Word Study Intervention for Students with ASD: A MultipleBaseline Study of Data-Based Individualization
Michael Solı́sUniversity of California Riverside
Farah El ZeinBrescia University College
Marie Black, Alexandra Miller, William J. Therrien, and Marcia InvernizziUniversity of Virginia
Abstract: This multiple baseline across participants study examined the efficacy of a data-based individual-ization word study intervention for students with autism spectrum disorder (N � 5) and low word readingskills. An experienced interventionist provided 1:1 word reading instruction in 30-minute sessions five times perweek for an average of 10 sessions per participant. Intervention effects for directly taught words and words withsimilar spelling patterns were estimated using visual analysis and calculation of mean differences acrossbaseline and intervention phases. Results indicate immediate and consistent improvements in word readingoutcomes across all participants.
Recent reports suggest that as many as one inthree students with autism spectrum disorder(ASD) have reader profiles of low decodingskills and comprehension (Nation, Clarke,Wright, & Williams, 2006; Wei, Christiano, Yu,Wagner & Spiker, 2014). Nation et al. (2006)reported 42% of students with ASD in theirsample as having low decoding skills in addi-tion to low reading comprehension. More re-cently, Wei et al. (2014) reported a similarfinding of 32% of students with ASD havinglow decoding. Despite this being the case,reading interventions for students with ASDover the last 10 years have primarily focusedon reading comprehension (Chiang & Lin,2007; El Zein, Solis, Vaughn, & McCulley,2014; Whalon, Al Otaiba, & Delano, 2009).The overwhelming focus on comprehension
instruction is not surprising considering theinfluential reader profile studies conductedover several decades which described averageor above average decoding skills with low com-prehension for students with ASD (Frith &Snowling, 1983; Goldberg, 1987; Minshew,Goldstein, Taylor, & Siegel, 1994; O’Connor& Hermelin, 1994; Pattie & Lupinetti, 1993).While there does appear to be agreementamong many researchers on the pervasive dif-ficulties students with ASD experience withreading comprehension (Fluery et al., 2014),it is also important to note the needs of stu-dents with ASD that do not fit this typicalprofile. Beyond comprehension needs, it isalso important to investigate other challengesfaced by students with ASD with less typicalreader profiles such as those who face chal-lenges with decoding and word reading.
Code-Based Reading Intervention Research forStudents with Autism Spectrum Disorders
Whalon and colleagues (2009) conducted asynthesis of reading interventions for studentswith ASD that identified eleven studies provid-ing instruction in decoding, fluency, vocabu-lary, and comprehension. Of these 11 studies,only three pre/posttest design and one single-case design study investigated code-based in-
The authors would like to thank Hannah Mathewsfor her hard work on materials development and datacollection. This research was supported by the Instituteof Education Sciences, U.S. Department of Education,through Grant R324A160299 to the University of Cal-ifornia Riverside. The opinions expressed are those ofthe authors and do not represent the views of theInstitute or the U.S. Department of Education. Corre-spondence concerning this article should be ad-dressed to Michael Solis, Graduate School of Edu-cation, University of California Riverside, Riverside,CA 92521. E-mail: [email protected]
terventions (Basil & Reyes, 2003; Coleman-Martin, Heller, Chihak, & Irvine, 2005;Heimann, Nelson, Tjus, & Gilberg, 1995; Tjus,Heimann, & Nelson, 1998). Three of the stud-ies used modified versions of the same com-puter assisted instructional software program(Basil & Reyes, 2003; Heimann et al., 1995;Tjus et al., 1998). The software program pro-vides opportunities for students to selectwords and word groups to form sentences withtasks over time requiring increasingly morecomplex grammatical structures. Across thesestudies, the findings showed some promise yetwere inconclusive due to a number of meth-odological issues including small sample sizeswith large age ranges, unclear screening proce-dures, and lack of comparison or control condi-tions.
Basil and Reyes (2003) investigated the soft-ware program along with a scaffolding ap-proach in a pre/posttest design. Two studentswith ASD, ages 8 and 14, received 12 hours ofinstruction in 30 min. sessions twice per week.While both students “mastered” the lessonsaccording to the software-based proximalmeasure, only one student made significantgains from pre to post-test on measures ofphonological awareness (PA) and word spell-ing. Heimann et al. (1995) investigated thesoftware program with a larger sample size ofstudents with ASD (N � 11) ranging in agefrom 6 to 14 years old. Significant gains inreading was reported, however, no significantgains were detected on PA measures. Tjus etal. (1998) also reported mixed results for stu-dents with ASD with significant gains on PAand reading measures reported on initial post-tests but no significance differences found ondelayed post-test measures.
Utilizing a multiple condition design withdrop-down baselines, Coleman-Martin et al.(2005) investigated a computer-assisted in-structional (CAI) approach using a nonverbalreading approach for word identification forone student with a dual diagnosis of ASD andmoderate intellectual disability. Researcherscompared baseline phases to teacher only,teacher plus CAI, and CAI only. The student,‘Carrie’, had low mean scores on percent ofwords identified correctly (M � 13.3%) dur-ing baseline. The remaining baseline phasesstayed consistently low (M � 0.0%). Acrossthe three treatment conditions the teacher
only phase had the steepest slope with a clearupward trend ending with Carrie scoring 80%correct on three of her last four scores. Ini-tially during the teacher plus CAI conditionthe scores dropped to 60% correct. However,at the end of this phase Carrie scored 100%correct on her last two scores. During the CAIonly phase, scores ranged from 60% to 80%correct with the exception of a few outliers.
Following the synthesis by Whalon et al.(2009) some additional studies were con-ducted that investigated word study interven-tions for students with ASD. Infantino andHempenstall (2006) investigated a direct in-struction intervention of word reading thatused standardized protocols for the basis ofinstruction (Infantino & Hempenstall, 2006).After 23 hours of instruction, one student withASD did not show significant gains in decod-ing skills with no differences in percentilerank, age and grade equivalents. In a multiplebaseline across probes design, Whitcomb,Bass, and Luiselli (2011) reported improvedaccuracy with word lists after one student withASD participated in five intervention sessionsprovided through computer assisted instruc-tion. Similarly, in a multiple baseline acrossprobes study, Yaw et al. (2011) reported animmediate increase in sight word readingacross 16 intervention sessions delivered toone grade 6 student through computer as-sisted instruction.
Rationale and Hypotheses
Initial data on reading performance for thisstudy were taken from a larger observationstudy investigating issues of reading instruc-tion for students with ASD (Solis, Black, Ro-mig, & Miller, in development). The districtpersonnel informed us of their concern withstudents’ word reading skills rather than read-ing comprehension, which were further con-firmed from our testing results. We thenworked with district personnel to design astudy that would integrate in with their cur-rent efforts to address the problem ratherthan supplant any attempts at intervention.Utilizing the techniques described by theNational Center on Intensive Interventions(2013), we adapted the word study readingprogram that was in use by the school district,Words Their Way (Bear, Invernizzi, Templeton,
288 / Education and Training in Autism and Developmental Disabilities-September 2018
& Johnston, 2016), by infusing explicit in-struction and corrective feedback through adata-based individualization (DBI) approachto intervention. The Words Their Way DBI(WW-DBI) approach was designed to addressthe most intense instructional needs for stu-dents who have not responded to previousinterventions (National Center on IntensiveInterventions, 2013).
The purpose of this study was to determineif adaptations of the district adopted wordstudy curricula would increase performanceword study skills for words directly taught andfor words that followed the same spelling pat-terns. The study addressed the following re-search hypotheses: The WW-DBI interventionwill result in improved performance on wordsdirectly compared to baseline performance.The WW-DBI intervention will result in im-proved performance on words not taught butthat followed the same word patterns as thosetaught compared to baseline performance.
Method
Setting and Instructor
All participants were from one rural schoollocated in a southeastern state. According todata from the State Department of Education,the racial and ethnic population of studentsin the district at the time of the study includedthe following: Caucasian, 73.9%; African Amer-ican, 13.0%; Hispanic, 6.90%; two or moreraces, 5.0%; Asian, 0.50%; Native American,0.30%. All intervention sessions were conductedin a small room adjacent to the school library.No other students were present during the in-tervention. Sessions were held during students’intervention or resource period and scheduledfor 30 minutes per day, five days per week.
The instructor was a retired special educa-tor who previously worked at the school dis-trict for 30 years. She had extensive back-ground in teaching reading to children withdisabilities. She was hired, trained, and super-vised by the research team.
Participants
Students. Four male students and one fe-male student with ASD (as identified by theschool district through the multi-disciplinary
team process) in grades 5–7 participated inthe study (N � 5). All students were Cauca-sian. According to school district personnel,students all received a combination of generaleducation and special education classes withminimal behavior supports provided as needed.Parent consent and student assent were ac-quired for all participants as approved by theInstitutional Review Board from the universityof the first author. As part of the screeningprocedure, participants were administered twostandardized reading measures: the Test ofSentence Reading Comprehension (TOSREC)(Wagner, Torgesen, Rashotte, & Pearson, 2010)and the Test of Word Reading Efficiency(TOWRE) (Torgesen, Wagner, & Rashotte,1999). The Kaufmann Brief Intelligence TestVerbal (KBIT-2) (Kaufman & Kaufman, 2004)was also administered to provide descriptivedata. To qualify for the study, students neededto meet the following criteria: TOSREC stan-dard score � 85, TOWRE standard score � 80.To determine the appropriate starting point forintervention, each students was administeredthe spelling inventory placement test fromWords Their Way to identify their current wordpart knowledge. See Table 1 for a summaryof participant information including ageand grade placement.
Materials
Individualized pool of unknown words. Weadapted the procedures described by Ferkis,Belfiore, and Skinner (1997) to guide the de-velopment of the word lists. Using words se-lected from Words Their Way (Bear et al.,2016), we developed a pretest probe to deter-mine unknown words for each participant.Prior to administration of the pretest, thewords were leveled for frequency using theCorpus of Contemporary American English(COCA; http://corpus.byu.edu/coca/) withhigh frequency words being removed. Theseunknown words, in turn, were used to developthe word banks included in the baselineprobes, intervention materials, and interven-tion probes of taught words. The initial wordprobe was delivered on an iPad using stan-dard-sized PowerPoint slides. Each slide con-tained 10 words from a single developmentallevel of the Words Their Way scope andsequence (Bear et al., 2016). Each student’s
level was determined by the Words Their Wayspelling inventory placement test (Bear et al.,2016) (see Table 1), and each student’s scopeand sequence was individualized accordingly.
Scope and sequence development. In order todevelop each student’s individualized scopeand sequence, we compiled and sorted allunknown words by spelling feature. Using theprinciples of word study and minimally dis-criminant pairs, we developed weekly wordsorts that followed the developmental spellingsequence established in Words Their Way (Bearet al., 2016). Each sort contained 15 words, 10words established through the pre-assessmentas unknown to the student, and a set of fivegeneralization words following the same spell-ing pattern. This scope and sequence was usedto develop the lessons, baseline and interven-tion probes, and curriculum-based measures.
Measures
Screening and descriptive measures. Threestandardized measures were administered tostudents prior to baseline data collection. TheTest of Sentence Reading Comprehension(TOSREC), Test of Word Reading Efficiency–Second Edition (TOWRE–2), and the Kauf-man Brief Intelligence Test – Second Edition(KBIT-2).
Test of Silent Reading Efficiency and Compre-hension (TOSREC; Wagner et al., 2010). TheTOSREC is a 3 min, group-administered as-sessment of reading fluency and comprehen-sion. Students are presented with a series ofshort sentences and asked to read silently andassess whether the sentences are true or false.
Average alternate-form coefficients range from0.84 to 0.95.
The Test of Word Reading Efficiency (TOWRE:Torgesen et al. 1999). The TOWRE consists oftwo individually administered 45-second sub-tests of sight word reading and phonemic de-coding efficiency. Each list of words and non-words starts with the least-difficult items andgradually increases in difficulty. The alternate-forms reliability coefficients were reported as0.91 to 0.97 (Torgesen et al., 1999).
Kaufman Brief Intelligence Test–Second Edition(KBIT-2; Kaufman & Kaufman, 2004). TheKBIT-2 is individually administered in approx-imately 15 min; it assesses both verbal andnonverbal ability in people from 4 through 90years of age. The KBIT-2 is composed of twoseparate scales. The Verbal Scale contains twokinds of items---Verbal Knowledge and Rid-dles---both of which assess crystallized ability(knowledge of words and their meanings).The items cover both receptive and expressivevocabulary, and they do not require readingor spelling. Composite internal consistency re-liabilities were ranged from 0.89 to 0.96. Va-lidity studies yielded moderate to high corre-lations with both construct and concurrentvalidity studies (Kaufman & Kaufman, 2004).
Dependent measures. Based on the individu-alized pool of unknown words, daily 20-wordprobes were administered during the baselinephase. During the intervention phase, eachprobe contained 10 words that were directlytaught as a part of the word sort, five wordsthat were not included in the word sort butfollowed the same spelling patterns as those inthe sorting activity, and five distractor words
TABLE 1
Participant Information
Participant Age (years) Grade KBIT Verbal* TOSREC* TOWRE* WWSI
Edward 11 6 79 68 77 Late syllable with affixesTony 13 7 78 76 76 Early within wordRichard 13 6 87 85 63 Late within wordJulia 12 6 66 61 54 Middle letter nameJustin 10 5 80 64 77 Early within word
Note: * � Reported as standard scores; KBIT � Kaufman Brief Intelligence Test; TOSREC � Test of SentenceReading Efficiency and Comprehension; TOWRE � Test of Word Reading Efficiency; WWSI � Words TheirWay Spelling Inventory placement information.
290 / Education and Training in Autism and Developmental Disabilities-September 2018
used to control for cueing during the probes.The daily probes contained two dependentmeasures. The first measure was students’identification of the 10 explicitly taughtwords. These words served as a measureof students’ acquisition of directly taughtwords. The second measure consisted of fivewords not directly taught. These wordsserved as a measure of students’ ability togeneralize taught spelling patterns to novel,unknown words.
Probes were delivered using a single wordlist each day. Directly taught words remainedthe same for the 5-day instructional cycle, gen-eralization words varied across days of theweek. When recording word reading accuracy,the tutor allowed 5 seconds per word for a stu-dent response and allowed for unprompted,self-corrections within those 5 seconds. Onepoint was awarded for each correctly pro-nounced word. See Figure 1 for an annotatedexample of a daily probe.
Procedure
Tutor training. We trained one tutor, a re-tired special educator with up-to-date teachercertification, in two 2-hour sessions. The firstday of training consisted of reviewing the prin-ciples of word study (Bear et al., 2016), theprinciples of explicit instruction (Archer &Hughes, 2011), and error correction proce-dures established by the first author of thestudy. The second day of training consisted of
detailing the specific instructional routineand materials and having the tutor conductmock sessions of the intervention until shedemonstrated mastery of the instructional se-quence. Throughout the training, the tutorwas encouraged to balance following the in-structional routines along with maintainingsome instructional flexibility which is a corefeature built into Words Their Way (Bear et al.,2016).
Intervention. The word study interventionWW-DBI consisted of daily teacher-directedword sorts and guided practice in word fea-ture analysis. The intervention was an adaptedversion of the word study instructional routinedetailed by Bear et al. (2016). To better alignto the needs of students with ASD, we madethe following modifications: instruction wasprovided one to one instead of in groups,increased teacher modeling and guided feed-back (Archer & Hughes, 2011) during explo-ration, additional instructor prompts and scaf-folds during guided practice, and increasedintervention duration. The tutor used the cur-rent behavior management system of positivereinforcement as outlined on each student’sindividualized education plan.
Students met individually with the tutor forfive days per week for 15- to 20-minutes ofinstruction (30 min with probe administra-tion). Words used in each sort were 50%known and/or high frequency words and 50%unknown to the student. During instruction,the following 7-step process was followed.
Figure 1. Teacher copy of the daily word probe.
Word Study Intervention and ASD / 291
(1) The instructor delivered a goal statementfor the lesson, and after session one, a reviewwas provided on the previous day’s word sort.(2) The tutor reviewed the new sort by estab-lishing word categories and reviewing thewords’ pronunciations and meanings. (3) Thetutor demonstrated how to sort the words bysound and repeated the process for sorting bysight or spelling patterns. (4) Using a teacher-directed, closed sort, the tutor guided the stu-dent through the process of sorting the wordsunder the given headers, working through thewords one at a time. Using the sentenceframes provided as a meta-cognitive guide,the instructor introduced each word andprompted discussion about the words regard-ing position of spelling features within theword, frequency, and related words. (5) Thetutor gradually released this responsibility tothe students, providing praise and correctivefeedback. (6) The students led the sort (inde-pendent practice) while the tutor promptedthe student to say the words and comparethem as they sorted. (7) Following completionof the sort, the tutor guided the studentsthrough a reflection process. During the re-flection process, the tutor prompted the stu-
dent to compare and contrast the words andtheir features. Sentence frames such as “I no-tice that . . .” and “When I read these words, Ihear . . .” were used to promote students’ useof feature analysis. To encourage students tothink about their rationale for sorting wordsand making connections to the words’ sounds,spellings, and meanings, the tutor asked thestudent to explain why they sorted the wordsas they had. This explanation of students’thinking was recorded on a flip chart to beused during the following day’s lesson.
Procedural Integrity and Reliability
All intervention and assessment sessions wereaudio-recorded. We used these recordings tocheck a random sample of 20% of the inter-vention sessions for procedural integrity andreliability. From the audio recordings, a re-searcher familiar with the intervention usedan implementation validity checklist of thecore instructional steps of the intervention todetermine the percent of expected instruc-tional steps. See Figure 2 for an example oftreatment integrity protocol. The integrity
Figure 2. Fidelity checklist for day one of intervention instruction.
292 / Education and Training in Autism and Developmental Disabilities-September 2018
and reliability of implementation was 82% forthe coded sessions.
Reliability of assessment data was assessedfor 28% of the sessions by rescoring the sessions’dependent measures and comparing the scoresobtained to the scores reported by the tutor. Atotal of 35 sessions were dual coded by tworaters. Interobserver agreement was then deter-mined by calculating the total number of agree-ments divided by the total number of agree-ments and disagreements multiplied by 100.The mean agreement across observers was96.1%.
Experimental Design and Data Analysis
A single-case multiple baseline design acrossparticipants was used to evaluate the effects ofthe intervention on participants’ abilities toread words in isolation. The advantage of amultiple baseline design, unlike a reversal de-sign, is it allows for the empirical examinationof dependent measures (i.e., word reading)that do not reverse upon removal of the inter-vention (Tawney & Gast, 1984). Furthermore,the sequential implementation of the inde-pendent variable parallels the practices ofteachers and generalization of the behaviorchange is monitored through the design (Gast,Lloyd, & Ledford, 2014).
Researchers have traditionally used the vi-sual analysis method to interpret single casestudy results (Kratochwill et al., 2010). Cur-rently, there is not consensus regarding onestatistical analysis procedure to interpret re-sults from single-case design studies (Kratoch-will et al., 2010). For these reasons, we ana-lyzed data based on visual inspection of thegraph for each participant. Visual inspectionof the graphs was based on the (a) level, (b)trend, (c) variability, (d) overlap, (e) immedi-acy of the effect, and (f) consistency of datapatterns across similar phases (Kratochwillet al., 2010).
Results
Edward
Edward’s word reading data are presented inFigure 3. During baseline, Edward’s scores ondaily reading probes ranged from 20% to 60%with a mean of 45%. Upon implementation of
WW-DBI intervention, Edward’s level of taughtand untaught word reading accuracy increasedto means of 86% and 68% respectively. Scoresfor taught words ranged from 60% to 100%,and scores for untaught words ranged from0% to 100% (See Table 2). Upon introduc-tion of the intervention, an immediate increasein Edward’s score is observed for taught wordsand a gradual increase is observed for untaughtwords. Visual inspection of the graph reveals anupward trend for both taught and untaughtwords during the intervention phase. Calcula-tion of overlapping data between interventionand baseline phases shows 5% overlap fortaught words and 58% overlap for untaughtwords.
Tony
Tony’s word reading data are presented inFigure 3. Tony’s scores on daily readingprobes ranged from 20% to 50% during base-line phase, with a mean of 38%. Upon imple-mentation of WW-DBI intervention, Tony’slevel of taught and untaught word readingaccuracy increased to means of 98% and 71%respectively. Scores for taught words rangedfrom 90% to 100%, and scores for untaughtwords ranged from 40% to 100% (See Table2). Upon introduction of the intervention, animmediate increase in Tony’s score is ob-served for both taught and untaught words.Visual inspection of the graph reveals an up-ward trend for both taught and untaughtwords during the intervention phase. Calcula-tion of overlapping data between interventionand baseline phases shows 0% overlap fortaught words and 8% overlap for untaughtwords.
Richard
Richard’s word reading data are presented inFigure 3. During baseline, Richard’s scores ondaily reading probes ranged from 15% to 55%with a mean of 29%. Upon implementationof WW-DBI intervention, Richard’s level oftaught and untaught word reading accuracyincreased to means of 95% and 74% respec-tively. Scores for taught words ranged from70% to 100%, and scores for untaught wordsranged from 40% to 100% (See Table 2).Similar to Edward’s results, upon introduction
Word Study Intervention and ASD / 293
of the intervention phase, an immediate in-crease in Richard’s score is observed fortaught words and a gradual increase for un-
taught words. Visual inspection of the graphreveals an upward trend for both taught anduntaught words during the intervention
Figure 3. Word reading accuracy reported as percentage of words read correctly.
294 / Education and Training in Autism and Developmental Disabilities-September 2018
phase. Calculation of overlapping data be-tween intervention and baseline phases shows0% overlap for taught words and 10% overlapfor untaught words.
Julia
Julia’s word reading data are presented inFigure 3. During baseline, Julia’s scores ondaily reading probes ranged from 5% to 60%with a mean of 31%. Upon implementation ofWW-DBI intervention, Julia’s level of taughtword reading accuracy increased to a mean of70%, and her untaught word reading accuracyincreased to a mean of 37%. Scores for taughtwords ranged from 40% to 90%, and scoresfor untaught words ranged from 20% to 60%.Upon introduction of the intervention, agradual increase then drop in Julia’s score isobserved for both taught and untaught words.Visual inspection of the graph reveals a neu-tral trend for both taught and untaught wordsduring the intervention phase. Calculation ofoverlapping data between intervention andbaseline phases shows 43% overlap for taughtwords and 100% overlap for untaught words.
Justin
Justin’s word reading data are presented inFigure 3. During baseline, Justin’s scores ondaily reading probes ranged from 25% to 80%with mean of 58%. Upon implementation ofWW-DBI intervention, Justin’s level of taughtand untaught word reading accuracy in-creased to means of 87% and 93% respec-tively. Scores for taught and untaught words
ranged from 80% to 100%. Upon introduc-tion of the intervention, a gradual increase inJustin’s score is observed for taught and un-taught words. Visual inspection of the graphreveals an upward trend for untaught wordsand a neutral trend for taught words. Calcu-lation of overlapping data between interven-tion and baseline phases shows 67% overlapfor taught words and 33% overlap for un-taught words.
Discussion
The purpose of this study was to examine theimpact of an adapted version of a widely usedword study program on the word reading skillsof students with ASD. Our aim was to examinethe effectiveness of the intervention of wordsdirectly taught and words not directly taughtthat followed the same spelling patterns whenimplemented by a tutor hired, trained, andsupervised by researchers. We hypothesizedthat with adaptations to the program, this ap-proach is likely to improve performance forstudents with a history of very low perfor-mance with word reading (TOWRE, standardscore M � 69.4).
Effectiveness of Intervention
Results indicate that the WW-DBI interventionwas generally effective for improving perfor-mance with words directly taught and moder-ately effective with untaught words for stu-dents with ASD with low word reading skills.These gains in word reading skills are similarto findings reported in previous work (Cole-man-Martin et al., 2005; Infantino & Hempen-stall, 2006; Whitcomb et al., 2011; Yaw et al.,2011). During the baseline condition, stu-dents on average read 45.8% words correctly.During the intervention condition, studentsread an average of 87.2% words directlytaught correct and an average of 68.6% ofwords not directly taught. All five participantshad mean scores of words taught well abovetheir baseline performance, and four of thefive participants had mean scores of words notdirectly taught above their baseline perfor-mance.
There was a clear immediacy effect frombaseline to intervention across all five partici-pants. The average baseline score just prior to
Note: Scores represent percentage of words readcorrectly during the daily reading probes.
Word Study Intervention and ASD / 295
beginning intervention was 36% for wordstaught with the average score of words taughtwhen starting intervention was 78%. The dataindicate clear intervention effect at five pointsin time across baseline and intervention witheach phase having three or more data points.In fact, the data “meets standards” as outlinedby Kratochwill et al. (2010) for demonstratingan effect by having four phases with at leastfive data points per phase. Across four of thefive participants the variability was relativelylow for words directly taught with more vari-ability with words not directly taught. It shouldalso be noted that findings from the screeningmeasures indicated much lower performancefor the one participant that did not performsimilarly to the others. Across both dependentmeasures, there was a neutral trend duringthe intervention phase of higher performancecompared to baseline for four of the five par-ticipants. These findings suggest that whendifficult words are targeted and explicitlytaught to students, their performance in com-pleting the task almost doubles. Even morecompelling is evidence that the particular skillbeing instructed did in general transfer towords not directly taught. In other words, theadaptations to the word study program usedby the school district appear to have improvedperformance for all participants.
We expected the procedural integrity to beabove 90% during the sessions. The lowerscore of 82% still indicates that overall theinstructional routines were followed the ma-jority of the time. The lower score may bereflective of the flexibility that was empha-sized during the training. Because of the het-erogeneity and behavioral issues that are oftentypical of students with ASD, we wanted tomake sure that the tutor felt empowered witha certain level of flexibility to address theuniqueness of each student.
Limitations and Implications for Research
There are several limitations to be consideredwhen interpreting the results of this study.First, the tutor was very experienced with boththe word study program that was adapted andwith working students with ASD. The rapportbetween the students and the tutor may haveresulted in higher outcomes and may also ex-plain why little to no behavior management
issues were reported. Future studies shouldconsider use of school-based personnel in-cluding paraprofessionals that would likely beutilized for providing one to one interven-tions. Second, the length of the interventionwas short considering the severity of students’word reading difficulties. To overcome stan-dard scores that are on average 2.0 SD belowthe normative sample on the TOWRE, stu-dents would most likely need intensive inter-ventions of 80 sessions or more (Wanzek et al.,2013). An intervention with longer durationshould also consider adding more distal out-comes measures such as weekly probes andpre/posttest standardized measures. Third,due to time constraints, we were unable tosystematically capture social validity data fromthe students’ perspective. While the tutor didreport that students appeared to enthusiasti-cally participate, not having a social validitymeasure limits the external validity of thestudy. Finally, while we did not have access toassessment data defining the samples with re-spect to executive functioning variables (e.g.,self-regulation, attention, memory), future re-searchers may want to consider how customizinginterventions for students with ASD to align withtheir basic reading processes or executive func-tioning needs might influence students wordreading outcomes.
Implications for Educational Practice
This study provides evidence for the value ofadapting previously developed programs inorder to increase the intervention’s intensityvia smaller group size, lengthening interven-tion sessions, and developing an individual-ized instructional scope and sequence basedon detailed assessment of previous perfor-mance. School personnel should considerthese adaptations as an alternative to the useof separate programs that may incidentallycause confusion because of differences ininstructional approach and the sequence ofskills taught. With systematic and explicitinstruction, students with ASD with low wordreading skills in the upper grades can im-prove their word reading ability. Practitio-ners should keep in mind the intensity nec-essary to fully remediate word readingdeficits and should consider providing moreintensive interventions.
296 / Education and Training in Autism and Developmental Disabilities-September 2018
References
Archer, A. L., & Hughes, C. A. (2011). Explicit in-struction: Effective and efficient teaching. New York,NY: Guilford Press.
Basil, C., & Reyes, S. (2003). Acquisition of literacyskills by children with severe disability. Child Lan-guage Teaching and Therapy, 19, 27–48.
Bear, D. R., Invernizzi, M., Templeton, S., & John-ston, F. (2016). Words their way: Word study forphonics, vocabulary, and spelling instruction. Colum-bus, OH: Pearson.
Chiang, H. M., & Lin, Y. H. (2007). Reading compre-hension instruction for students with Autism spec-trum disorders: A review of the literature. Focus onAutism and Other Developmental Disabilities, 22, 259–267. doi:10.1177/10883576070220040801
Coleman-Martin, M. B., Heller, K. W., Cihak, D. F.,& Irvine, K. L. (2005). Using computer-assistedinstruction and the nonverbal reading approachto teach word identification. Focus on Autism andOther Developmental Disabilities, 20, 80–90.
Corpus of Contemporary American English (n.d.).Retrieved October, 2016, from https://corpus.byu.edu/coca
El Zein, F., Solis, M., Vaughn, S., & McCulley, L.(2014). Reading comprehension interventionsfor students with autism spectrum disorders: Asynthesis of research. Journal of Autism and Devel-opmental Disorders, 44, 1303–1322. doi:10.1007/s10803-013-1989-2
Ferkis, M. A., Belfiore, P. J., & Skinner, C. H. (1997).The effects of response repetitions on sight wordacquisition for students with mild disabilities. Jour-nal of Behavioral Education, 7, 307–324. doi:10.1023/A:1022875506110
Fleury, V. P., Hedges, S., Hume, K., Browder, D. M.,Thompson, J. L., Fallin, K., . . . Vaughn, S. (2014).Addressing the academic needs of adolescentswith autism spectrum disorder in secondary edu-cation. Exceptional Children, 35, 68–79.
Frith, U., & Snowling, M. (1983). Reading formeaning and reading for sound in autistic anddyslexic children. British Journal of DevelopmentalPsychology, 1, 329–342. doi:10.1111/j.2044-835X.1983.tb00906.x
Gast, D. L., Lloyd, B. P., & Ledford, J. R. (2014).Multiple baseline and multiple probe designs. InD. L. Gast & J. R. Ledford (Eds.), Single caseresearch methodology: Applications in special educationand behavioral sciences, (pp. 251–296). New York,NY: Routledge.
Goldberg, T. (1987). On hermetic reading abilities.Journal of Autism and Developmental Disorders, 17,29–44. doi:10.1007/BF01487258
Heimann, M., Nelson, K. E., Tjus, T., & Gillberg, C.(1995). Increasing reading and communicationskills in children with autism through an interac-
tive multimedia computer program. Journal of Au-tism and Developmental Disorders, 25, 459–480.
Infantino, J., & Hempenstall, K. (2006). Effects of adecoding program on a child with Autism Spec-trum Disorder. Australasian Journal of Special Edu-cation, 30, 126–144.
Kaufman, A. S., & Kaufman, N. L. (2004). KaufmanBrief Intelligence Test, Second Edition (K-BIT-2). Min-neapolis, MN: Pearson Assessment.
Kratochwill, T. R., Hitchcock, J., Horner, R. H.,Levin, J. R., Odom, S. L., Rindskopf, D. M., &Shadish, W. R. (2010). Single-case designs techni-cal documentation. Retrieved from What WorksClearinghouse website: https://ies.ed.gov/ncee/wwc/Document/229
Minshew, N. J., Goldstein, G., Taylor, H. G., & Sie-gel, D. J. (1994). Academic achievement in highfunctioning autistic individuals. Journal of Clinicaland Experimental Neuropsychology, 16, 261–270. doi:10.1080/01688639408402637
Nation, K., Clarke, P., White, B., & Williams, C.(2006). Patterns of reading ability in childrenwith autism spectrum disorder. Journal of Autismand Developmental Disorders, 36, 911–919. doi:10.1007/s10803-006-0130-1
National Center on Intensive Intervention. (2013).Data-based individualization: A framework for inten-sive intervention. Washington, DC: Office of Spe-cial Education, U.S. Department of Education.
O’Connor, N., & Hermelin, B. (1994). Two autisticsavant readers. Journal of Autism and DevelopmentalDisorders, 24, 501–515. doi:10.1007/BF02172131
Patti, P. J., & Lupinetti, L. (1993). Brief report:Implications of hyperlexia in an autistic savant.Journal of Autism and Developmental Disorders, 23,397–405.
Solis, M., Black, M., Romig, J., & Miller A. (in prog-ress). Reading instruction for students with ASD:An observation study.
Tawny J. W., & Gast D. L. (1984). Single subjectresearch in special education. Columbus, OH:Merrill.
Tjus, T., Heimann, M., & Nelson, K. E. (1998).Gains in literacy through the use of a speciallydeveloped multimedia computer strategy. Autism,2, 139–156.
Torgesen, J. K., Wagner, R. K., & Rashotte, C. A.(1999). Test of word reading efficiency. Austin, TX:PRO-ED.
Wagner, R. K., Torgesen, J. K., Rashotte, C. A., &Pearson, N. A. (2010). Test of silent reading efficiencyand comprehension. Austin, TX: Pro-Ed.
Wanzek, J., Vaughn, S., Scammacca, N., Metz, K.,Murray, C., Roberts, G., & Danielson, L. (2013).Extensive reading interventions for studentswith reading difficulties after grade 3. Review ofEducational Research, 83, 163–195. doi:10.3102/0034654313477212
Wei, X., Christiano, J. W., Yu, J. W., Wagner, M., &Spiker, D. (2014). Reading and math achievementprofiles and longitudinal growth trajectories of chil-dren with an autism spectrum disorder. Autism, 78,89–106. doi:10.1177/1362361313516549
Whalon, K. J., Al Otabia, S., & Delano, M. E. (2009).Evidence-based reading instruction for individu-als with autism spectrum disorders. Focus on Au-tism and Other Developmental Disabilities, 24, 3–16.doi:10.1177/1088357608328515
Whitcomb, S. A., Bass, J. D., & Luiselli, J. K. (2011).Effects of a computer-based early reading pro-
gram (Headsprout®) on word list and text read-ing skills in a student with autism. Journal of Devel-opmental and Physical Disabilities, 23, 491–499.
Yaw, J. S., Skinner, C. H., Parkhurst, J., Taylor, C. M.,Booher, J., & Chambers, K. (2011). Extendingresearch on a computer-based sight-word readingintervention to a student with autism. Journal ofBehavioral Education, 20, 44–54.
Received: 21 June 2017Initial Acceptance: 16 August 2017Final Acceptance: 17 October 2017
298 / Education and Training in Autism and Developmental Disabilities-September 2018
