs designed for learners with limited speech often focus on sight-word reading instruction alone (Browder, Wakeman, Spooner, Ahlgrim-Delzell, & Algozzine, 2006), and do not address the basic decoding skills (e.g., phonemic awareness, LSCs) that are critical to the development of the ability to read and spell a wide variety of words (Adams, 1990; Barker, Bridges, & Saunders, 2014).

More recently, a small number of researchers have investigated adapted instruction in early literacy skills for persons with ASD and limited speech, and the development of activities that do not require spoken responses and support participation via alternative response modes (e.g., pointing to letters and pictures, the use of an augmentative and alternative communication [AAC] device). For example, Bailey, Angell, and Stoner (2011) reported that a structured intervention package resulted in some improvement in early literacy skills for three adolescents with ASD and limited speech. Browder, Ahlgrim-Delzell, Flowers, and Baker (2012) described the positive impact of a multicomponent early literacy program for children with severe developmental disabilities in grades Kindergarten through Grade 4, some of whom had a diagnosis of ASD, and some of whom were described as having little or no speech.

Although there is evidence that literacy intervention programs have produced positive results for children and adolescents with limited speech (Barker, Saunders, & Brady, 2012; Machalicek et al., 2010), there is little information as to whether these interventions would provide an effective, efficient, and appropriate intervention for very young children with ASD and limited speech in a preschool setting. Despite the challenges involved in providing reading instruction to these learners, instruction that focuses on LSCs is critical not only for developing independent reading skills but also for providing a flexible and powerful means of communication for those with limited speech. Instruction in LSCs can serve as a starting point for the development of early spelling skills (Adams, 1990; Ball & Blachman, 1991); once learners acquire LSCs, they can then learn how to apply this knowledge to read and spell words, which enables individuals with limited speech to create novel messages rather than relying on preselected symbols or words (Beukelman & Mirenda, 2013; Koppenhaver, 2000; Light & McNaughton, 2009).

Purpose

As a first step in developing a more comprehensive approach to literacy instruction for preschoolers with ASD who have limited speech, the current study focused on teaching LSCs to this population. The instructional activities were based on procedures demonstrated to be effective and efficient in teaching early reading skills to persons with limited speech (Fallon, Light, McNaughton, Drager, & Hammer, 2004; Light, McNaughton, Weyer, & Karg, 2008) with special adaptations to provide appropriate instruction to very young children with ASD and limited speech. This study investigated the (a) effectiveness, (b) efficiency, and (c) social validity of the identified literacy activities on the acquisition and maintenance of LSC skills by young children with ASD and limited speech.

Method

Participants

Participants were recruited from rural communities within a northeastern state, and met the following selection criteria: (a) were between the ages of 3,0 and 5,11, (b) had a diagnosis of autism provided by a licensed psychologist, (c) did not demonstrate speech skills adequate to meet their daily communication needs as reported by the teacher, (d) demonstrated the ability to follow familiar one-step directions, (e) demonstrated an interest in books and/or letters, (f) demonstrated ability to select target pictures (e.g., photos, line drawings) from a field of at least four choices in response to a spoken request, (g) lived in homes in which English was the first language, and (h) demonstrated vision and hearing skills within normal limits. As a final selection criteria, candidates participated in a Screening Task (which followed the same procedures as the LSC probe, described in detail in the section on “Measures”). In brief, the researcher presented a letter sound orally (e.g., /m/) and asked the participants to select the target letter from a field of six choices. Candidates who identified fewer than 10 letter sounds were eligible for participation in the study.

The three participants who met all of the selection criteria were White non-Hispanic, from middle-class families, and from rural areas in central Pennsylvania. They all attended a preschool program that specialized in providing ECSE services to children with an ASD or a developmental delay.

Cameron

Cameron was 3 years, 6 months old at the beginning of this study. He received a diagnosis of ASD when he was 2 years, 4 months old. When provided with an opportunity to select a preferred activity, Cameron chose listening to music, playing games on the computer, coloring pictures, and looking at books that included pictures of animals.

Cameron’s teacher reported that he could follow familiar, oral one-step directions (e.g., “sit down,” “stand up,” “put in,” “get one”). On request he would point to photos of family members (e.g., mom, dad, grandma) and some animals (e.g., dog, cat, pig), and to some body parts (e.g., nose, eyes, toes). Cameron made use of gestures (e.g., waving his hand to indicate “no” to a particular item or activity), a small number of sign approximations for some words (e.g., “help,” “more,” “all done”), and Picture Communication Symbol line drawings (Mayer-Johnson, 1992) to communicate. Cameron also used a speech generating AAC device to request preferred items and to participate in learning activities when prompted by an adult. As part of the services provided at the ECSE program, Cameron had been provided with an Auggie (see Note 1) programmed with approximately 250 vocabulary items. Cameron primarily used his AAC device during group activities (e.g., circle time) and for speech therapy sessions.

Cameron’s teacher reported that he never initiated interaction with others at school and that he mostly communicated to express his needs and wants and to respond to questions when prompted by an adult. In addition, Cameron’s teacher reported that he did not typically communicate to share information, express social closeness, or participate in social etiquette routines (e.g., communicate a greeting). In a typical interaction involving the use of his AAC device, Cameron would respond to a question from an adult (e.g., “What do you want?”) by pointing to a picture in his system. Cameron’s teacher reported that he did not use natural speech to communicate, although he did vocalize when he was excited or frustrated. Cameron typically obtained desired items independently, and he demonstrated frustration (e.g., loud vocalizations and/or crying) when communication breakdowns occurred.

Zach

Zach was 5 years, 6 months old when the study began. He received a diagnosis of ASD when he was 3 years, 2 months old. When provided with an opportunity to select a preferred activity, Zach would often choose to look at books, color, or play with puzzles or toy cars.

Zach’s teacher reported that he could follow familiar one-step and two-step directions and that he demonstrated understanding of his own name, his family members’ names (e.g., mom, dad, brother, sister), names of people in his classroom, and names of animals, body parts, and familiar objects (e.g., chair, table, scissors, book, crayons, etc.). His teacher also reported that he could respond appropriately within some conversational routines (e.g., he would say “hi” with prompting after someone initiated a greeting). In addition, Zach could respond to simple wh- questions (e.g., “What do you want?”) by pointing to an appropriate line drawing. Although Zach did not have a personalized AAC system at the time of the study, the teacher provided the students in the classroom with communication boards, which included 2 to 10 vocabulary items represented by Picture Communication Symbol line drawings presented within a grid format, to support their participation during specific activities such as circle time and snack time.

Zach’s natural speech consisted of one- to two-word phrases (e.g., “hi,” “look,” and “that one”) that were often repetitions of what people had said, but sometimes seemed to be used with communicative intent. His teacher indicated that Zach occasionally used his natural speech during familiar tasks and routines but that his speech was limited and often difficult to understand.

According to his teacher, Zach rarely initiated interaction at school, and he mostly communicated to express needs and wants and to respond to questions. Zach did not typically communicate to express social closeness or to participate in social etiquette routines. Zach’s teacher reported that his natural speech was often difficult for all communication partners to understand but that he did not demonstrate frustration when others did not understand him.

Cole

Cole was 4 years and 11 months old at the beginning of this study. He received a diagnosis of ASD when he was 1 year, 10 months old. When given an opportunity to select an activity, Cole would often choose to look at books or play with blocks.

Cole’s teacher reported that he could follow familiar one-step and two-step directions and that he demonstrated understanding of his own name, as well as names of family members, people in his classroom, and of animals, body parts, and familiar objects. His teacher also noted that he could participate appropriately within some conversational routines with support (e.g., he would say “hi,” “bye,” “yes,” and “no” when prompted to do so).

The methods of communication most commonly used by Cole at school included gestures (e.g., waving his hand to indicate “no” to a particular item or activity), line drawings, and some natural speech, but his speech was very limited and almost never understood by others. Cole did not have a personalized AAC system; however, he used the communication boards his teacher provided to support his participation during specific activities. These communication boards included 2 to 10 vocabulary items represented by Picture Communication Symbol line drawings presented within a grid format.

It was reported that at school Cole mostly communicated to express his needs and wants and to respond to questions. In addition, Cole’s teacher noted that he would often smile to express social closeness with others and would say “hi,” “bye,” and “thank you” when prompted by an adult. Cole’s teacher reported that he rarely initiated interaction and that he demonstrated frustration during communication breakdowns (e.g., loud vocalizations and/or crying, stomping feet, waving arms).

Setting

The study took place in two preschool classrooms that specialized in ECSE services for children with a diagnosis of ASD between 3 and 6 years of age. Each classroom used a schedule, which divided the day into activities at short learning centers (e.g., speech therapy, fine motor skills, water play); the instructional sessions for this study were scheduled as a learning center activity. Although the children participated in some early literacy activities as part of the daily routine in their classroom (e.g., listening to a story read aloud at circle time, completing puzzles that included letters of the alphabet), none of the children were participating in individualized literacy instruction activities. For this study, the researcher who provided all assessment and instructional activities was a certified special education teacher who was experienced in providing literacy instruction to children with learning disabilities.

Materials

Early instructional sessions included both instructional activities and an extension activity and were followed by a choice activity. The instructional materials used during the instruction and extension activities were based on the LSCs targeted during instruction. Choice activities were traditionally used in the classrooms to “reward” children for their participation at the end of an instructional session and were based on the preferences identified for the child by the classroom teacher.

Six LSCs (o, t, r, l, u, p) were targeted during the intervention and were represented as lowercase letters in 83-point Arial font on 2 × 2-in. cards. The researcher selected these six LSCs from a pool of letters that were identified as unknown to all three participants based on the LSC Screening Assessment. The decision to select these six letters from the pool of unknown letters was based on the sequencing guidelines recommended by Carnine, Silbert, Kame’enui, and Tarver (2010) and Light and McNaughton (2009), which include (a) teach letters that are more frequently used letters (e.g., a, m, t, s, etc.) before less frequently used letters (e.g., q, x, y, z), (b) separate letters that are visually similar (e.g., b and d), and (c) separate letters that are auditorily similar (e.g., g and k).

Instruction

Instructional materials included letter cards, a box for depositing the cards as they were used during instruction (decorated to look like a popular children’s cartoon character, Elmo), and a place mat. When introducing a new LSC, the researcher also presented 3 × 3-in. picture cards for words that began with the target letter sound (e.g., rug, run, and red for the letter r).

Extension activities

Extension activity materials were based on activities that were familiar to and preferred by the children (as identified by the classroom teacher), including (a) a letter-sound book (color photographs or line drawings that represented words that began with the target letter sound), (b) an I-Spy task (a page that presented a collage of items, which included letters “hidden” within the collage), and (c) a color-and-paste task in which participants first selected the target letter that corresponded to the first sound in the target word (e.g., p for penguin) and then completed pasting and coloring activities.

Choice activities

Choice activity materials included puzzles, books, and stickers.

Dependent Variable

The primary dependent variable in this study was the participants’ performance on an LSC probe, which tested both acquisition and maintenance of the six LSCs (o, t, r, l, u, p) targeted during the intervention. An LSC probe included two trials for each of the six targeted LSCs, which provided an opportunity for 12 correct responses. When assessing the participants’ performance, the researcher presented a letter sound orally (e.g., /l/) and asked the participants to select the target letter from a field of six choices (i.e., the six letters targeted during the intervention: o, t, r, l, u, p). For the response to be considered correct, the learner had to select the correct letter on the first attempt within 5 s.

Experimental Design

This study used a multiple-probe design across three participants (Horner & Baer, 1978; Kennedy, 2005). The independent variable was instruction in LSC; the dependent variable was the frequency of correct performance on the LSC probes. The participants remained in the baseline condition until the prior participant demonstrated a treatment effect (i.e., an increase of at least two correct responses from the highest point observed at baseline for two consecutive LSC probes), at which time the intervention was introduced to a new participant.

Measures

The researcher conducted an LSC probe before every second instructional session. On a day that included a probe session, the probe was always delivered as the first task on that day; thus, the probe session served as an assessment of what the participant had retained from the previous instructional session. Each LSC probe contained 2 trials for each of the six target letters (i.e., o, t, r, l, u, p), which equaled a total of 12 trials per probe. In addition, the response field of six letters always included the six target letters.

Procedures

The study included four phases: baseline, intervention, maintenance, and generalization.

Baseline

Each participant completed at least five LSC probes during baseline. Each probe targeted each of the six LSCs twice in random order. When introducing the LSC probe task, the researcher provided oral instructions and modeled the task to assist the participant in understanding the expectations. Next, the researcher said the letter sound orally and paused to provide the child with an opportunity to respond (up to 5 s). The researcher did not provide the participant with corrective feedback regarding the accuracy of the response but did encourage and thank the participant for participating (e.g., “Thank you for touching a letter!”).

Intervention

Every intervention session included an instructional session, and every second intervention session began with an LSC probe prior to instruction. At the beginning of the intervention, each intervention session lasted approximately 20 min, which included approximately 15 min of instructional activities (10 min of instruction, 5 min for the extension activity) and 5 min for choice activities. The choice activities were designed to reinforce the participant’s engagement during the sessions. As the children quickly demonstrated engagement and increased interest in the instruction and extension activities, the use of choice activities was not needed to promote engagement—It was provided when time permitted; however, it was omitted when the classroom schedule necessitated moving on to a new activity.

The researcher typically provided one instructional session per day, averaging five sessions per week. This study was conducted during the last two months of the school year; therefore, to accommodate the teachers’ end of the year schedules, as well as to ensure that all participants received an adequate number of instructional sessions before the end of the school year, the researcher provided two sessions on a small number of days.

Instructional activity

Instructional activities for teaching LSC included the following components: (a) introducing and modeling the LSC, (b) modeling the task, (c) providing guided practice, and/or (d) providing independent practice. The LSC targeted for instruction was either (a) a new LSC that was being introduced for the first time or (b) a recently introduced LSC for which the learner had previously received instruction, but for which the learner had not yet demonstrated criterion. The instructional components included in each session varied depending on the participant’s performance.

Introducing the LSC

To introduce a new LSC, the researcher presented the letter card, pointed to the letter (e.g., r), and stated its sound (/r/). Next, the researcher showed the participant picture cards of words that began with the target sound (e.g., run, rabbit, ring, etc.) and emphasized and extended the target sound as appropriate (e.g., “rrrrun,” “rrrrabbit,” “rrrring”).

Modeling the task

The instructor then modeled the task (i.e., the participant’s expected performance). When modeling the task, the researcher presented two letter cards to the participant, which included the target letter (e.g., r) and a foil (e.g., t). The researcher then described the first step of the activity by saying, “I am going to say a sound.” The researcher stated the target sound (e.g., /r/) and explained how to select the corresponding letter (e.g., “Now I look at the letters and point to the letter that makes the sound /r/”). The researcher then modeled the next steps in the task by (a) looking at each of the letters, (b) pointing to the target letter, and (c) stating the target sound (e.g., /r/). During each model, the researcher presented a different letter to serve as a foil (one of the five remaining nontarget letters) and varied the location of the target letter.

Providing guided practice

The researcher began guided practice by saying, “Let’s do one together.” The researcher then explained the task: “I am going to say a sound. Touch the letter that makes that sound.” The instructor then stated the sound and paused. The researcher provided an opportunity for the participant to respond by pausing for 3 to 5 s before prompting the participant to select the corresponding letter. If the participant did not initiate a response, the researcher repeated the verbal prompt (e.g., “Touch the letter that says, /r/) and modeled the correct response. After providing the correct response, the researcher prompted the participant, saying “touch the letter with me” and then prompted the participant to touch the target letter independently (e.g., “Now it’s your turn. Touch the letter that says, /r/”).

During instructional activities, the researcher provided instructional scaffolding to support correct methods of responding. Instructional scaffolding included starting with a smaller number of response options (i.e., only presenting two cards as options at first, and gradually building to the presentation of six response options as the participant demonstrated increased competency), pausing and looking expectantly for a response, reaching for (but not touching) the correct response, and/or lightly lifting the hand of the participant to indicate that a response was expected.

The researcher provided corrective feedback based on the participant’s response. When the participant provided an incorrect response, the researcher immediately provided corrective feedback for the participant, which included (a) modeling the correct response, (b) prompting the participant to select the correct response with the researcher, and (c) telling the participant to select the correct response independently. If the participant continued to demonstrate incorrect responses, the researcher increased the amount of support to assist the participant in responding correctly (e.g., decreasing the field size, providing more prompts, decreasing the length of the pause before modeling the correct response for the participant, etc.). This support was, again, faded as the participant demonstrated correct responses independently. This assisted the participant in maintaining high rates of success while also supporting the participant’s independence in completing the task.

Once the participant demonstrated at least two consecutive correct responses independently, the researcher gradually reduced scaffolding support by (a) increasing the length of the pause provided for the participant to independently respond (up to 5 s) and (b) increasing the field size of options by one additional letter card until the learner could select the target letter from a field of six letters.

Guided practice for a new LSC always began with massed trials of the new LSC—that is, the participant had multiple opportunities to practice selecting the most recently introduced LSC. As the participant demonstrated increasing accuracy with the recently introduced LSC, the researcher also presented trials with previously learned LSCs in random order. Once the participant selected the target LSC from a field of six choices for three consecutive trials including the target letter, with minimal support from the researcher, the participant moved into the independent practice phase of instruction.

Providing independent practice

Independent practice included 15 trials: 10 trials of the target (most recently introduced) LSC and 5 trials of any previously learned letter sounds (see Note 2). During independent practice, no prompts were provided after the researcher presented the directions (e.g., “Touch the letter that says /t/”), and response options were always a field of six letter cards.

Independent practice provided further review for both the newly learned and previously learned LSCs. The participant had to reach the criterion of at least 80% accuracy (8 correct trials out of 10 trials) for identifying the target LSC for two consecutive independent practice sessions for the LSC to be considered acquired. Once a participant demonstrated that the target LSC had been acquired, the researcher introduced the next LSC.

The information from the LSC probe was also used to make a decision about the provision of booster instruction. Booster instruction was to be presented when a participant missed both trials on a probe for an LSC that had previously been acquired and was to include a model followed by practice trials. However, no participant missed both trials on a Probe for an LSC that had previously been acquired, so no booster instruction was used in this study.

Extension activity

Instructional sessions ended with an activity that provided the participants with an opportunity to practice learned LSCs in the context of an age-appropriate preferred activity. Example extension activities included the use of a letter-sound book, an I-Spy task, or a color-and-paste task, using preferred topics (e.g., animals, cartoon characters) and emphasizing the targeted LSC. The extension activity was typically 5 min in length.

Maintenance

Maintenance data were collected after the intervention was completed. For Cameron and Cole, the researcher conducted maintenance probes on 3 consecutive days, approximately 4 weeks after the intervention ended. For Zach, maintenance data were collected approximately 5 weeks after the intervention ended; due to the fact that Zach had moved out of the area and his availability was limited, three maintenance probes were collected on a single day, with a 10-min break between each probe.

Generalization

During the last week of the intervention, the researcher presented two generalization tasks to assess whether the participants could make generalized use of their LSC knowledge to new materials and task formats: (a) a Generalization Book Probe, in which all six target letters could be viewed at one time on two pages (three letters presented vertically on the left page and three letters presented vertically on the right page) rather than as a series of six cards and (b) a Generalization Alphabet Probe, which presented the target letter in a field of six letters that included letters that were randomly selected from the entire alphabet rather than from the six letters targeted during the intervention. The researcher assessed the participants’ performance on these tasks on two occasions.

Procedural Integrity and Interobserver Agreement

To ensure the accuracy and consistency of both procedural integrity and data scoring, all sessions were videotaped. To calculate procedural integrity, both the researcher and a trained research assistant used an experimenter-created checklist, which provided a summary of the steps in each phase of instruction and data collection, to check a randomly selected 20% of the instructional sessions and LSC probes on a weekly basis. The percentage of instructional and probe steps completed correctly equaled 100%. To calculate interobserver agreement for data scoring, the researcher randomly selected 20% of the independent practice sessions and LSC probes on a weekly basis. Next, the research assistant then scored the students’ responses, while blind to the condition (baseline or instructional). Interobserver agreement averaged 99% throughout the intervention for both independent practice (range = 93%-100%) and the LSC probe (range = 92%-100%).

Data Analysis

Data were graphed and analyzed visually for trend, level, and variability (Gast, 2010). Data were also analyzed using Nonoverlap of All Pairs (NAP), a recommended effect size measure of data overlap for single-subject studies that addresses some of the concerns with previous effect size methods used in single-subject research (Parker & Vannest, 2009). When using NAP, all baseline data points are compared with all intervention data points. NAP is calculated using the following formula: $NAP = (N_{A} \times N_{B}) - (O + . 5 [T]) / N_{A} \times N_{B}$ . In the equation, $N_{A}$ = the number of baseline data points, $N_{B}$ = the number of intervention data points, O = the number of overlapping data points from A and B phases, and T = the number of comparisons where the data are equivalent (Parker & Vannest, 2009). Scores of 0 to .65 using NAP are considered to have weak treatment effects, .66 to .92 medium effects, and .93 to 1.0 strong effects (Parker & Vannest, 2009).

Social Validity

After the intervention was completed, three ECSE professionals who worked with the participants completed a brief questionnaire to collect information on the perceived acceptability of the intervention in early childhood settings. The questionnaire included a 5-point Likert-type scale for three questions and two open-ended questions. The questionnaire asked the ECSE professionals to indicate their level of agreement (1 = strongly disagree to 5 = strongly agree) regarding their perceptions of the intervention and the participants’ performance. In addition, the questionnaire asked the teachers to describe any benefits of the intervention they observed and any aspects of the intervention they would change.

Results

Effectiveness of Instruction

All three participants demonstrated an increase in the total number of correct responses provided on the LSC probe after instruction. Figure 1 presents the graphs representing the number of correct responses provided by the three participants on the LSC probe during (a) baseline, (b) intervention, and (c) maintenance phases.

Figure 1.

Number of correct responses (out of 12) on LSC probes.

Cameron

During baseline, Cameron demonstrated an average score of 1.3 correct responses on the LSC probes (11% accuracy), below chance levels (2/12 or 17%). Cameron demonstrated a treatment effect (i.e., an increase of at least two correct responses from the highest point observed at baseline for two consecutive LSC probes) after nine instructional sessions, and he continued to demonstrate improvement as new LSCs were introduced. By the end of the intervention, Cameron had received instruction for all six of the targeted LSCs and he correctly provided correct responses for 11 of the 12 trials (92% accuracy) on the final LSC probe. Using NAP, a strong intervention effect of .99 was calculated for Cameron.

Zach

During baseline, Zach demonstrated an average score of 2 correct responses on the LSC probes (17% accuracy), at chance levels. During baseline and the initial intervention probes, Zach typically selected the letter o for every trial, which resulted in a score of 2 for every LSC probe (every LSC probe contained 2 trials for each letter). Zach demonstrated a treatment effect after receiving 13 instructional sessions—He consistently selected the correct response for the first two letters that were introduced for instruction (o and t) for two consecutive probes. Zach continued to demonstrate improved performance as subsequent letters were introduced. By the end of the intervention, Zach had been received instruction for five LSCs (o, t, r, l, u) and demonstrated acquisition for four (o, t, r, l). He provided correct responses for 10 out of the 12 trials (83%) on the final LSC probe.

When reviewing Zach’s performance during the later stages of instruction (including generalization and maintenance measures), it is important to take into consideration that 2 of the 12 trials included an LSC (u) for which Zach had received instruction but had not yet demonstrated acquisition, and 2 trials for an LSC (p) that the researcher had not yet introduced to Zach. When looking at Zach’s performance on the LSCs for which he had demonstrated acquisition at the time the final LSC probe was administered, he correctly identified the LSC for 97% of these trials. Using NAP, a medium intervention effect of .90 was calculated for Zach.

Cole

During baseline, Cole’s performance demonstrated an average score of 2.3 trials correct on the LSC probes (19% accuracy), close to chance levels (2/12 or 17%). Cole demonstrated evidence of a treatment effect after 11 instructional sessions. By the end of the intervention, Cole demonstrated acquisition for four LSCs (o, t, r, l) and he provided a correct response for 10 out of the 12 trials (83%) on the final LSC probe. Like Zach, it is important to note that 2 of the 12 trials on the final LSC probe included an LSC (u) for which Cole had not yet reached criterion during the instructional sessions and 2 trials for an LSC (p) that the researcher had not yet introduced to Cole. On the final LSC probe, Cole correctly identified 87% of the LSCs for which he had received instruction and demonstrated acquisition. Using NAP, a medium intervention effect of .88 was calculated for Cole.

Maintenance

The researcher implemented three LSC probes to measure the participants’ maintenance for identifying the LSCs acquired during the intervention, 4 weeks after the final instruction session for Cameron and Cole, and 5 weeks after the last instructional session for Zach. Although both Cameron and Cole demonstrated a slight decrease in some of their scores on the maintenance LSC probes, all three participants demonstrated evidence of maintenance 4 to 5 weeks after not receiving any instruction, with a mean average score of 9.6 (range = 7–12).

The maintenance scores reflect the total number of correct responses for all six LSCs. As previously noted, because of the end of the school year Zach and Cole had only received instruction and demonstrated acquisition for four of the six LSCs included in the LSC probes. When taking into account only the letters for which Zach and Cole had demonstrated acquisition, Zach scored 88%, 100%, and 100% (an average of 96%), and Cole scored 100%, 100%, and 88% (an average of 96%).

Generalization

All three participants demonstrated evidence of generalization across materials and tasks during the generalization probes. The three participants demonstrated scores between 8 and 11 trials correct on the Generalization Book Probe (Cameron: 8, 11, 11; Zach: 9, 9, 10; Cole: 10, 10, 10). The three participants demonstrated scores between 9 and 12 trials correct on the Generalization Alphabet Probe (Cameron: 10, 12; Zach: 10, 9; Cole: 11, 12). When only taking into account the LSCs for which Zach and Cole had demonstrated acquisition, they both demonstrated scores of 100% (i.e., 8 out of 8) for all generalization probes.

Efficiency of Instruction

Because of time limitations associated with the end of the school year, the three children varied in the number of instructional sessions received. There also was some variation observed in the number of instructional session needed to reach acquisition for a letter.

Cameron

Cameron demonstrated acquisition for all six LSCs after receiving 36 instructional sessions. Given that each session consisted of approximately 10 min of instruction and 5 min of extension activities, Cameron acquired six letters after approximately 360 min of explicit instruction and 180 min of extension activities for a total of 540 min (9 hr) of instruction. Cameron’s rate of acquisition for reaching criterion on an LSC ranged from 3 to 12 instructional sessions. During independent practice, Cameron demonstrated an average of 82% accuracy on previously acquired LSCs.

Zach

Zach demonstrated acquisition for four LSCs after participating in 31 instructional sessions and had also begun instruction for the letter u during this time. The school year ended before Zach demonstrated acquisition for u and before he received instruction on p. Zach demonstrated acquisition for four LSCs after approximately 390 min (6.5 hr) of intervention (instruction and extension activities). Zach’s rate of acquisition for reaching criterion ranged from 4 to 9 instructional sessions for each LSC. During independent practice, Zach demonstrated 100% accuracy on previously acquired LSCs.

Cole

Cole demonstrated acquisition for four LSCs after receiving 18 instructional sessions. During these sessions, Cole received instruction on five of the six target LSCs. The school year ended before Cole demonstrated acquisition for the LSC, u, and before receiving instruction on p. Cole acquired four letters after approximately 210 min (3.5 hr) of intervention (instruction and extension activities). Cole’s rate of acquisition for reaching criterion ranged from 3 to 4 instructional sessions for each LSC. During independent practice, Cole demonstrated 100% accuracy on previously acquired LSCs.

Social Validity

After instruction was completed, the researcher provided a Social Validity Questionnaire to three ECSE professionals, all of whom worked with the participants. In response to the statement, “The student’s reading skills improved as a result of the intervention,” two professionals indicated “strongly agreed” and one indicated “agreed.” All three ECSE professionals indicated that they “agreed” with the following statements: (a) “The reading intervention could be implemented in the ECSE classroom or natural environment by a teacher/therapist” and (b) “The reading intervention would be beneficial for other children with autism.”

No formal data on the participants’ views of the activities were collected, although there is also some anecdotal evidence that the activity was viewed as preferred by the children who participated. The participants were observed to laugh and smile as they engaged in the activities, and it was not found to be necessary to provide reinforcement (i.e., choice activities) to maintain engagement. Although it is difficult to interpret a single behavior with confidence, on one occasion a participant saw the instructor/researcher outside at recess, led her by the hand into the classroom, and sat down with her at the desk that was used for reading instruction, which may be interpreted as a self-initiated request for the reading instruction activity.

Discussion

The current study examined the effects of adapted reading instruction on the acquisition of LSCs for young learners with ASD and limited speech. The results provided evidence that the instructional package was an effective and efficient means of teaching LSCs to young learners with ASD and limited speech and was viewed as appropriate by the ECSE personnel. The three children who participated in this study learned new LSCs, maintained their knowledge of the LSCs after instruction was completed, and generalized this information to modified task formats.

The current study extends the previous research related to reading instruction for young learners with ASD and learners with limited speech. While past literacy research on persons with autism and limited speech has typically included school-age children and adolescents (Flores, Shippen, Alberto, & Crowe, 2004), the results from this study provide evidence that implementing adapted reading instruction with very young children with ASD and limited speech can be an effective, efficient, and appropriate method of teaching LSCs.

Effectiveness

There are several factors that may have contributed to the positive impact of instruction on acquisition of LSC with very young learners with ASD and limited speech. Specifically, the instruction included (a) instructional content recommended by the National Early Literacy Reading Panel (Lonigan & Shanahan, 2008); (b) evidenced-based teaching practices, including explicit and systematic instruction, instructional scaffolding, and immediate and corrective feedback (Archer & Hughes, 2011; Lonigan & Shanahan, 2008); and (c) adaptations for participants with limited speech that eliminated the need for an oral response during instruction (Fallon et al., 2004; Light & McNaughton, 2009; Light et al., 2008). Following the completion of instructional activities, all three participants demonstrated maintenance of previously learned LSCs, with an average score of 82% to 100% accuracy on previously acquired LSCs presented during independent practice. Participants also demonstrated the ability to make use of their LSC skills in two generalization activities.

Efficiency

The approach described here also appears to provide an efficient method of instruction. Variation was observed in the number of sessions needed to support acquisition of a letter, ranging from 3 to 12 sessions for each LSC (i.e., a total of 45–180 min per LSC). Variation was also observed across participants for acquisition of LSCs: Cameron averaged 6 instructional sessions for each LSC, Zach averaged 6.5 instructional sessions, and Cole averaged 3.5 instructional sessions per letter sound.

Although there are only limited data available, it is also interesting to note that the participants in the current study did not appear to be slowing down in their rate of acquisition as new letters were introduced. For all three participants, the number of sessions required for the last letter acquired (3, 4, and 3 for Cameron, Zach, and Cole, respectively) was fewer than the number of sessions needed for the first letter acquired (5, 4, and 4), providing some evidence that the rate of acquisition was not decreasing as new letters were acquired. One possible explanation is that as the children became more familiar with the instructional activities, and developed a better understanding of the alphabetic principle (i.e., that letters can be used to represent sounds), the number of sessions needed to learn additional letters was reduced. Cameron did require an extended number of instructional trials (12) to demonstrate acquisition for the LSC for u (the fifth of six letters he acquired); however, this may have represented some difficulty with that particular letter.

Other researchers have investigated the relative efficiency of instruction for the acquisition of LSCs for learners with significant disabilities (Blischak, Shah, Lombardino, & Chiarella, 2004; Fallon et al., 2004; Flores et al., 2004; Millar, Light, & McNaughton, 2004), although it is challenging to make comparisons across studies. Flores et al. implemented systematic and explicit instruction to teach LSCs to elementary students (7–13 years of age) with moderate intellectual disabilities, which included one learner with autism. These students reached criterion for identifying LSCs after receiving approximately 3 to 16 instructional sessions per LSC. The student with autism (age 7) required 3 to 5 sessions per LSC. The number of minutes focused on teaching LSCs was not stated in the Flores et al. study; therefore, it cannot be determined if sessions were equal in length to those reported here. The instructional procedures in Flores et al. also required participants to produce oral responses.

In another study, Millar and colleagues (2004) provided explicit instruction on five LSCs to three children (7–10 years of age) with developmental disabilities and limited speech. The researchers found that two out of three participants acquired five LSCs after receiving 10 to 14 instructional sessions (i.e., ranging from 2 to 4 instructional sessions per letter). In the Millar et al. study, however, the researchers reported that a session typically included 30 to 45 min of instruction (which included instruction on LSCs, identifying initial phonemes in words, and a writer’s workshop-type activity). These sessions were 2 to 3 times as long as the sessions conducted in this study and provided instruction on additional literacy skills.

Social Validity

The provision of formal literacy instruction at an early age requires the use of effective and efficient techniques that can have a powerful impact on literacy skills (Lonigan & Shanahan, 2008), while acknowledging the importance of providing motivating and engaging activities that are viewed as appropriate for very young children (Connor, Morrison, & Katch, 2004). In this study, we focused on teaching specific reading behaviors using instructional activities frequently observed with older students but which needed to be adapted to address the interests and skills of young learners with autism and limited speech. The ECSE teachers who observed these instructional activities reported that they saw these activities as consistent with activities in an ECSE classroom (and indeed, one specifically commented that she liked the fact that the instruction was systematic and structured). The use of materials that were engaging to the children (e.g., pictures of cartoon characters, animals) while introducing LSCs and practicing LSCs during extension activities also helped to promote the acquisition of important early literacy skills while acknowledging the developmental level and interests of the children.

Limitations

Although this study provides evidence of the positive impact of teaching LSCs to young children with ASD and limited speech, there are limitations that should be considered. First, this study only included three children with ASD and limited speech; therefore, more research is required to determine the generalizability of these results to other young learners with ASD and limited speech. Second, it was necessary to end instruction before all learners had demonstrated acquisition of all LSCs because of the end of the preschool year; while both Zach and Cole were continuing to demonstrate acquisition of new LSCs and maintenance of those LSCs that had previously been introduced, more information on their performance with all of the targeted LSCs (and indeed with all LSCs) would be of interest. Finally, although it was the specific purpose of this study to focus solely on teaching LSCs, it is important to note that teaching other reading skills (e.g., phonemic awareness) is appropriate and often recommended (Lonigan & Shanahan, 2008), as well as the integration of reading instruction into a wide variety of language and literacy activities (Connor et al., 2004; Lonigan & Shanahan, 2008). Although our instruction focused on the acquisition of a specific skill within a specific task, we taught this skill with the knowledge that a wide variety of literacy activities (e.g., shared read alouds, vocabulary development activities) were also taking place in the classroom and that a wide variety of print awareness and other early literacy skills were addressed at these times. We also worked to incorporate the child’s interests into the activities, both by using familiar cartoon characters in the instructional activity and through the creation of books on preferred topics (e.g., animal alphabet books, I-Spy books) in the extension activities. Future research should investigate both other ways to further incorporate children’s interests into the instructional activities and ways to better integrate language and literacy instruction into the entire day, including (as appropriate) ways to support family involvement (Connor et al., 2004; Romski, Sevcik, Barton-Hulsey, & Whitmore, in press). One special area of interest may be the participation of both children with and without disabilities in early literacy instruction activities—The skills targeted in this intervention are appropriate for same-age peers without disabilities, and the use of inclusive small group instructional activities would have both social and educational benefits (Koppenhaver & Erickson, 2003; Mirenda, 2014).

Conclusion

The results of this study provide evidence that young children with autism and limited speech can acquire LSCs using a structured intervention program that was positively received by the ECSE teachers and the children themselves. In addition to developing early reading skills and the traditional benefits of literacy, learning LSCs at a young age can support the development of spelling and writing skills, which individuals with limited speech can then use as part of their AAC system to maximize participation in a variety of environments (Light & McNaughton, 2009; Lund & Light, 2006; McNaughton & Light, 2013).

Footnotes

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

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Adams

M. J.

(1990). Beginning to read: Thinking and learning about print. Cambridge, MA: MIT Press.

Archer

A. L.

Hughes

C. A.

(2011). Explicit instruction: Effective and efficient teaching. New York, NY: Guilford Press.

Bailey

R. L.

Angell

M. E.

Stoner

J. B.

(2011). Improving literacy skills in students with complex communication needs who use augmentative/alternative communication systems. Education and Training in Autism and Developmental Disabilities, 46, 352–368.

Ball

E. W.

Blachman

B. A.

(1991). Does phoneme awareness training in kindergarten make a difference in early word recognition and developmental spelling? Reading Research Quarterly, 26, 49–66.

Barker

R. M.

Bridges

M. S.

Saunders

K. J.

(2014). Validity of a non-speech dynamic assessment of phonemic awareness via the alphabetic principle. Augmentative and Alternative Communication, 30, 71–82. doi:10.3109/07434618.2014.880190

Barker

R. M.

Saunders

K. J.

Brady

N. C.

(2012). Reading instruction for children who use AAC: Considerations in the pursuit of generalizable results. Augmentative and Alternative Communication, 28, 160–170. doi:10.3109/07434618.2012.704523

Beukelman

Mirenda

(2013). Augmentative and alternative communication: Supporting children and adults with complex communication needs (4th ed.). Baltimore, MD: Paul H. Brookes.

Blischak

D. M.

Shah

S. D.

Lombardino

L. J.

Chiarella

(2004). Effects of phonemic awareness instruction on the encoding skills of children with severe speech impairment. Disability and Rehabilitation, 26, 1295–1304.

Browder

D. M.

Ahlgrim-Delzell

Flowers

Baker

(2012). An evaluation of a multicomponent early literacy program for students with severe developmental disabilities. Remedial and Special Education, 33, 237–246. doi:10.1177/0741932510387305

10.

Browder

D. M.

Wakeman

S. Y.

Spooner

Ahlgrim-Delzell

Algozzine

(2006). Research on reading instruction for individuals with significant cognitive disabilities. Exceptional Children, 72, 392–408.

11.

Carnine

D. W.

Silbert

Kame’enui

E. J.

Tarver

S. G.

(2010). Direct instruction reading (5th ed.). Upper Saddle River, NJ: Prentice Hall.

12.

Coleman-Martin

M. B.

Heller

K. W.

Cihak

D. F.

Irvine

K. L.

(2005). Using computer-assisted instruction and the nonverbal reading approach to teach word identification. Focus on Autism and Other Developmental Disabilities, 20, 80–90.

13.

Connor

C. M.

Morrison

F. J.

Katch

L. E.

(2004). Beyond the reading wars: Exploring the effect of child-instruction interactions on growth in early reading. Scientific Studies of Reading, 8, 305–336.

14.

Dynia

J. M.

Lawton

Logan

J. A.

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, 142–153. doi:10.1177/0271121414536784

15.

Fallon

K. A.

Light

McNaughton

Drager

Hammer

(2004). The effects of direct instruction on the single-word reading skills of children who require augmentative and alternative communication. Journal of Speech, Language, and Hearing Research, 47, 1424–1439.

16.

Flores

M. M.

Shippen

M. E.

Alberto

Crowe

(2004). Teaching letter sound correspondences to students with moderate intellectual disabilities. Journal of Direct Instruction, 16, 173–188.

17.

Foley

B. E.

Wolter

(2010). Literacy intervention for transition-aged youth: What is and what could be. In McNaughton

Beukelman

(Eds.), Transition strategies for adolescents and young adults who use AAC (pp. 35–68). Baltimore, MD: Brookes.

18.

Gast

D. L.

(2010). Single subject research methodology in behavioral sciences. New York, NY: Routledge.

19.

Horner

R. D.

Baer

D. M.

(1978). Multiple-probe technique: A variation of the multiple baseline. Journal of Applied Behavior Analysis, 11, 189–196.

20.

Kennedy

C. H.

(2005). Single-case designs for educational research. Boston, MA: Pearson.

21.

Koppenhaver

D. A.

(2000). Literacy in AAC: What should be written on the envelope we push? Augmentative and Alternative Communication, 16, 270–279.

22.

Koppenhaver

D. A.

Erickson

K. A.

(2003). Natural emergent literacy supports for preschoolers with autism and severe communication impairments. Topics in Language Disorders, 23, 283–292.

23.

Koppenhaver

D. A.

Evans

Yoder

(1991). Childhood reading and writing experiences of literate adults with severe speech and motor impairments. Augmentative and Alternative Communication, 7, 20–33.

24.

Light

McNaughton

(2009). Addressing the literacy demands of the curriculum for conventional and more advanced readers and writers who require AAC. In Soto

Zangari

(Eds.), Practically speaking (pp. 217–245). Baltimore, MD: Paul H. Brookes.

25.

Light

McNaughton

(2014). Communicative competence for individuals who require augmentative and alternative communication: A new definition for a new era of communication? Augmentative and Alternative Communication, 30, 1–18. doi:10.3109/07434618.2014.885080

26.

Light

McNaughton

Weyer

Karg

(2008). Evidence-based literacy instruction for individuals who require augmentative and alternative communication: A case study of a student with multiple disabilities. Seminars in Speech and Language, 29, 120–132.

27.

Lonigan

C. J.

Shanahan

(2008). Executive summary of the report of the National Early Literacy Panel. In National Early Literacy Panel—Developing early literacy: Report of the National Early Literacy Panel (pp. v–xii). Washington, DC: National Institute for Literacy. Retrieved from http://www.nifl.gov/earlychildhood/NELP/NELPreport.html

28.

Lund

S. K.

Light

(2006). Long-term outcomes for individuals who use augmentative and alternative communication: Part I—What is a “good” outcome? Augmentative and Alternative Communication, 22, 284–299.

29.

Machalicek

Sanford

Lang

Rispoli

Molfenter

Mbeseha

M. K.

(2010). Literacy interventions for students with physical and developmental disabilities who use aided AAC devices: A systematic review. Journal of Developmental and Physical Disabilities, 22, 219–240. doi:10.1007/s10882-009-9175-3

30.

Mayer-Johnson

(1992). The picture communication symbols. Solana Beach, CA: Mayer-Johnson.

31.

McNaughton

Light

(2013). The iPad and mobile technology revolution: Benefits and challenges for individuals who require augmentative and alternative communication. Augmentative and Alternative Communication, 29, 107–116. doi:10.3109/07434618.2013.784930

32.

Millar

D. C.

Light

McNaughton

(2004). The effect of direct instruction and writer’s workshop on the early writing skills of children who use augmentative and alternative communication. Augmentative and Alternative Communication, 20, 164–178.

33.

Mirenda

(2003). “He’s not really a reader . . .”: Perspectives on supporting literacy development in individuals with autism. Topics in Language Disorders, 23, 271–282.

34.

Mirenda

(2014). Revisiting the mosaic of supports required for including people with severe intellectual or developmental disabilities in their communities. Augmentative and Alternative Communication, 30, 19–27. doi:10.3109/07434618.2013.875590

35.

Parker

R. I.

Vannest

(2009). An improved effect size for single-case research: Nonoverlap of all pairs. Behavior Therapy, 40, 357–367.

36.

Romski

M. A.

Sevick

Barton-Hulsey

Whitmore

(2015). Early intervention and AAC: What a difference thirty years makes. Augmentative and Alternative Communication. Advance online publication. doi: 10.3109/07434618.2015.1064163

37.

Travers

J. C.

Higgins

Pierce

Boone

Miller

Tandy

(2011). Emergent literacy skills of preschool students with autism: A comparison of teacher-led and computer-assisted instruction. Education and Training in Autism and Developmental Disabilities, 46, 326–338.

38.

Whalon

K. J.

Al Otaiba

Delano

M. E.

(2009). Evidence-based reading instruction for individuals with autism spectrum disorders. Focus on Autism and Other Developmental Disabilities, 24, 3–16.

39.

Wodka

Mathy

Kalb

(2013). Predictors of phrase and fluent speech in children with autism and severe language delay. Pediatrics, 131, 1–7. doi:10.1542/peds.2012-2221