Abstract
This study examined the response of two students in the Southwestern United States with autism spectrum disorder and IQ in the intellectual disability range to a comprehensive, text-based reading intervention. The intervention, Friends on the Block, includes multiple strands of literacy providing explicit instruction in phonics, phonemic awareness, high-frequency irregular or temporarily irregular words, decoding, vocabulary, oral language, and comprehension. Multiple supports for students with intensive needs are embedded within the intervention, such as specially designed multi-criteria books, extensive scaffolding, and intensive cumulative review. Teachers customize the program to address the varying needs of students. We used single-case multiple baselines across levels of instruction design to assess growth on a measure of reading (words read correctly). Results indicated a positive, functional relation between reading intervention and word reading. Both students also demonstrated modest growth on measures of phonemic awareness, letter–sound knowledge, and word identification from pre- to post-intervention.
As the number of students identified with autism spectrum disorder (ASD) continues to rise (American Psychiatric Association [APA], 2013; Head et al., 2018), expectations for reading performance for these students have been increasing due to heightened accountability standards and research advances (Browder & Spooner, 2014). While students with ASD have been included in research on reading development, this research is often limited to students with ASD who have average or above average IQ (El Zein et al., 2016; Flores & Ganz, 2007). However, as of 2019, an estimated 33% of children diagnosed with ASD also have concurrent Intellectual Disability (ID, CDC, 2014, 2020). Understanding the response of students with ASD and ID to reading instruction is vital to ensure that they are developing foundational skills required for early reading (i.e., phonemic awareness, letter-sound knowledge, word identification, and language comprehension; Castles et al., 2018; Foorman et al., 2016; Gough & Tunmer, 1986; National Reading Panel & National Institute of Child Health and Human Development, 2000). These skills are all integral pieces of reading development and relate to a comprehensive approach to reading instruction.
Students diagnosed with both ASD and ID are likely to share many instructional needs with students with only ID, including difficulties with exposure to novel information (APA, 2013). These challenges relate to their language development and cognitive functioning, although both vary greatly among children with ASD (APA, 2013; Lord et al., 2004). Specific deficits related to language development that commonly occur in children with ASD and ID include difficulty with social and spontaneous language, social use of communication, and oral language/listening comprehension (Allor et al., 2020; APA, 2013; Ganz & Flores, 2009). Additional challenges are that many students with ASD and ID have limited background and vocabulary knowledge and often demonstrate difficulty transferring skills across tasks (Castles et al., 2018; Conners et al., 2006; Joshi & Aaron, 2000; Leach et al., 2003; Scarborough, 2001). These delays in language development are often related to delays in decoding and reading comprehension as well as difficulty participating in reading instruction (Ainsworth et al., 2016).
Emerging research supports the ability of children with ASD to master skills related to decoding and phonemic awareness, such as sound-spellings, blending, segmenting, and word reading (Ainsworth et al., 2016; Light & McNaughton, 2009). A few studies have investigated the effectiveness of systematic phonics instruction and/or Direct Instruction (DI) programs for students with ASD and ID and have yielded promising results (Flores et al., 2013; Flores & Ganz, 2007, 2009; Ganz & Flores, 2009; Head et al., 2018). These findings converge with evidence for the effectiveness of literacy interventions that address both sight words (high-frequency irregular or temporarily irregular words) and phonics instruction; for this population, these are similar to findings of studies that focus on effective reading instruction for students with only ID (e.g., Allor et al., 2014; Connor et al., 2014; Lemons et al., 2015). Researchers have called for additional research on both decoding and comprehension interventions for students with ASD and ID (Head et al., 2018; Nation et al., 2006).
The Present Study
The purpose of the present study was to focus on two students with ASD and IQs in the ID range (i.e., 55 and 70) from within a series of single-case research design (SCRD) studies examining the effects of Friends on the Block (Allor, Cheatham, & Al Otaiba, 2018) for 18 students with disabilities (see Allor et al., 2020; Allor, Gifford, et al., 2018). These two students were specifically chosen as they were the only two students in the study with IQ in the ID range and a diagnosis of ASD whose data have not been previously reported (Allor et al., 2020; Allor, Gifford, et al., 2018). The dependent variable included a researcher-created assessment of word reading, including both sight and decodable words. Across these 18 students, we found consistent moderate-to-strong positive effects, with more demonstrations for sight words than decodable words. This pattern of findings for foundational word reading skills is consistent with the early levels of the intervention, as the majority of students began the intervention in the early levels that focus on sight words, letter sounds, and phonemic awareness, building the skills necessary for later decoding. In addition to the summative study on all 18 students, we have reported more detailed data for eight of these students, including both graphs of word reading data and data from several pre-/post-assessments (Allor, Gifford, et al., 2018). Here, we reported that participants demonstrated at least some growth on curriculum-based pre-/post-measures. Findings from these studies are consistent with research that supports comprehensive literacy instruction for students with ID, including those with IQs as low as 40 (Browder et al., 2012; Connor et al., 2014; Lemons et al., 2015); however, additional research is needed to further examine variables other than word recognition. Teachers reported that the intervention was feasible and implemented the intervention with high degrees of fidelity (Allor, Gifford, et al., 2018).
The present study examines the relationship between participation in Friends on the Block, provided for 21 weeks by the classroom teacher, and improvement in reading for two students from the larger study (Allor et al., 2020) who were identified as having ASD and ID and for whom detailed data have not been previously reported. Word reading assessments to measure the dependent variable were researcher-created and pre-/post-assessments of reading included curriculum-based measures commonly used for screening and progress monitoring which assessed phonemic awareness, letter–sound knowledge, and word identification. We addressed the following research questions in regard to the two students:
Method
Participants and Setting
Participants were two students with ASD, speech-language impairments (SLI), and IQs in the ID range (i.e., 55 and 70). The ASD and SLI diagnoses were provided by the school and research staff administered the Wechsler Abbreviated Scale of Intelligence (WASI-II, Wechsler, 2011) to confirm IQ. Students were recruited following approval from a university institutional review board and met the following recruitment criteria: (a) IQ scores ranged from 40 to 79, (b) currently in Grades 1 through 4, (c) were able to use spoken English as primary communication means, and (d) had limited literacy skills (see Allor, Gifford, et al., 2018). The two students in the current study, Brent and Frank, attended the same public Title 1 school serving students of diverse ethnicities in a large metropolitan district in the U.S. Southwest. Brent’s teacher was an African American woman with 7 years of special education teaching experience, and Frank’s teacher was a Caucasian woman with 25 years of teaching experience. Both teachers held undergraduate degrees in Special Education. Prior to the intervention, research assistants collected baseline and descriptive data, including age, disability status, IQ, and receptive vocabulary as measured by the Peabody Picture Vocabulary Test-4 (PPVT-4; Dunn & Dunn, 2007).
Both students spent more than 50% of their school day in a special education classroom but were in different rooms due to age (7 and 9 years). The classrooms were designated “resource rooms” by the school and served students with a variety of abilities and disabilities. Some students remained in the room all day and others, including the two participants in this study, participated in general education instruction for music, art, and other activities designated as “specials” with the support of a paraprofessional. The number of students in the classroom varied throughout the day, ranging from three to no more than 15. At the time of the intervention, both students were also participating in another reading curriculum provided by the district to students in special education, which was computer-based and focused only on language comprehension.
Brent, a Caucasian male, was 7 years old at the start of the intervention and diagnosed with ASD and SLI; he had an IQ of 55 (WASI-II). At the pretest, his receptive vocabulary, assessed on the PPVT-4, reflected an age equivalent of 3 years, 4 months (standard score of 51). He initially had limited literacy skills as he was unable to segment or blend phonemes or consistently recognize any words. He frequently exhibited externalizing behaviors during instruction, such as cursing, hitting, spitting, and refusing to work (i.e., ignoring or escaping). He began instruction at Level 1. According to teacher reports and anecdotal information from video recordings for treatment fidelity, his behaviors were exacerbated when he was presented with novel or difficult information.
Frank, an African American male, was 9 years old at the start of the intervention. He was also diagnosed with ASD and SLI, he had an IQ of 70 (WASI-II). At the pretest, his receptive vocabulary, assessed on the PPVT-4 indicated an age equivalent of 5 years and 3 months (standard score of 78). His initial literacy skills were limited. He struggled to segment and blend individual phonemes; however, he was able to correctly identify onset sounds about a third of the time. He was unable to recognize any sight words, with the exception of “a” and “I” and, therefore, he also began instruction in Level 1. According to his teacher and fidelity videos, Frank did not exhibit many externalizing behaviors; however, he often became frustrated during reading instruction and refused to participate by putting his head down on his desk. Prior to beginning the intervention, he expressed to the researchers and to his special education teacher that he did not enjoy reading and would rather work on other academic subjects, such as mathematics or science.
Brent and Frank both received Friends on the Block instruction in their own special education classrooms in a one-on-one setting with their respective special education teachers. Friends on the Block includes activities that may be delivered via paraprofessional; however, due to Brent’s aggressive externalizing behaviors, his classroom teacher chose to provide this portion of the intervention as well. A paraprofessional with whom Frank was familiar provided the additional practice sessions and activities. Over the 21 weeks of participation in Friends on the Block, Brent and Frank received 20.5 and 31.10 hr of instruction, respectively.
The Intervention
Friends on the Block is a comprehensive intervention that includes multiple strands of literacy providing explicit instruction (i.e., routine delivery of instruction and repetitive model of corrective feedback that includes modeling, guided practice, and independent practice, “I, We, You”) in phonics, phonemic awareness, sight words (high-frequency irregular or temporarily irregular words), vocabulary, oral language, and comprehension. Following the model of DI programs, the intervention is organized around a systematic scope and sequence and provides extensive cumulative review and practice for the transfer of skills across instructional strands (Carnine et al., 2010). Each lesson begins with a straightforward, brief warm-up where key skills are explicitly introduced, reintroduced, and/or practiced (e.g., phonemic awareness blending, and letter sounds). With the exception of oral phonemic awareness activities, items are presented on paper for students to say letter sounds and read sight words. There are only a few items per skill. The teacher usually models the first item for each activity using the “I, We, You” procedure, although this can be faded as students develop the skill. When necessary, the teacher may use errorless learning, temporarily having students simply repeat the correct response. For example, the teacher points to the letter, says the sound, and then the students repeat the sound; this process continues for one row of about six letters. Student books include extensive repetition of words being learned and teachers model spoken language as they discuss stories with students and provide practice of basic comprehension strategies. Games provide further intensive practice and opportunities for generalization. A variety of games are used to practice word reading, such as trail games, Bingo, and matching card games (e.g., Go Fish). For example, one bingo game required students to practice blending and segmenting spoken words, placing Bingo markers on top of the pictures for the words. During a scavenger hunt game, students find words in the book and then read a sentence with the word and build the word with magnetic letters (or write the word on a dry-erase board). Games target the specific skills students are currently developing, and many games also provide cumulative review of previous skills in addition to promoting engagement, motivation, and success. Throughout lessons, corrective feedback and scaffolding are used as needed, including errorless learning and echo reading of sentences within books. The consistent procedures and routines are designed to benefit students with ASD who struggle with the presentation of novel information (APA, 2013).
Several features are distinctive and provide support for students with intensive needs, particularly the specially designed early reading books, which address multiple criteria such as decodability, repetition and cumulative review of high-frequency words, familiar topics, and similarity to early oral language. Examples of scaffolds within the books include ample practice with new words, cumulative review, pictures underneath a small number of content words (words that are difficult to decode but improve storylines and support comprehension), and helper text that is read by a teacher or tutor. Helper text fosters listening comprehension and draws student attention to text structure. The sequence of words also facilitates the development of word recognition skills and oral language. For example, target words in the early levels are common words in early spoken language (e.g., like, want) and, along with a few picture words, allowed us to write sentences that were similar to students’ spoken language (e.g., “I like apples” [picture of apples beneath the word]). Students begin reading books with support (i.e., echo reading or immediate corrective feedback) in the very first lesson.
In the first three levels of the intervention, students learn to recognize 22 words while they are developing skills needed to sound out words (i.e., letter sounds and phonemic awareness). Students are placed into the program based on their knowledge of the target words, beginning in Level 1 if they do not already know most of the target words in Level 1. Letter knowledge is not a prerequisite skill, although students with more letter knowledge will learn the target words more quickly and, therefore, move through the early levels more quickly. At the time of the study, the program consisted of eight levels. Teachers customize the program to address the varying needs of students by moving students to the next level when they have learned most of the target words in a level, rather than after a designated number of lessons and by selecting learning games based on student needs. For example, the teacher would select games to practice letter sounds more frequently for students experiencing difficulty with letter sounds. Each level includes multiple books and learning games to provide cumulative review and maintain student interest. (See Allor, Cheatham, & Al Otaiba, 2018 and friendsontheblock.com).
Lesson Structure
Lessons are intended to take place four to five times per week for at least 20 min per session. Additional practice sessions that include book reading and games are recommended and may be provided by the teacher or a paraprofessional. Lessons include: (a) brief warm-up activities, (b) book reading and comprehension instruction, and (c) learning games. First, students participate in brief warm-up activities practicing key phonemic awareness, phonics, and word recognition skills. These activate prior knowledge and introduce new skills, briefly reviewing strategies that will be applied to book reading. High-frequency words, phonemic awareness, and common letter sounds are introduced and practiced during the warm-up activities in Levels 1 through 3. Second, students read the books and participate in discussions during which they are guided to make predictions, use vocabulary, summarize, and make connections to the text. In-text comprehension questions are provided to enhance comprehension, vocabulary, and oral language and provide frequent opportunities to respond. Third, students participate in a learning game, which is designed to practice skills. Additional practice sessions include book reading and learning games and may be led by either the teacher or a paraprofessional.
Training and Fidelity of Implementation
Both teachers participated in a 2-hr training session provided by the researchers. Training included an overview of the scope and sequence, lesson structure, descriptions and modeling of the steps involved in all lesson routines, and guided practice. Training also included modeling and practice on accurately pronouncing sounds in isolation and blending for both phonemic awareness and decoding instructional routines. Teachers were trained to complete daily lesson logs that included information regarding the duration of session and behavior ratings (i.e., poor, good, excellent). Research assistants further supported teachers with bi-weekly coaching in the delivery of the intervention during which they would observe the teachers during instruction and provide notes regarding the fidelity of implementation. They also demonstrated correct implementation of procedures, as needed, frequently demonstrating routines with that teachers’ students. Trained research assistants took monthly videos, totaling four to five videos per student, during teacher instruction and coded each session using a checklist to measure the fidelity of implementation (Allor, Gifford, et al., 2018). Interrater reliability was established by double-coding videos. Multiple trained research assistants participated in double coding, achieving 98% reliability across 20% of the videos, overall.
Assessment Integrity and Interobserver Agreement
During each observed session, the research assistants recorded instructions and completed an observation checklist developed by the primary researcher to ensure correct implementation. During training, the research assistants were taught to use the coding form and achieved 90% agreement with the lead researcher on videos of reading instruction. During data collection, 20% of the videos were double-coded, resulting in one video for each student in this study.
Results from the fidelity checklist indicated that Frank’s teacher implemented Level 1 with 88% fidelity, Level 2 with 94% fidelity, and Level 3 with 96% fidelity. His video for Level 3 was blind-double coded by another trained research assistant, who reached 93% agreement with the original researcher. Brent’s teacher implemented Level 1 with 95% fidelity, Level 2 with 96% fidelity, and Level 3 with 92% fidelity. The video for Level 2 was blind-double-coded by an additional trained research assistant who reached 99% agreement with the original score.
Placement Test
Research assistants administered a placement test that consisted of 24 words (three randomly selected words from each of the eight levels of the intervention). Students were given approximately 5 s to say each word. The placement test was discontinued when the students made four errors in a row. Students began the intervention in the first level in which they were unable to identify all three words. We chose this criterion to increase the likelihood of initial success in the curriculum, as students would move to the next level as soon as they demonstrated mastery of the words in the current level.
Measure of the Dependent Variable
Reading progress during the intervention was measured through weekly word reading assessments, created and collected by the researchers, aligned with the curriculum. Although the curriculum is comprehensive and lessons addressed skills other than word recognition, we chose word recognition as the dependent variable because instructional decisions to move to subsequent levels of the curriculum were based on recognition of the target words. Thus, we used this assessment as the primary measure within our research design. Each word reading assessment contained six words targeted at that level and a baseline measure of six words introduced in the next level. See Table 1 for a list of the six words on the word reading assessment for Levels 1 through 3. The assessment for each level was functionally similar in that each included a new set of target words that were taught and practiced within that level of the intervention. Each week, the students took the assessment administered by the same research assistant so the student would be familiar with the examiner. Words were presented in random order one at a time on paper flashcards (Brent) or the iPad (Frank) based on student performance (Jones et al., 2019). Students were told to read the words as they appeared. Each word was visible for 5 s. Students moved to the next level once they had mastered the words in the current level, mastery was defined as correctly reading four of the six words during two different administrations of the word reading assessment. Using this as the only criterion for movement is possible because the intervention includes extensive cumulative review.
Key Skills for Levels 1–3 of the Text-Based Reading Intervention.
Pre- and Post-Measures
Pre- and post-assessments included both curriculum-based measures (CBM) and standardized assessments, individually administered by trained research assistants. We anticipated the gradual development of these skills and thus did not use these measures to establish a functional relation between the intervention and these outcomes. The 1-min CBM assessments were Achievement Improvement Monitoring System: Letter Sound Fluency Test (AIMSweb, Pearson, 2012), Formative Assessment System for Teachers: Onset Sound and Word Blending tests (FAST, Christ, 2015), Fuchs Word Identification Fluency (WIF, Fuchs et al., 2004), and Dynamic Indicators of Basic Early Literacy Skills: Phoneme Segmentation Fluency (DIBELS PSF, Good et al., 2007). See Allor, Cheatham, & Al Otaiba (2018), or friendsontheblock.com for detailed information about the measures. All assessments were delivered in a familiar, quiet setting.
Analytic Methods
Each of the participants in the study are considered within an independent multiple baseline across levels SCRD. Each study is a comparison of baseline and intervention phase per level (1, 2, and 3). We began by collecting two concurrent data points in all levels and continued collecting continuous baseline data on Levels 1 and 2. However, we paused baseline data collection on Level 3 to increase feasibility for implementers and to avoid student frustration with repeatedly incorrect responses. We began the intervention with Level 1 after collecting five baseline data points per student. After a student met the mastery criterion for Level 1, which was four out of six correct words across at least two data collection points, we began intervention in Level 2 and resumed continuous baseline data collection for Level 3. After a participant met the mastery criterion for Level 2, which was the same as Level 1, we began intervention in Level 3. The study ended after the participants met the mastery criterion for Level 3.
We conducted visual analysis of the data from the word reading assessments, examining each student’s graphs for changes in trend, slope, and variability from baseline to intervention phases (Horner et al., 2005; Kratochwill et al., 2010). Graphs for Levels 1, 2, and 3 are displayed in Figures 1 and 2. The dependent variable was the number of target sight words the student read correctly (the y-axis) during each assessment (the x-axis). Both students completed the minimum of three observations per level as required by What Works Clearinghouse guidelines (Kratochwill et al., 2010). In addition to visual analysis, we used baseline and intervention data to calculate Taubc effect sizes. Taubc is a popular effect size used in SCRD, it is a rank-order correlation between experimental phase and repeated measures (Allor, Gifford, et al., 2018; Brossart et al., 2014; Parker, Vannest, & Davis, 2011; Parker, Vannest, Davis, & Sauber, 2011). We chose the “baseline corrected Tau” (Taubc) for this study (Tarlow, 2017) because it takes into account baseline-to-intervention non-overlapping data analysis and is bounded by the traditional limits of −1 and +1 (Brossart et al., 2014; Tarlow, 2017).
Brent’s Response to Intervention.
Frank’s Response to Intervention.
Results
RQ1: Is There a Functional Relation Between the Intervention and Sight Word Reading?
Both students in this study began the intervention in Level 1 (based on the placement assessment) and mastered Levels 1, 2, and 3 during their 21 weeks of participation. Overall, both students demonstrated improvement in sight word reading during the intervention phase over the baseline, indicating a functional relation between the intervention and word reading ability. Additionally, Taubc effect sizes indicate medium-positive effects for each of the six cases (or six tiers, which are levels mastered per student). See Figures 1 and 2 for graphs and Table 2 for Taubc effect sizes.
Sight Word Effect Size Tau Estimates by Study (Level).
p ≤ .05. **p ≤ .0.
Brent
Visual inspection of the data was conducted for Brent across the three levels of the intervention. Results indicate that Brent read more words during the intervention phases than baseline phases, according to the researcher-developed word reading measure. During baseline there was some variability in Levels 1 and 3; however, all baselines were stable for at least three sessions prior to the intervention phase. During the intervention, an immediate level change or upward trend was established for Brent over all three levels and the data were stable. For Level 1, Brent reached criterion mastery at the third data collection point. This level remained stable for further data points. For Level 2, he reached criterion mastery at the seventh data collection point, and his scores continued to increase across the remaining data points. For Level 3, he achieved mastery at the sixth data collection point. These data support a functional relation between participation in the intervention and sight word reading as measured by the researcher-created assessment.
Frank
Visual inspection of the data was conducted for Frank across the three levels of the intervention (see Figure 2). Results indicate that he was also able to read more words during the intervention phases than the baseline phases, as measured by the word list. During baseline, a stable trend was observed in each level, indicating that sight word reading on the word reading measures prior to intervention did not change over time. During the intervention, an upward trend was established for Frank over all three levels. For Level 1, he reached criterion mastery on the fourth data collection point, remaining stable for the next four data collection points. For Level 2, he reached criterion mastery at the seventh data collection point and remained stable for further data collection. For Level 3, he achieved mastery at the sixth data collection, the remaining data points had some variability (range: 4–6 words read correctly) but remained above criterion levels. These data support a functional relation between participation in the intervention and sight word reading as measured by the word reading assessment.
The Taubc Effect Sizes
A Taubc effect size was estimated for each tier (i.e., intervention level) of the two multiple baseline designs. For each tier, results indicate medium-positive effects for sight word reading. Table 2 provides a summary of the sight word studies per tier. The Taubc percentages for each of the six tiers (i.e., levels, three per student) ranged from 74% to 90%. The mean effect size of Taubc across the six tiers ranged from 0.74 to 0.90, and all were statistically significant (p ≤ .01). Medium-positive effects related to sight word reading are supported, as Parker, Vannest, and Davis (2011) suggest that Taubc estimates between 66-92% should be considered medium, with any percentage above 92% considered strong and below 66%, weak.
RQ2: Did Students Make Gains on Reading Measures After Participating in the Intervention?
Scores on reading measures were examined from pre- to post-intervention across a variety of skills. As seen in Table 3, both students made gains across each measure. While the demonstrated growth was small to modest, growth is evident for both students across measures.
Growth on Distal Measures for Brent and Frank Pre- and Post-Intervention.
Note. FAST = Formative Assessment System for Teachers (Christ, 2015); DIBELS PSF = Dynamic Indicators of Basic Early Literacy Skills: Phoneme Segmentation Fluency (Good et al., 2007); AimsWeb = Achievement Improvement Monitoring System: Letter Sound Fluency Test (Pearson, 2012).
Maximum score is 16. bRaw score. cMaximum score is 10. dFuchs et al. (2004).
Both students’ phonemic awareness abilities improved from pre- to post-assessment on 1-min CBMs. On the FAST, Brent improved from identifying one initial sound to correctly identifying eight initial sounds and Frank improved from five to 12 (maximum score is 16). On the FAST-blending assessment, the student is asked to blend individual sounds into a word (maximum score is 10). At pretest, both students scored 0 on blending. At posttest, Brent and Frank blended four and six words, respectively. On the DIBELS PSF measure, Brent was unable to segment any sounds at pretest and was able to correctly segment seven sounds (per minute) at posttest (isolating the first sound consistently). Frank displayed stronger growth and grew from segmenting one to 34 sounds on this measure isolating all phonemes (2–4) within most words.
Although differences from pre- to post-assessment on the AIMSweb (measure of letter–sound knowledge) were not as great as that of the phonemic awareness measures, both students also improved on this measure. At pretest, neither student identified any letter sounds correctly, but at post-intervention, Brent identified three correct sounds and Frank correctly identified two sounds. During the lessons, both students demonstrated more progress in this area according to anecdotal information captured during observation and shared by teachers during the interview and the increase in scores on the AIMSweb, identifying about 14 letter sounds by the end of the study.
Both students also grew in word identification. Improvement on the Fuchs WIF, which measures sight word recognition from a predetermined list of words similar to a first-grade Dolch list, was modest, with initial scores of 0 for both students and posttest scores for Brent and Frank of five and four, respectively.
Discussion
Few studies have investigated the responsiveness to a comprehensive literacy curriculum of children with ASD and IQ in the ID range (Flores & Ganz, 2009), the current study aims to address this research gap, inform schools and teachers about responsiveness, and add to existing research on Friends on the Block by focusing on students with ASD and ID (Allor et al., 2014, 2020; Allor, Gifford, et al., 2018). Consistent with previous research, results from this study support a positive functional relation between Friends on the Block and sight word reading measures with moderate effects. Small to moderate gains were also seen on pre–post measures, and although these do not support causal conclusions, these gains indicate progress on letter–sound knowledge and phonemic awareness, both necessary for developing decoding skills.
Our primary research hypothesis was confirmed, over 21 weeks of the intervention, there was a positive functional relation between participation in Friends on the Block and scores on a measure of sight word reading. Results from pre-post measures of reading in this study support emerging research, which suggests that children with ASD and IQ in the ID range can master skills such as letter sounds, blending, segmenting, and word reading (Ainsworth et al., 2016; Flores et al., 2013). Additionally, the teachers indicated that participation in Friends on the Block had a positive effect on the behavior and confidence of both students.
An important implication of this study is that these two students with ASD and an IQ in the ID range made gains on measures of word reading as part of an intervention that focused on a full range of literacy skills. Prior to beginning the intervention, both students were reading well below expectations for their ages (i.e., Brent was 7 years old and Frank was 9 years old). The only two sight words either of the students was able to identify on the initial placement test were “a” and “I,” and they scored zero on nearly all commonly administered progress monitoring measures of phonemic awareness, letter sounds, and word identification (see Table 3). They had previously been participating in but had not adequately responded to, an individual, online reading curriculum, which focused primarily on listening comprehension, designated by the district for students who were classified as nonreaders. Prior research has demonstrated that students with ID can learn to identify words through different approaches (i.e., errorless learning, system of prompts, and constant time delay; Allor et al., 2014; Browder et al., 2012); however, the current study finds that these students can learn to identify words within a broader, comprehensive reading intervention. This is important given recent research on providing comprehensive reading instruction for students with ID (e.g., Allor et al., 2014; Connor et al., 2014; Lemons et al., 2015).
Both students made improvements in sight word reading, according to the researcher-created word reading measure, these improvements enabled them to participate more regularly during group literacy instruction in their special education classrooms. Brent and Frank’s sight word reading data demonstrated immediate increases following intervention implementation and increasing trends across each intervention phase. Each of the students began instruction in Level 1, knowing zero and two words, respectively, on a researcher-created placement assessment of reading, and demonstrated similar improvement in Levels 1, 2, and 3, as seen in Figures 1 and 2. Baseline data on untaught words remained low and relatively stable following intervention at each level. Thus, these data provide a strong demonstration of experimental control within single-case multiple baseline design. Both students finished the intervention in Level 3, which incorporates text in books that provide more opportunities to address decoding through phonemic awareness and letter sound knowledge. Instruction focused on decoding words begins at a later level, so it was not assessed.
Pre- and post-reading measures were incorporated to assess students’ growth in phonemic awareness and letter sound knowledge, as well as word identification, and both students improved from pre- to post-assessment on each of the five measures (see Table 3). Although the intervention is comprehensive and addresses multiple skills, based on prior research, we anticipated that these skills would develop relatively slowly and gradually (Allor et al., 2014). When Nation et al. (2006) assessed a group of 41 children with ASD with a broad range of abilities, the researchers found that many children had delays in decoding and suggested additional research in programs that could close this ability gap. While both students in our study were still performing well below grade level at post-assessment, they demonstrated modest growth, across each measure. However, considering their ages and scores at the pretest were zero for most measures, indicating non-response to prior instruction, this growth is encouraging and is likely to have been due to participation in this comprehensive reading intervention. Furthermore, during lessons, the researchers and teachers observed that the students made progress on phonemic awareness and letter sound knowledge, consistently blending and segmenting individual phonemes within three-phoneme words and consistently identifying all 14 letter sounds that had been presented.
In our prior research, teachers and parents shared in focus groups that they were highly satisfied with student progress and the feasibility of the intervention (Allor, Gifford, et al., 2018). Although the dosage was lower than anticipated, it was sufficient to support growth (see Allor, Gifford, et al., 2018 for more information). Anecdotally, both teachers who participated in this study stated that their students seemed to enjoy the routine and structure of the intervention and that student attention and behavior improved during participation. Brent’s teacher noted that when he first began the intervention, he would frequently lose focus; however, after participating for just a few weeks, his ability to sustain focus during reading instruction markedly improved, possibly due to the repetitive design and structure of the lessons, as well as his success. Brent particularly enjoyed book reading and would often choose to engage with eBook versions of the Friends on the Block books that were available for the students to use during independent work time. He would read the eBooks without distracting classmates for small portions of time, something that was not observed prior to participating in Friends on the Block. Frank’s teacher reported that he became more engaged in whole group activities, which she attributed to his growth in confidence related to this intervention.
It is possible that a different approach to placement in the intervention and movement from level to level might result in better outcomes. Our current approach is to determine the starting level and movement to subsequent levels based solely on student knowledge of the target words, rather than multiple skills. Thus far, we have observed that students with limited letter knowledge, like those in this study, move through the levels more slowly than other students who begin the program in Level 1 with more letter knowledge (see intervention.com for additional case studies). Waiting for students to master other intervention targets before moving to the next level would further slow their progress. Furthermore, there is evidence that Brent and Frank made gains in applying phonemic awareness and letter–sound knowledge in their reading, even if they were not yet demonstrating mastery of the skills during assessments. For example, at posttest, Brent and Frank appeared to be in the partial alphabetic phase of word recognition, using some alphabetic knowledge to recognize the target words (see Ehri, 2014). This is evident in that they recognized the words out of context and were able to distinguish between words that shared at least one letter (e.g., like and look, in and no, yes and you).
Finally, it is important to note that we hypothesize that including supported book reading made the instruction overtly purposeful and increased motivation for Brent and Frank in ways that isolated practice on subskills without application to connected text would not have. This observation has implications for the future use of comprehensive approaches to literacy with students having ASD and ID and other more complex developmental disabilities. It is likely that this addition supported the generalization to measures used in the posttest assessment.
Limitations and Directions for Future Research
As with all school-based research, several limitations should be considered within this study. One limitation is that the classroom teachers provided less dosage than we intended. The lessons were designed to be implemented four to five times a week for about 20 min, but due to a variety of reasons (e.g., limited staff, participants’ externalizing behaviors, testing) lessons were not implemented this frequently (three to four times per week for these students). The total dosage over 21 weeks was 20.5 hr for Brent and 31.20 for Frank. The desired total dosage would have been around 52 hr. However, both students made promising growth despite less than ideal dosage. Although the intervention was implemented individually for this study, the intervention may be implemented in small groups as was the case with some of the other participants in the larger study (Allor et al., 2020). Also, due to discontinuous baseline data collection in Level 3, we were not able to detect whether responding remained stable during Level 1 intervention. However, participants responded to zero words correctly during the initial two sessions, responded to zero words correctly when we resumed baseline data collection, and did not have exposure to the target words prior to intervention in Level 3.
A second limitation is that the dependent variable measure for this SCRD study focused on sight word reading although the intervention was comprehensive, addressing a wide range of literacy skills. We did not include multiple skills in our dependent variable measure (e.g., phonemic awareness, letter sounds, and comprehension) as we would have lost flexibility due to students needing to master all skills before moving to the next level. Students demonstrated modest pre- to post-test growth on CBMs of phonemic awareness and letter sounds, and relatively more limited growth on word reading on the Fuchs WIF, however, during the lessons, researchers and teachers observed that by the end of the study they were able to consistently blend and segment words with three phonemes and identify all 14 taught letter sounds. They could also recognize sight words they had been taught and that they practiced in text and games.
Although Friends on the Block addresses comprehension in every lesson and incorporates multiple techniques (e.g., helper text read by the teacher, in-text comprehension questions), measures of comprehension were not included in this study; however, parents and teachers reported increased engagement, ability to make connections, and other behaviors indicative of comprehension growth during participation in the intervention. Additional research is needed to explore multiple variables, including language, listening, and reading comprehension.
Behavior, which can interfere with academic instruction for all children, including those with ASD, was not formally assessed in this study. During the intervention, the researchers and teachers noticed improvements in student behavior (e.g., increased time on task and fewer externalizing behaviors). Behavior ratings were tracked on daily dosage sheets by the teachers, but these were not consistently completed due to logistical challenges. In the future, researchers might consider ways to increase the feasibility of this type of data collection or supplement teacher report with direct observations of students’ behavior.
Instructional Implications
Our findings indicate that the comprehensive approach and consistency in the program worked well for the two students in this study. Multiple basic skill areas and supported text reading are included within each lesson from the beginning of the intervention. Although growth was slow, as is consistent with our own research and the research of others (e.g., Allor, Gifford, et al., 2018; Allor et al., 2014, 2020; Connor et al., 2014; Flores & Ganz, 2007), students demonstrated clear growth across multiple skills. Teachers of students with ASD and ID may need to be flexible with pacing across skills, as growth in phonemic awareness and phonics skills may be slower than the growth of sight words, which is consistent with the results from our pre-post measures (see Tables 3 and 4) and supported in other studies of Friends on the Block (Allor et al., 2014; Allor, Gifford, et al., 2018). Our method of movement through the intervention by level based on learning the targeted sight words worked well as the intervention includes extensive cumulative review. Other programs may not be designed this way, and teachers should consider adding a cumulative review of a broad range of skills while focusing on the acquisition of new skills.
Teachers of both students managed challenging behavior during literacy instruction, particularly Brent’s teacher, as Brent often exhibited externalizing behaviors when he became frustrated. Brent’s teacher managed challenging behavior in several ways. She provided breaks during instruction when appropriate. She also incorporated reward systems with which Brent was already familiar into the Friends on the Block lessons. She also made use of props that took the form of Brent’s favorite story characters during reading time and word reading assessments. The consistent routines embedded in the intervention were likely another contributing factor to the increased attention span of both students, which is consistent with previous research demonstrating that individuals with ASD struggle with novel information or change in routines (APA, 2013; Ganz & Flores, 2009; Head et al., 2018).
Although both students received this intervention individually, the intervention may be delivered in small groups as in our other studies (Allor et al., 2020; Allor, Gifford, et al., 2018), which teachers should consider when determining how to increase reading intervention intensity. Additional practice may also be provided by paraprofessionals, peer tutors, or family members and might include rereading of books and playing learning games to reinforce skills and promote transfer. These activities provide examples of how intensity may be increased.
In summary, this study demonstrates a positive relation between participation in a comprehensive reading program (Friends on the Block) and the reading growth of students with ASD and IQ in the ID range. A clear relation between the intervention and sight word reading was demonstrated, and there was descriptive evidence of positive growth on several additional measures of early literacy. The unique and repetitive features of the curriculum make it a promising intervention for students with ASD who may struggle with transitions during academic instruction or novel material. Future research is needed to replicate and expand these findings, particularly with additional measures of decoding, comprehension, and behavior. Longitudinal tracking of students ongoing reading development is also warranted. We are currently conducting a randomized control trial in response to Friends on the Block in which growth on multiple variables is being measured, including decoding, comprehension, and behavior. Additional research analyzing the various features of the intervention is also needed to determine which features are essential and to further refine the intervention.
Footnotes
Declaration of Conflicting Interests
Jill Allor and Stephanie Al Otaiba acknowledge a small financial interest in the Friends on the Block books and curriculum. Any inquiries should be directed to the Office of Research Compliance at Southern Methodist University.
Funding
This research was supported by grant RA324A130102 from the Institute of Education Sciences (IES). The content is solely the responsibility of the authors and does not necessarily represent any views of IES. Carlin Conner received support from a Postdoctoral Research Fellow Training Program Grant from the National Center for Special Education research at the Institute of Education Sciences (R324B180034) during the completion of this manuscript.
