Reading Instruction for Learners with Intellectual Developmental Disabilities and Complex Communication Needs

Abstract

Communication incorporates the transfer of information and knowledge amongst individuals. Communication takes many different forms including but not limited to printed text, images, and verbal communication. Literacy instruction strengthens and develops communication skills by developing foundational skills necessary for reading. Individuals with intellectual and developmental disabilities (IDD) process information and instruction differently than their neurotypical counterparts and have less access to phonics reading intervention. Mainstream phonics instruction and curriculum do not incorporate adaptations or evidence-based practices for individuals with IDD. Many individuals with IDD also have complex communication needs and use multimodal and augmentative methods of communication. Assistive technology devices and strategies allow learners with complex communication needs to access instruction and provide means for communication. The following research study uses a match-to-sample procedure to develop letter–sound correspondence using single-output recordable buttons. Participants increased correct responding following intervention and demonstrated initial generalization when auditorily matching letter sounds to the matching letter presented in the augmentative and alternative communication device.

1. Introduction

Communication is vital for transferring information from one person to another. Communication fuels collaboration, understanding, trust, and knowledge (Bruce et al., 2026). Individuals with intellectual and developmental disabilities (IDD) often face challenges in communication (Westover, 2010). An intellectual disability can be defined as “a group of developmental conditions affecting the cognitive functioning of a person and, consequently, their capacity to learn and function independently” (Unugrean, 2021, p. 248). Intellectual disabilities are determined with below average intellectual functioning and significant impairments in at least two of the following adaptive skills: practical skills, social skills, or conceptual skills (Unugrean, 2021). Some individuals with IDD have complex communication needs and may use alternative methods and technology to communicate (Light & Drager, 2007). Without functional communication, individuals with disabilities often face continuous challenges and obstacles into adulthood.

Individuals with complex communication needs often rely on others to provide symbols and pictures to express their ideas (Ulriksen et al., 2023). There are multiple different types of devices to assist individuals with communication. Augmentative and alternative communication (AAC) devices and other audio-output assistive technology devices increase access to spoken language and allow users to demonstrate comprehension of spoken and written language. AAC can be low-tech, mid-tech, or high-tech (Light & Drager, 2007). A high-tech device uses some form of technology often an internet capable tablet with communication applications specifically for nonverbal communicators.

Whole-body nonverbal communicators and AAC users process language differently than oral communicators. The difference in language processing requires alternative systems for teaching beginning literacy skills. Individuals with disabilities also face barriers to acquisition of literacy skills due to a variety of factors including expectations of skill development by parents or teachers as well as limited availability of evidence-based intervention programs and adapted materials for AAC users (Ulriksen et al., 2023). It is important that interventions include corrective feedback, repetition, and opportunities to generalize skills (Ulriksen et al., 2023).

Literacy skills enhance and provide alternate methods of communication. The foundational literacy skills of phoneme awareness and letter–sound correspondence are the two greatest predictors of future reading ability (Johnston et al., 2009). For individuals with complex communication needs who use alternative methods to communicate, there is significant evidence that phonological awareness facilitates beginning reading and spelling (Truxler & O’Keefe, 2007). Most programming for individuals with a significant intellectual disability access a modified curriculum in a separate classroom that focuses on functional life skills. Functional life skills can be defined as “skills that are most important for autonomy and independence in homes, communities, and jobs” (Ruppar et al., 2011, p. 100). Literacy includes oral or written communication experienced by individuals throughout their everyday lives and contexts; therefore, literacy fits the definition of a functional skill (Ruppar et al., 2011). Literacy exists in every domain of life from reading the menu at a restaurant, completing a medical document, understanding common signs in the community, and more. Literacy instruction is important for all individuals as literacy is necessary for an individual to attain any level of independent life.

1.1. The Science of Reading

The science of reading is based on a collection of research conducted over many years that outlines the key skills and methods needed for effective reading instruction. The National Center on Improving Literacy (2024b) defines the science of reading (SOR) as an approach to teaching reading that is based on five main ideas: fluency, vocabulary, phonics, comprehension, and phonemic awareness. Structured literacy instruction is informed by SOR and is an approach to reading instruction that is based on four key elements. Structured literacy instruction is evidence-based, systematic, diagnostic, and explicit. For instruction to be evidence-based, it should incorporate the elements of phonology, sound–symbol association, syllables, morphology, syntax, and semantics (National Center on Improving Literacy, 2024a).

Systematic instruction starts with instruction on the most basic concepts and progresses in difficulty of concepts upon mastery. Phonology and sound–symbol association are the first two concepts taught through structured literacy instruction. Phonology focuses on the sound structure of words also known as phonemes, and sound–symbol association gives the instruction on matching phonemes to specific letters or groups of letters (Reading Rockets, 2026). Diagnostic instruction is tailored to meet individual student needs through continuous data collection and assessment to extend knowledge or provide extra opportunities for practice as needed. Explicit literacy instruction implements direct teaching of skills with modeling, guided practice with immediate corrective feedback, and opportunities for students to independently practice while continuously progress monitoring and reteaching as needed (National Center on Improving Literacy, 2024b).

1.2. Statement of the Problem

Most literacy instruction and curriculums are designed for the neurotypical population and do not include adaptations or modifications, rendering the instruction inaccessible for many individuals with IDD. Mainstream phonics instruction and the teaching of letter–sound correspondence for the general education demographic is heavily reliant on the ability to verbally produce letter and digraph sounds. A misconception regarding literacy instruction is that individuals need to be able to orally produce speech sounds to accurately sound out words when learning to read (Yorke et al., 2020). Individuals with complex communication needs are frequently unable to produce sounds or words to receive feedback which serves as the most common type of phonics instruction method (Johnston et al., 2009). Not only are individuals with complex communication needs unable to participate fully in general education phonics instruction, when compared to typically developing peers, they are less likely to receive literacy learning activities or develop advanced literacy skills (Johnston et al., 2009).

Historically literacy and reading instruction for students with cognitive disabilities have been underemphasized (Browder et al., 2006). This lack of emphasis on literacy instruction may be attributed to negative perceptions of individuals with cognitive and communicative disabilities and their ability to learn. Prior expectations of this population were that they may be able to develop a sight word bank, but further reading skills were unattainable (Browder et al., 2006). Sight word recognition is a small part of literacy instruction but is not included as one of the predictors of future reading as it does not teach a student how to decode words but rather automatically know a small set of words based on memorization (Browder et al., 2006). Sight word instruction is important; however, it should be used in complement to other literacy domains to create a holistic approach. Letter–sound correspondence does provide an indicator for reading skills as it develops skills for decoding and blending words.

Current general education phonics and literacy programs are fast paced and often not able to be differentiated to meet the needs of diverse learners (Yorke et al., 2020). Students who use AAC to communicate may also present characteristics that require more intensive and individualized instruction to acquire a new skill (Browder et al., 2006). When considering the population of individuals with IDD, it is important to consider the content, and the strategies used to teach the content. Therefore, a need exists for evidence-based strategies for teaching phonics to individuals with complex communication needs.

To address this need, researchers have conducted studies incorporating AAC devices into reading instruction based on the science of reading. Yorke et al. (2020) completed a systematic review of the literature surrounding reading interventions for individuals who use AAC and found that the interventions with the most benefits for participants were those that included direct instruction, consistency between skills explicitly targeted during instruction that were then evaluated with dependent variable measures, and interventions that focused on a single skill rather than five or more skills at once. Direct instruction is an evidence-based strategy for teaching individuals with significant cognitive delays in the acquisition of new skills as well as a key feature of structured literacy instruction informed by the SOR. Direct instruction utilizes different supports and prompting strategies for students. This includes most-to-least prompting hierarchy beginning with errorless learning and slowly fading prompting as the student begins to gain independence through repeated opportunities for practice (Light et al., 2008). Errorless prompting strategies that incorporate a model and guided response have been shown as effective for individuals with and without disabilities (Browder et al., 2006; Light et al., 2008).

There is a growing body of research which addresses some of the shortcomings of standard literacy instruction for students with complex communication needs. These studies use elements of direct and systematic instruction and align with the SOR. Choi et al. (2015) used a match-to-sample procedure intervention to teach listener literacy and echoic clarity to three elementary school students in a self-contained special education classroom. The participants were below grade level in reading and writing. Two students were autistic, and the other participant was diagnosed with attention-deficit hyperactive disorder. All participants used inexact echoic responses to adult prompts and had decreased accuracy in response to adult-spoken instructions. The match-to-sample (MTS) procedure targeted matching auditory sample sounds. The procedure used computer-based software that displayed three red circles on a screen. There was one circle on the top of the screen, and two red circles positioned side-by-side on the bottom half of the screen with a solid horizontal black line separating the top and bottom circles. When touched or clicked, the top button would produce a spoken sample or word, while the bottom buttons had one produce an identical target sound and the other with a distractor sound. Participants were taught to match the top sample sound with one of the comparison buttons. All directions for participants were provided vocally by the researcher. The authors found that following mastery of the MTS procedure all participants had an increased correct response to listener literacy probes. Participants also demonstrated increased correct full and partial echoic responses.

In a similar study, Ahlgrim-Delzell et al. (2014) used AAC to measure phoneme identification and blending sounds to form words with and without picture referents. High-tech devices were used in the Alghrim-Dezell study and the study presented in this paper. The study implemented a phonics curriculum that used stories made of CVC words with phonemes from the lessons. The instruction targeted three phonemes while addressing phoneme identification, determining beginning sounds, phonemic segmentation of CVC words, blending words to form words with and without picture referents, sight word recognition and reading comprehension connected to text. Instruction was delivered to three elementary students with moderate to severe intellectual disabilities who were currently or had recently accessed a foundational literacy program, could not identify at least five letter sounds through vocalizations, and had the physical capacity to use an AAC device. The results found that all three participants showed improvement in accuracy of correct responses to probes for phoneme identification and blending sounds with and without picture referents. These studies are encouraging but should be replicated in different settings with different student populations to increase the generalizability of the findings.

1.3. Research Questions

The research presented within this study systematically combines evidenced-based strategies for teaching individuals with complex communication needs with a match-to-sample procedure developed by Choi et al. (2015). The purpose of this study was to implement explicit and systematic reading instruction targeting foundational skills (e.g., phonology, letter–sound correspondence) for elementary students with complex communication needs and provide additional evidence that students with intellectual and developmental disabilities can successfully acquire and benefit from these skills.

We designed the following research questions to meet this goal.

• What are the effects of using a match-to-sample and errorless learning procedure to teach phonological awareness for students with complex communication needs?
• What are the effects of the intervention for repeated acquisition of phonological awareness skills?
• What are the effects of intervention on phonics skills (i.e., letter–sound correspondence)?

2. Materials and Methods

We used a multiple probes across participants with embedded multiple probes across sounds design (Gast et al., 2018) to measure the effects of the intervention package on phoneme identification and generalization to letter–sound correspondence.

2.1. Participants

The participants were enrolled in a public-school program servicing students with extensive support needs. Students within the program had a special education eligibility of multiple disabilities or intellectual disability as determined through school-based assessment. Within the classroom, the students accessed a modified curriculum with a focus on functional academics. The inclusion criteria to participate in the study were (a) intellectual disability in the moderate to severe range; (b) daily use and access to an augmentative and alternative communication device or strategies; (c) complex communication needs or limited verbal language; (d) inability to recognize at least five letter–sound correspondences; and (e) elementary school age. The participants were selected from a pool of thirty-five students recommended for intervention by their education specialist.

Consent and Assent Procedures

The participants were students at the elementary school where the first author taught. Parents/guardians of potential participants were contacted via email by the second author to describe the study, and provide an opportunity for questions, then consent forms and information about the study were sent home for parents/guardians to review. Four families were contacted as potential participants, and four participants submitted parental consent and were enrolled in the intervention.

To obtain assent from the participants directly, students were read an assent script and given an opportunity to use their visual communication strips to say yes or no to participating and to request “more” or to be “all done.” All four students provided assent following the script. To measure ongoing assent and engagement, the researcher observed students for signs of physical or emotional distress before, during, and after each session. Examples of physical and/or emotional distress include but were not limited to crying, screaming, attempting to walk away from the instructor/instructional setting, throwing materials, and hitting or attempting to hit themselves or others. If participants engaged in these behaviors prior to a session, the researcher delayed the session until a time when the participant was not displaying signs of distress. If the participant engaged in these behaviors during a session, the researcher discontinued the session. If two consecutive sessions ended this way, it was viewed as withdrawal of assent to participate in the study. Based on this, one participant was withdrawn from the study, and their data have been removed from the research and results of the study.

In addition to monitoring for indications of withdrawal of assent, the first author noted several instances of participant behavior which could be interpreted as ongoing assent. For example, Student 2 would pick up the box of materials needed for the intervention and bring them to the first author. Student 1 would sit at the table where research was conducted and would call the first author to come to the table. When other participants were engaged in the intervention, Student 3 would state “my turn” to request access to the intervention.

2.2. Setting and Materials

Participants received the intervention within scheduled academic times during the school day. Baseline and intervention sessions were conducted in the classroom where students regularly attended and received instruction. No school personnel or students not normally present were present during sessions. Sessions appeared similar in structure to normal instructional practices, so there was no indicator for potential observers that a different procedure was being implemented. Other students within the classroom accessed different small group instruction at other tables within the classroom. Intervention occurred in ten-to-fifteen-minute one-on-one daily sessions with the researcher. Participants engaged in no more than two intervention sessions per day.

During intervention sessions, participants sat at a table in a child-sized chair, and the researcher sat to the right of the participants. Three single-output switches were on the table for every trial. The buttons were pre-recorded with a letter sound that would play when pressed. Reinforcers were used to sustain participant engagement and motivation. Positive praise and high-fives were used with all participants. Each participant had a repertoire of individualized reinforcers based on specific interests. Student 1 brought a preferred toy to the table to earn, chose coloring or typing, and engaged in dance breaks. Student 2 used a token board exchange system where five trials would earn a preferred toy. When responding correctly, Student 2 could press a separate pre-recorded button that said “good job.” Student 3 requested tickles and making silly faces.

2.3. Measures

The dependent variables for this study were alternative communication responses of sounds and letters. During pre-baseline, baseline, phase 1, and phase 2 of intervention, participants used single-output AAC switches to produce target sounds. During phase 3 of the intervention, participants selected letters using their high-tech speech generating devices.

Sounds and letter–sound correspondence targets were selected based on instructional sequences described by Patching et al. (1983) and due to frequency of usage in CVC words used within literacy instruction programs for decoding and blending (Ahlgrim-Delzell et al., 2014). Specifically, the sounds /s/, /ă/, /t/, /i/, and /p/ were targeted along with the corresponding letters s, a, t, i, and p.

When conducting trials, responses were recorded using +c/f, −c/f, or inv in a Microsoft Excel spreadsheet. A correct response was recorded as a + when the student pressed the desired button with their hand and then did not interact with any other buttons. A correct response was also recorded if the student pressed the desired button repeatedly or picked up the button and then pressed it. An incorrect response was recorded as a—when the student pressed the undesired button with their hand and then did not interact with any other buttons. An incorrect response was also recorded if the student pressed the undesired button repeatedly or picked up the button and then pressed it. An invalid response was recorded as inv when the student interacted with the buttons in any way that is not what is listed above. This includes pressing multiple buttons, not pressing any buttons, pushing the buttons, swiping the buttons, and lifting multiple buttons.

2.3.1. Randomization and Target Sound Selection

During pre-baseline, all five target sounds were probed with each participant. Following pre-baseline data collection, the three sounds with the lowest average correct responses were selected to be utilized for each student during the baseline and intervention phases of the study. Following baseline, intervention order for each sound was randomized using a web-based random number generator (i.e., a virtual six-sided dice). Each sound was assigned two numbers, and the sound associated with the randomized number was introduced first. Randomization of tiers improves the rigor of the design by decreasing the likelihood of bias in selection of intervention order (Kratochwill & Levin, 2014).

2.3.2. Data Collection Procedures

Three switches were presented to participants. When pressed, the top button would produce a speech sound. The participants were taught to match the speech sound when provided with a target and distractor sound response choice. Correct responses were reinforced through delivery of praise or tangible reinforcers while incorrect responses were corrected through research guided modeling and were not reinforced. Baseline data were collected in the same setting as the intervention. During baseline neither corrective feedback nor reinforcement were provided for correct or incorrect answers.

2.4. Reliability and Fidelity

Sessions were observed via teleconference software, were not recorded, and only viewed by approved project personnel from a secure device on a university campus in accordance with the protocol approved by the university institutional review board. Interobserver agreement (IOA) was measured using Trial-by-Trial IOA. This was done by the smaller of the two counts being divided by the larger and then multiplied by 100 (Cooper et al., 2020). Using this method, the third author watched sessions via Zoom and collected data. The data was marked as “c,” “f,” or “inv” on a Microsoft Excel sheet. During the intervention phases, “gp” and “pp” were also used to mark gestural prompting and physical prompting respectively. The third author then compared their data with the first author and totaled the number of agreements. The number of agreements was divided by the number of opportunities. During training sessions, there were necessary clarifications and changes that needed to be made to the procedure prior to IOA being collected.

Across all phases and students, there were a total of 149 sessions. IOA was collected during 59 of these sessions across students and phases. This information was then added to a different Microsoft Excel sheet. IOA was collected during 39.6% of sessions across phases and students. The average IOA was 97.45% across all students and phases. During baseline, the average IOA was 96.91%; during intervention, the average IOA was 97.45%. For Student 1, IOA ranged from 62.5 to 100%. For Student 2, the IOA ranged from 60 to 100%. For Student 3, the IOA ranged from 78.57 to 100%.

Fidelity was collected utilizing a task analysis. The third author watched the session via Zoom and collected fidelity data. A correct step was marked with an “x” while an incorrect or missed step was left blank. If a trial had to end in the middle due to behavior or other circumstances, the following steps were marked with “n/a.” During training sessions, the task analysis was refined. The third author also reviewed procedures after the training sessions with the first author to review and discuss any frequently missed steps that happened during training sessions. During baseline, fidelity of intervention was 97.06%. During intervention, fidelity of intervention was 99.70%.

2.5. Treatment Package

The intervention consisted of a treatment package of multiple effective interventions combined to create a literacy skills development program (Cooper et al., 2020). The intervention replicates elements of Choi et al. (2015) and Ahlgrim-Delzell et al. (2014; i.e., match-to-sample procedures) as well as errorless learning practice sessions.

Table 1. Study replication elements.

	Ahlgrim-Delzell et al. (2014)	Choi et al. (2015)	Present Study
Design	Multiple Baselines Across Participants	Delayed Multiple Probes	Multiple Baselines Across Sounds and Participants
Methodology
Procedure	Match-to-sample	Match-to-sample	Match-to-sample
Setting	One-on-one phonics setting	Self-contained 3rd grade classroom	Elementary IDD classroom
Dependent Variable & Measurement
Targeted Skill	Letter/word/phrase–sound correspondence, Comprehension questions	Letter/word/phrase–sound correspondence	Auditory matching, letter–sound correspondence
Dependent Variables	Phoneme identification, blending sounds with and without picture referents	Correct responses to advanced listener literacy probe trials, echoic clarity	Correct responses to matching trials across AAC modalities
Intervention Implementation Materials	GoTalk Phonics	Computer-based auditory match-to sample software	Pre-programmed buttons, high-tech AAC letters

outlines procedural similarities and differences from the replicated studies.

The intervention incorporated a match-to-sample procedure modified to meet the specific and unique needs of the learners. When creating a literacy instructional program for learners with complex communication needs it is important to include “(a) systematic targeting of appropriate skills, known to contribute to positive literacy outcomes; (b) implementation of effective evidence-based instructional procedures, known to result in successful skill acquisition; and (c) development of appropriate adaptations to accommodate the unique needs of individuals with complex communication needs” (Light et al., 2008, p. 126). The letter–sound correspondence intervention combines multiple techniques and strategies to combat the lack of adequate phonics instruction for individuals with IDD. The letter–sound correspondence intervention incorporated errorless learning sessions before probing trial data. Systematic correction procedures involved a model–prompt–check sequence where the instructor provided an explanation of the correct response, the model, leads the student in making a correct response, the prompt, and, finally, provided the student with an opportunity to select the correct response independently, the check (Fallon et al., 2004). The letter–sound MTS intervention incorporated direct instruction and systematic correction sequences through prompting and modeling by the researcher. Skills were targeted daily with participants, and reinforcement was provided to support explicit instruction.

2.5.1. Pre-Baseline

During pre-baseline, probe data were collected for the selected five sounds (i.e., /s/, /ă/, /t/, /i/, /p/). Each sound was probed three to five times over nine consecutive school days. The distractor sound during probe sessions was one of the four remaining sounds not being targeted. The distractor sound was selected at random by taking out one switch from a box without looking at the sound prior to selection. After all probe data were collected, target sounds were identified based on the lowest average performance. For Student 1, the average percentage correct for the five sounds during pre-baseline ranged from 8.33 to 44.44%. For Student 2, the average percentage correct for the five sounds during pre-baseline ranged from 8.33 to 41.67%. For Student 3, the average percentage correct for the five sounds during pre-baseline ranged from 11.11 to 44.44%. Target sounds for each student were as follows: /ă/, /i/, and /p/ for Student 1; /s/, /i/, and /p/ for Student 2; /ă/, /t/, and /p/ for Student 3.

2.5.2. Baseline

During baseline the switches were configured in a triangle-like manner on the table (Appendix A). One switch was placed at the top across from the participant. The top switch, further referenced as the exemplar, played the target sound when pressed. There were two switches placed in front of the participant horizontally spaced out approximately five inches apart from each other. One of the bottom switches had the target sound, and the other had a distractor sound. The researcher invited participants to the table and explained that they were going to practice matching letter sounds. To conduct the baseline procedure, the researcher first modeled the procedure before probing data. The researcher pressed the exemplar, the switch farthest from the researcher, the closest switch, then the top exemplar again. The researcher presented the verbal stimulus “match” and pressed the bottom switch with the target sound. To probe for data collection, the researcher pressed the exemplar, then the farthest switch, the closest switch, and finally the exemplar again. The researcher stated “match” and waited for a participant response. The location of the target sound and distractor sound were moved after every trial to prevent position control. If the participant pressed the target sound, the trial was marked as correct. If the participant pressed the distractor sound, the trial was marked as incorrect. If the participant pressed the exemplar or multiple switches, the trial was marked as invalid. If the participant did not press any switches, the trial was marked as no response. No response and invalid responses were scored as incorrect when graphing and analyzing data. No corrective feedback or response-specific praise was provided during the baseline procedure.

2.5.3. Phase 1 Auditory Matching Following Errorless Trials with Pre-Session Errorless Practice

Phase 1 was run with a similar technical procedure as baseline with the switches placed in the same configuration. The key difference between baseline and phase 1 was the incorporation of errorless teaching and the use of corrective feedback or reinforcement. The intervention phase incorporated a more intensive teaching session prior to probing data collection trials. By introducing pre-session errorless teaching sessions, the participants increased access to repeated opportunities to practice, which is an evidence-based practice for students with IDD. These techniques involve the gradual transfer of stimulus control (Cooper et al., 2020).

To conduct the first phase, the researcher invited the participants over individually and explained the objective of the session was to practice matching letter sounds. The researcher modeled the procedure by pressing the exemplar, pressing the far switch, pressing the close switch, pressing the exemplar, and saying “match” followed by pressing the switch with the target sound. The participants engaged in seven to ten errorless trials before data collection. Immediately following the seven to ten errorless practice trials, the participants engaged in five to ten data collection trials. The researcher pressed the exemplar, the far switch, the close switch, the exemplar, and then stated “match.” If the participant did not initiate response within three seconds of the stimulus being provided, the researcher provided the least intrusive prompt to evoke an accurate response from the participant. If the participant lifted their hand to choose the distractor sound, the researcher provided a partial physical prompt to guide the participant by the elbow to the target sound. Corrective feedback was immediately provided for incorrect responses, and reinforcement was used for correct responses. The location of the target sound was moved after every trial to prohibit position control.

2.5.4. Phase 2: Stimulus Shape Transformation

Once participants demonstrated mastery criterion for their three target sounds, they entered the letter–sound correspondence phase. The letter–sound correspondence phase was designed to pair the letter with the sounds the participants were auditorily matching. Phase two included the labeling of the target and distractor sound switches. The exemplar remained unlabeled while the two bottom switches were labeled with the letter associated with the sound they output. The researcher first invited participants to the learning area and described the procedure as matching letter sounds. The researcher pressed the exemplar, the far switch, the close switch, the exemplar, and provided the verbal instruction to “match.” The researcher then pressed the switch with the target sound. The participant completed five to seven trials of errorless practice before data collection trials. For the first three trials of repeated errorless practice, the researcher described the correlation between the letter label and the sound. The researcher pressed the exemplar, the far switch, the close switch, the exemplar, stated “match”, and provided a prompt for the participant to press the target sound. The researcher stated “letter __ says /__/” for the first three errorless trials. For example, if the exemplar and target sound were P, after providing a prompt to evoke a correct response the researcher stated, “letter P says /p/.” The participant continued with prompted trials while receiving reinforcement. For the data collection trials, the researcher pressed the exemplar, the far switch, the close switch, the exemplar, and stated “match.” The researcher provided consistent wait time and provided reinforcement or corrective feedback dependent on participant response. The location of the target sound was switched after every trial to minimize position control.

2.5.5. Phase 3: AAC Generalization

The final stage of the letter–sound correspondence intervention was designed to measure generalization of the skills developed during the MTS protocol. Phase three incorporated participants’ individual AAC devices. The researcher invited participants to the table and explained that the objective of the session was to practice letter sounds. The individual participant’s AAC device was brought to the table by the participant or set up when the participant joined the table. The AAC device was open to the keyboard or alphabet page. The researcher first modeled the protocol for all three target sounds before probing data collection trials. The researcher stated, “what letter says ___.” During the pause, the researcher pressed the switch with the recorded target sound. The researcher then selected the correct letter on the AAC device. The researcher then positioned the AAC in front of the participants. The researcher cued “what letter says__” while pressing the target sound switch then waited for the participants’ response. A correct response was characterized through the selection of the corresponding letter on the AAC device. An incorrect response was determined through either no response or the selection of any letter other than the corresponding letter. In the case of an incorrect response, the researcher provided immediate corrective feedback. The researcher selected the correct letter on the AAC and stated “letter __ says __ (pressed target sound switch) __.” The target sound changed after every trial. Accuracy was determined through the average correct response rate per target sound.

3. Results

Primary experimental control and decisions to move through intervention were based on the multiple probes across participants design (Gast et al., 2018) targeting the first sound.

3.1. Design and Analysis

Figure 1 presents data for Sound 1 for all three participants using a multiple probes across participants design. Baseline data for all three participants are displayed to the left of the phase change line. Phase one intervention data (errorless practice sessions prior to data collection) are displayed to the right of the phase change lines. Following the logic of the multiple-probe design, each tier of intervention is introduced at different points in time to allow for the demonstration of a functional relation.

Figure 2, Figure 3 and Figure 4 also apply the same design logic as Figure 1 but use a multiple probes across behavior (i.e., specific sounds) design. The first tier of Figure 2, Figure 3 and Figure 4 correspond to the data included in Figure 1. Data to the left of the phase change lines in Figure 2, Figure 3 and Figure 4 represent baseline measures for target sounds, and data to the right of the phase change lines represent phase one intervention data across sounds.

The use of these nested single-case designs provides an opportunity to demonstrate a functional relation between the intervention and the dependent variables both across participants (Figure 1) and within participants (Figure 2, Figure 3 and Figure 4). For Figure 1, Figure 2, Figure 3 and Figure 4, the y-axes represent the percentage of correct responses per session. The x-axes represent separate sessions conducted; for points 1–30 only a single session was conducted per day and the remaining points represent two possible sessions per day (e.g., a student with a point on 31 A and the next point had two sessions conducted on day 31).

Figure 5 includes additional data for each student and displays data for phase 2 and phase 3 for Student 1 and Student 2. Data points to the left of the phase change line are from phase 2 (stimulus shape generalization) and data to the right of the phase change line are from phase 3 (AAC generalization). Data for Student 3 only represent phase 2 of the intervention. X- and y-axes scaling and interpretation are identical to Figure 1, Figure 2, Figure 3 and Figure 4.

3.2. Across Participant Data

Figure 1 displays results for all three participants. For all three students, baseline data are variable with some data points overlapping intervention data. On average, intervention data are at a higher level than baseline data for all students. It is also of note that all students had multiple data points at 100% accuracy during intervention and no data points of that significance during baseline. Similarly, all students had multiple data points at 0% accuracy during baseline and only the initial two data points for Student 2 were at 0%. Taken collectively, these data suggest a possible functional relation between the intervention and correct auditory matching behavior. Additional evidence and context for the functional relationship is demonstrated at the individual student level.

Participant Response Variability

Participant response patterns introduced variability in performance data. Student 2 demonstrated challenges with transition from baseline to intervention with the change in expectation of structured responses which contributed to variability within the intervention phase. During baseline, Student 2 frequently picked up switches to listen close to their ears. During intervention, the expectation was to keep the switches in the correct position. Errorless teaching with full-physical prompting was implemented to teach appropriate response methods. During pre-baseline and baseline, the first author observed Student 3 to select the left switch for most trials. During the intervention phase, Student 3 demonstrated task-avoidant behaviors consistent with performance outside of the intervention work which affected variability of responses during the intervention phase of Sound 1. Student 3 received strategic behavioral interventions throughout the school day and continued to demonstrate assent while engaging in task-avoidant behaviors by touching switches and requesting “more”.

3.3. Student 1

Figure 2 displays data across sounds for Student 1. During baseline probes of the three target sounds, Student 1 selected the correct switch in an average of 9.26% of trials. Sound /ă/ and sound /p/ had the highest average (11.11%), and sound /s/ had the lowest average (8.33%). The baseline data was variable at a low level for all three sounds. Following randomization, sound /ă/ was selected for intervention. Student 1 worked with sound /ă/ during the intervention for three sessions before demonstrating a stable increased level of accurate responses. An increased stable level of responses was characterized by at least five sessions of 80.00% accuracy or higher. Once sound /ă/ was at an increased stable level, intervention was introduced for the next target sound. Throughout intervention of sound /ă/, once hitting the mastery response rate (initial data collection session of 80% correct responses), Student 1 demonstrated accurate selection of target sound /ă/ in 91.48% of trials. During intervention of sound /p/, Student 1 worked with sound /p/ for three sessions before seeing an increased level of stable responses. From the point of reaching a mastery criterion level to the end of intervention for sound /p/, Student 1 selected the accurate switch in 100% of trials. After sound /p/ was at a stable, increased level of response, sound /s/ was introduced into intervention. Student 1 showed immediate mastery of sound /s/ and accurately selected the target sound in 96% of trials. The correct response average of all trials across the intervention phases was 84.65%. The intervention data are consistently at a higher level than baseline data points across all three sounds. The student also reached mastery criteria and high level at a faster rate for each sound.

3.4. Student 2

Figure 3 displays data across sounds for Student 2. During baseline probes of the three target sounds, Student 2 selected the correct button in an average of 19.44% of trials. Sound /s/ had the highest average (33.33%), and sound /i/ had the lowest average (8.33%). During baseline, the level was low for all three sounds; on occasion, baseline data points overlapped with intervention data. Following randomization, sound /s/ was selected for intervention. After moving into intervention, Student 2 received intervention on sound /s/ for two sessions before initially hitting the mastery criterion of 80% correct response rate. Once reaching the mastery criterion Student 2 demonstrated an accurate response rate of 88.76% with sound /s/. Once sound /s/ was at an increased stable level for at least five sessions, intervention was introduced for sound /p/. After two sessions, Student 2 reached a stable, increased level of correct response rate and maintained an average of 75.3% accuracy rate. Once sound /p/ was at a stable level, intervention for sound /i/ was introduced. Sound /i/ reached a stable increased level after one session and Student 2 demonstrated 100% correct response rate after demonstrating the mastery criterion level for the remaining trials. For sounds 1 and 2, the student had variability prior to mastery criterion being met. Once mastery criterion was met, the data remained at a high level. For Sound 3, the student had no variability and remained at a high level throughout all trials. Student 2 had an average correct response rate of 81.88% across all intervention trials.

3.5. Student 3

Figure 4 displays data across sounds for Student 3. During baseline probes of the three target sounds, Student 3 selected the correct button in an average of 37.47% of trials. Sound /t/ had the highest average (44.22%), and sound /ă/ had the lowest average (33.33%). Baseline data had large amounts of variability across trials with the level being predominantly low. Following randomization, sound /ă/ was selected for intervention. After moving into intervention, Student 3 demonstrated an overall average accuracy of 73.48% with sound /ă/. After three sessions of intervention with sound /ă/, Student 3 demonstrated an increased stable response rate of sound /ă/. Once sound /ă/ was at an increased stable level for at least five sessions, intervention was introduced for sound /t/. The student immediately responded at the mastery criterion level and demonstrated an overall average accuracy of 87% with sound /t/. Once sound /t/ was at a stable increased response rate level, intervention was implemented for sound /p/. Sounds 1 and 3 show variability, but, on average, the level is high. For Sound 2, the level is high with low variability between trials. Again, the student demonstrated an immediate response to intervention and had an average correct response selection of 77.78% of opportunities presented for sound /p/. Student 3 had an average correct response rate of 79.42% across all intervention trials.

3.6. Phase 2: Stimulus Shape Transformation

Table 2. Correct responses by phase, sound, and student.

Student	Sound	BL	Phase 1	Phase 2	Phase 3
1	1	8.33%	79.39%	100.00%	60.71%
	2	8.33%	78.57%	96.00%	92.86%
	3	11.11%	80%	96.43%	86.43%
2	1	10.00%	68.56%	93.33%	85.00%
	2	24.24%	77.07%	95.00%	100.00%
	3	6.41%	100%	100.00%	100.00%
3	1	33.33%	73.48%	95.00%	N/A
	2	43.06%	87.00%	72.00%	N/A
	3	35.15%	77.78%	92.00%	N/A
Across Students Mean		20.00%	80.21%	93.31%	87.50%

displays data across sounds and phases for all students, and Figure 5 shows response rates across phases 2 and 3 for all students and sounds. Once Student 1 demonstrated stable levels of accurate selection of the target sound during intervention, Phase 2 was initiated with auditory matching while switches were labeled with corresponding letters. Student 1 demonstrated immediate high levels of accurate selection of the labeled target sound. In phase 2, Student 1 had an accurate response rate of 100% with sound /ă/, 96% with sound /p/, and 100% with sound /s/. Within phase 2, Student 1 demonstrated a 98.67% accurate selection response rate across all trials.

Student 2 entered phase 2 after showing stable levels of response with all three target sounds in the intervention phase. Target sounds labeled with the corresponding letters were introduced to Student 2 in the same order as he received intervention. Student 2 had one session in phase 2 below the mastery criterion for sound /s/ and all other sessions were at a 100% correct selection rate. Student 2 had 100% accuracy for sounds /p/ and /i/. Out of all trials within phase three, Student 2 correctly selected the target sound in 96.88% of opportunities.

After demonstrating stability in performance during the initial intervention stage, Student 3 moved into the second phase with labeled sound selection options. Student 3 immediately demonstrated mastery of matching the labeled letter sounds to the target sound. Student 3 selected the target sound /ă/ 85% of the time. For sounds /p/ and /t/ the correct switch was selected 100% of the time. Across all trials in phase 2, Student 3 responded with a correct response in 92.06% of opportunities presented.

3.7. Phase 3: AAC Generalization

Phase 3 was the final generalization and expressive understanding phase of the intervention protocol and utilized the student’s AAC device to select the letter corresponding to the target sound presented. Student 1 demonstrated mastery criterion of 80% accuracy per target sound on the second session of phase 3. For sound /s/, the correct response rate was 92.86%. For sound /ă/, the accuracy average was 60.74%. Finally, for sound /p/, Student 1 had an 86.43% correct response rate. Across all trials, Student 1 provided the correct letter selection on his AAC 79.27% of the opportunities presented. Student 2 was immediately successful in the final phase of intervention. Student 2 responded correctly to the sound /s/ 95% of the time, to sound /i/ 92% of the time, and sound /p/ 72% of the time. Across all trials in phase 3, the correct letter was selected in 86.33% of opportunities presented.

4. Discussion

Communication is fundamental for collaboration between individuals and participating in education. Individuals with complex communication needs face challenges in both general communicative interactions and more complex applications like reading. The purpose of this study was to determine the effects of an intervention using match-to-sample combined with errorless learning procedures to teach phonological awareness for students with complex communication needs. We used the MTS letter–sound correspondence intervention with three students. We found that following intervention, students increased correct auditory matching, and these skills could generalize to letter correspondence and use of individual AAC devices. This study connects and expands the extant research by demonstrating student learning through embedded multiple-probe designs across participants and behaviors, providing evidence for teaching foundational literacy skills incorporating AAC methods, and demonstrating the importance for expanding literacy instruction beyond high-frequency and sight words for individuals with complex communication needs.

First, results from this study expand and provide context to the extant literature surrounding literacy instruction for students with complex communication needs. Similar to Choi et al. (2015), our research saw an improvement in the accuracy of responses between pre-intervention and post-intervention probes. Pre-intervention or baseline probes yielded low accuracy percentages while post-intervention or probes at the end of intervention phases yielded significantly higher accuracy percentages. Therefore, our research findings agree with Choi et al. (2015) that MTS can lead to improvements in sound matching procedures for students using alternative methods of communication. The methods used in the letter–sound correspondence intervention differ from Choi et al. (2015) through the implementation of the MTS intervention using single-output switches rather than computer-based formats. We also extended upon the methods of Choi et al. (2015) by focusing research on the effectiveness of the MTS intervention and its use in an academic context (i.e., a busy special education classroom).

Second, this study presents additional evidence of how foundational literacy skills can be taught and adapted using AAC methods in ways that mirror reading instruction for students that communicate vocally. The MTS letter–sound correspondence intervention was founded on the same principles of the SOR which are used to develop general education literacy instruction (National Center on Improving Literacy, 2024b). The letter–sound intervention was systematic in the way it followed the logical order of language. Systematic instruction is key for structured literacy which is based on SOR. The letters and sounds implemented throughout instruction were phonemes that are high frequency in text. Structured literacy also emphasizes the importance of explicit instruction or the deliberate teaching of concepts (National Center on Improving Literacy, 2024a). The adapted letter–sound intervention implemented errorless researcher-led guided practice during every session for all participants. Phonology and phonics instruction are elements of SOR and structured literacy. Phonics is one of the five target areas of instruction within the SOR (National Center on Improving Literacy, 2024b). The use of letter–sound correspondence and the focus on individual phonemes demonstrates phonics within the adapted intervention. The research findings indicate that once participants mastered skills in phase one, they more readily adapted to phase two and three, showing mastery and success after fewer sessions.

Finally, results from this study demonstrate the importance of expanding reading instruction beyond high-frequency and sight words for students with complex communication needs. Sight word and high-frequency vocabulary instruction are often implemented in classroom servicing individuals with IDD because of the correlation to functional and daily living reading (Browder et al., 2006). In addition to this functional reading approach, research should be continued in other aspects of literacy instruction for individuals with complex communication needs. Past studies and the research presented in this article demonstrate the ability to teach phonics skills to individuals with complex communication needs and IDD (Browder et al., 2006). Letter–sound correspondence and similar phonics skills are indicators for future reading abilities. Therefore, a combination of sight-word and phonics instruction would create a holistic approach to reading instruction which would develop skills for literacy and functional reading.

4.1. Limitations and Future Directions

Limitations to the study include a small sample size. Future researchers should consider a larger sample size for more opportunities to replicate the findings. The current study started with a small sample size with one participant who withdrew, leading to only two participant replications. The research was implemented late in the school year, which created a small timeframe. Another limitation within the timeframe was the unpredictability of student attendance. Student attendance varied, and absences limited opportunities for data collection days. A longer timeframe to implement the research would decrease data collection restraints and allow for more robust research on the generalization of letter–sound correspondence skills. Another limitation to the research is the absence of social validity data. Social validity is the measure of social significance of the goals of the intervention, the social appropriateness of the intervention procedures, and the social importance of the effects of the intervention (Carter & Wheeler, 2019). Social validity data collection could be implemented by future practitioners and researchers using surveys and observational records. Parent and teacher questionnaires could provide insight for the effectiveness and appropriateness of the letter–sound correspondence instruction. Observational records of student engagement, behavioral responses during intervention, and student performance on phonics-based tasks post-intervention would indicate the effects and appropriateness of the intervention for students with IDD and CCN. A methodological limitation of this study is the absence of fading out the instructional practice sessions prior to trial sessions during phases 1 and 2. During the first and second phases, participants engaged in errorless practice sessions prior to trial sessions. The structure of the errorless practice before probing was intended to provide direct and explicit instruction. During the generalization phase, no practice trials were included. The absence of fading errorless sessions could be seen as a methodological limitation due to the correlation between errorless teaching and prompt dependency. Future studies should examine response patterns in the absence of this support, as well as across a greater number of sounds. Finally, the study did not include a formal maintenance phase due to the school year ending before these data could be collected. Future studies should incorporate maintenance measures to provide clearer evidence that behavioral repertoires are sustained beyond the intervention.

4.2. Implications for Practitioners

The findings of this study suggest that individuals with complex communication needs can learn letter–sound correspondence skills through systematic instruction. There is a functional relationship between a tangible MTS intervention protocol and the mastery of auditorily matching letter sounds. An adapted intervention that eliminates the need to orally produce speech sounds can be used to target letter–sound correspondence. There is a need for further research and development of interventions that teach phonics and early reading skills for individuals with complex communication needs. Modified direct instruction can increase literacy skills within this population. Within the context of a self-contained special education classroom, the intervention can be implemented during one-on-one instruction sessions. The intervention could be implemented by teachers or other trained educational staff and could be conducted in a variety of school settings.

5. Conclusions

Literacy skills are important for all individuals, including individuals with complex communication needs. Individuals with IDD benefit from an adapted curriculum with explicit instruction, corrective feedback, and repeated opportunities for practice. The research findings of our letter–sound correspondence intervention indicated that MTS procedures can be used to adapt curriculum and provide further access to phonics instruction for students with IDD. The literature surrounding phonics instruction and interventions targeting literacy skills for individuals with IDD and complex communication needs remains scarce and further research should be conducted to further evaluate the effectiveness of adapted instruction.