To wait in Tier 1 or intervene immediately: a randomized experiment examining first-grade response to intervention in reading.
Despite general support for multitier models, researchers and practitioners have expressed serious concern about the lack of research guidance for implementation. For example, despite the relatively robust evidence for Tier 2 interventions (e.g., Gersten et al., 2009; Wanzek & Vaughn, 2007), to date a fairly limited number of studies have reported effects of multitier intervention at the elementary level that includes what may be termed a Tier 3 intervention (Denton, Fletcher, Anthony, & Francis 2006; Gilbert et al., 2013; O'Connor, Harty, & Fulmer, 2005; Vaughn, Wanzek, Linan-Thompson, & Murray, 2007; Vaughn et al., 2009; Vellutino, Scanlon, Zhang, & Schatschneider, 2008; Wanzek & Vaughn, 2010). From this set of studies, only O'Connor et al. (2005) and Gilbert et al. (2013) allowed students to move up or down tiers within a study year, whereas the remaining provided Tier 3 only after tracking response to Tier 2 for a year or more. Further differences in how students were identified for intervention and how response was defined complicate direct comparisons across studies.
There is also a lack of guidance from a legal and policy perspective, as documented in a review by Zirkel and Thomas (2010), who found marked variability in state laws and guidelines informing local education agencies about how to implement RTI. This variability about RTI procedures, particularly for Tier 3, was further validated by a survey of 40 elementary schools conducted by Mellard, McKnight, and Jordan (2010). The lack of consistency led Vaughn, Denton, and Fletcher (2010) to propose that "schools should consider placing students with the lowest overall initial scores in the most intensive interventions" (p. 442). Given that some students might be in a Tier 2 intervention that does not meet their intensive needs for too long, the authors argued against allowing RTI to become another type of "wait to fail" model, referring to historical criticisms of the IQ--achievement discrepancy model. Vaughn and colleagues argued that immediate intensive interventions may be the most appropriate for some students because it is increasingly possible to predict poor response by students' preintervention scores (e.g., Al Otaiba & Fuchs, 2002; Nelson, Benner, & Gonzalez, 2003) and because it will be very difficult for schools to achieve catch-up growth for children who are persistently weak responders (Al Otaiba & Fuchs, 2006; Denton et al., 2006; Wanzek & Vaughn, 2008).
However, there is a case to be made for waiting to reduce the cost of false positives (providing intervention to students who would have responded to Tier 1). Evidence is accumulating that differentiated or individualized general education reduces the incidence of reading difficulties (Connor et al., in press). Further, Compton and colleagues have conducted an important series of studies to improve classification of which students will need Tier 3: They propose a two-stage screening within Tier 1 that might prevent (a) false positives (students receiving Tier 2 who do not really need it); (b) false negatives (students being missed for Tier 2 who really need it); and, perhaps most important, (c) waiting-to-fail students, who are not likely to respond to Tier 2 and who immediately need the most intensive and extensive interventions (Compton et al., 2010; Compton et al., 2012; D. Fuchs, Fuchs, & Compton, 2012; L. S. Fuchs & Vaughn, 2012). In a study conducted at first grade (D. Fuchs et al., 2012), the model that best predicted who would need special education utilized 6 weeks of progress monitoring in Tier 1 using word identification fluency (D. Fuchs et al., 2004); response to Tier 2 intervention did not add uniquely.
Finally, a recent study experimentally examined the impact of providing to first graders who made inadequate progress to Tier 2 either an additional 7 weeks of Tier 2 or 7 weeks of Tier 3 intervention (identical to Tier 2 intervention but provided in one-to-one tutorials). Gilbert and colleagues (2013) found no significant differences in reading outcomes between these two groups, but 7 weeks is a relatively brief intervention window, and the quality of Tier 1 was not addressed. It was concerning that after following students to the end of third grade, students' standard scores dropped, and only 40% of the students who received the extended Tier 2 intervention and 53% of the students who received the Tier 3 intervention achieved grade-level reading scores (above the 30th percentile). These findings led the authors to call for additional research ensuring that good-quality Tier 1 was implemented and that it experimentally tested the impact of fast-tracking students with the highest risk to Tier 3.
Purpose of the Study
These concerns about RTI implementation guided our large-scale project comparing the efficacy of two RTI models (funded by the Learning Disabilities Center of the National Institute of Child Health and Human Development). The first model, which we termed dynamic RTI, provided students entering first grade with the weakest skills the most intensive early literacy resources; essentially, we fast-tracked students immediately to Tier 2 or Tier 3 intervention depending on their reading skill profile. The second model, which we called typical RTI, was consistent with district policy that required students to begin in Tier 1 and progress through subsequent tiers based on continued weak skills and slow growth. Our typical condition could be considered most similar to Compton and colleagues' two-stage approach. Although we did not monitor weekly progress in Tier 1, we used slope (difference scores) and performance on screeners at the end of each 8-week period to make decisions about intervention sessions. Although many models of RTI are currently in use and some of these models may be similar to our dynamic condition, to our knowledge no similar experiments have been conducted, and research is needed to compare the efficacy of different models. Therefore, we conducted a randomized controlled experiment with matched pairs of students (see Method for details on matching variables) within classrooms assigned to either dynamic or typical RTI to compare the efficacy of the two approaches to answer the following research questions:
1. What are the effects of dynamic RTI and typical RTI on student reading outcomes by the end of first grade?
2. Does assignment to specific tiers predict gains on standardized assessments, and does this differ when comparing dynamic and typical RTI groups?
We contacted the school district in a midsize city in the Southeast in its first year of RTI implementation, which nominated seven schools; all principals and all 34 first-grade general education teachers agreed to participate, allowing us to randomly assign children to conditions within classrooms. One school was a high-performing blue ribbon school serving a fairly high socioeconomic neighborhood (15.8% of students at the school received free and reduced-price lunch); six schools served an economically diverse range of students (participation in free and reduced-price lunch ranged from 42.8% to 89.9%). Few students were limited English proficient (0.4%-2.8%). Prior to conducting the research, we received Institutional Review Board approval as well as approval from the school district's Research Review Board.
Classroom teachers and students. A majority of the teachers (n = 22, 64.7%) were Caucasian; 9 (26.5%) were African American; 1 (3%) was Hispanic; 1 (3%) was Asian; and 1 (3%) was multiracial. Nine teachers held graduate degrees (26.5%), and the rest held bachelor's degrees (77.3%). On average, teachers had taught for about 15 years (M = 14.54, SD = 9.74); seven were relatively new to the profession ([less than or equal to] 5 years); and 15 were veterans (> 15 years). None of the classrooms employed co-teaching, and none of the special education teachers in the schools participated, because special education occurred after Tier 3.
Teachers assisted us in recruiting all their students; subsequently, we received consent from 562 parents (85% of parents agreed). During the study, 35 students moved to schools not participating in the study; these students were from different schools and classrooms and were evenly distributed across conditions. One student exited the study by teacher request (she felt that the child had improved, and she wanted her to receive to Tier 1 only); five students exited after consultation with the school principal because of timing conflicts with special education services that began during the study (all attended the school with the highest socioeconomic status). No students who could be tested were excluded from the study; hence, some children received special education services. Table 1 shows demographics, IQ scores, and descriptive statistics on screeners by initial tier eligibility and by condition.
Typical and Dynamic RTI Conditions
We carefully designed the two RTI conditions to be identical in the following three ways: First, well-trained project staff provided the Tier 2 and 3 interventions, which supplemented classroom teacher-provided Tier 1 reading and language arts instruction; second, students could move up to a more intensive tier if needed and move down to a less intensive tier when they were successful for two screening periods; and, third, the standard protocols for intervention at Tier 2 and 3 (described later) were identical across conditions. Thus, the only difference between conditions was when students were provided supplemental intervention sessions.
The first RTI condition, typical RTI, was designed to mimic two-stage school RTI decision rules recommended by Compton and colleagues and to be consistent with district implementation. In this condition, all children, regardless of their initial screening scores, began the first session in Tier 1. Then, students who demonstrated insufficient response to the Tier 1 instruction at the next screening (8 weeks later) were provided Tier 2 intervention during the second session (decision-making criteria discussed later). Subsequently, students who demonstrated insufficient response to Tier 2 intervention were provided a more intensive Tier 3 intervention during the third session. By contrast, in dynamic RTI, students were provided Tier 2 or Tier 3 according to their initial screening or to criteria at subsequent screenings.
Initial screening for tier eligibility and subsequent decision rules for tier sessions. In fall of first grade, well-trained graduate students administered screening and pretest measures; screening measures were used to determine initial eligibility for Tiers 1-3, and all are described within the Measures section. Teacher ratings of the severity of students' reading difficulties relative to classmates (Speece et al., 2011) and four screeners of letter sound fluency, word reading fluency, and word attack fluency were used to determine initial eligibility (mean scores are shown in Table 1). Because schools had such variable socioeconomic status levels and reading scores, we used school/local norms or cut points to determine eligibility, but we also imposed norm-referenced exclusion criteria (students with word identification and passage comprehension scores above 95 were excluded because they had average reading scores).
Specifically, at the initial screening in September, students whose teachers reported that they had severe reading difficulties and who scored below the 40th percentile at the school level for all four screeners were considered as initially eligible for Tier 3. The initial criterion for Tier 2 eligibility was teacher reports of severe reading difficulties or scoring below the 40th percentile at the school level for three of four screeners. Local norms were used because we had one school with high socioeconomic status; the 40th percentile was used as a cut point for low-average performance. After the first 8-week session, all children (in Tiers 1-3) were again screened to redetermine local norms (this time frame corresponded to the report card period and appeared reasonable to schools for movement across tiers to occur). Students who remained below the 40th percentile on three of four measures and who demonstrated slopes of growth less than the mean for the entire sample moved to a more intensive tier in the next 8-week session (e.g., from Tier 2 to Tier 3 in dynamic RTI or from Tier 1 to Tier 2 for typical RTI). When students were successful (i.e., they scored above the 40th percentile and demonstrated slopes of growth at or above the mean) in a tier for two consecutive 8-week periods, they were exited to a less intensive tier.
Random assignment to condition. To assign students to condition, we calculated z-scores on the screeners, averaged them, rank ordered students within tier eligibility status within classroom, identified adjacent pairs in the ranking, and then randomly assigned one member of the pair to dynamic or typical RTI. Table 1 provides task means and standard deviations based on initial eligibility and condition. Then, after initial assignment, to determine that there were no significant differences across conditions, we conducted chi-square analyses on categorical variables and analyses of variance for continuous variables that revealed no significant differences by condition on gender, ethnicity, free and reduced-price lunch, and special education status. Except for visual disabilities (p = .06) and specific learning disability (SLD; p = .20), p values exceeded .50.
Tier 1 instruction: Classroom teacher training and observations of instruction. In all schools, there was a strong focus on first-grade reading and language arts instruction, which was provided for a minimum of 90 min. All teachers utilized Open Court as the core reading program (Bereiter et ah, 2002). To support teachers' understanding of their role in the project and to help them learn about RTI, we provided a 1-day professional development workshop. We used Gough's simple view of reading (Gough & Tunmer, 1986) as a theoretical framework supporting the need for code- and meaning-focused instruction; we described the research behind three-tier RTI models; explained the importance of individualizing Tier 1; and we provided a rationale for our experiment and our interventions.
To directly assess the quality of Tier 1 classroom instruction, staff used two digital video cameras with wide-angle lenses to videotape the language arts 90-min block in fall and winter (scheduled with teachers' prior knowledge). Then, to address the overall effectiveness of implementation of literacy instruction captured on the videotapes, we used a low-inference observational instrument based on prior research (Al Otaiba, Connor, et ah, 2011; Al Otaiba, Folsom, et ah 2011; adapted from Haager, Gersten, Baker, & Graves, 2003). Our scale ranged from 0 (content not observed) to 3 (1 = not effective, 2 = effective, 3 = highly effective).
Specifically, coders used this scale to rate each component of teachers' reading instruction (phonemic awareness, phonics, fluency, vocabulary, and comprehension), their warmth and classroom management, and the degree to which teachers individualized instruction. During the coding meetings, any disagreements were resolved by the master coder. Interrater reliability was then established on 10% of the tapes, and the mean Cohen's kappas for reliability were high (.975 and .972 for fall and winter, respectively). Averaged across both time points, teachers' overall reading instruction was rated as effective (M = 1.88, SD = 0.36; range, 1-2.69); classroom management ratings approached very effective (M = 2.53, SD = 0.46; range, 1.50-3), as did their warmth ratings (M = 2.51, SD = 0.43; range 1.50-3). Finally, although we trained teachers to individualize instruction by forming small groups and differentiating tasks, their ratings were lower for individualization (M = 1.40, SD = 1.04; range, 0-3).
Code and Meaning Components of Tier 2 and Tier 3 Interventions
Interventions supplemented Tier 1 instruction from October until the beginning of June, with movement across tiers each 8 weeks guided by screening data and condition. Trained research staff served as tutors for the project, who conducted intervention as a pullout in quiet classrooms or areas in the media center. We consulted with the district reading specialist to select evidence-based Tier 2 and Tier 3 that aligned with the existing Tier 1 core reading program. A standard protocol was used for content and sequencing of the lessons, but tutors were allowed to provide flexible pacing across groups to ensure mastery and they were trained to use positive behavior supports.
For Tier 2, students received intervention twice a week for 30 min in groups of four to seven. We selected code-focused activities for Tier 2 from Open Court's first-grade Imagine It! series (Berieter et al., 2002) and the Florida Center for Reading Research's K-3 Center Activities (http://www.fcrr.org). For the first 8 weeks, tutors spent 10 min explicitly teaching phonological awareness and letter sound skills and 10 min on decoding and sight word instruction. In the second 8-week session, as students mastered letter sounds and phonological awareness, relatively more time (15 min) focused on decoding, sight words, and fluency training. In the final 8-week session, virtually all instruction related to decoding, sight words, and fluency. Tutors provided meaning-focused instruction for about 10 min per day. In the first 8 weeks, interventionists read aloud high-interest trade books using dialogic reading techniques (e.g., Lonigan & Whitehurst, 1998). For the second 8 weeks, as students were able to read decodable books, they read Open Court decodable books to practice fluency, and they answered sentence-level comprehension questions. For the final 8 weeks, students read decodable books that we wrote to emphasize the text structure sequencing (i.e., first, next, last). Tutors used graphic organizers to model and guide students in retelling stories and in writing a brief retell of the story.
For Tier 3, students received intervention 4 days a week for 45 min in groups of one to three. The 30-min code-focused portion of Tier 3 consisted of Early Interventions in Reading lessons (Mathes, Torgesen, Wahl, Menchetti, & Grek, 1999; Mathes et al., 2005). Each lesson is thoroughly scripted and details exactly how to deliver explicit and systematic intervention that includes phonemic awareness, alphabetics and phonics, and fluency. Finally, tutors provided 15 min of meaning-focused activities that were identical to Tier 2, and they ensured that the decodable books were appropriate to the instructional level of Tier 3 group members.
Tutors, Training, and Treatment Fidelity
Intervention tutors included certified teachers and graduate students in special education who were research assistants. During the project, we took several steps to support implementation fidelity. The first author and three senior project staff trained tutors initially across two 8-hr days to orient them to the project and to provide training on interventions and positive behavioral supports. We provided coaching and modeling at school sites as needed and conducted weekly meetings with tutors to examine lesson plans and discuss student progress. Tutors were observed by senior staff, and they videotaped their own intervention every 8 weeks; tutors observed their videotapes and debriefed with one of two senior staff who coded the tapes for fidelity of implementation. At the end of each 8-week period, prior to each time that students were moving up or down tiers, meetings focused on orienting tutors to new activities and, if students were changing tiers or intervention group, allowing tutors to share information about their students (including praise and behavioral strategies that worked or favorite types of books).
Fidelity of Tier 2 and Tier 3 implementation. The videotapes of Tiers 2 and 3 were scored by a master coder and a senior staff member using a fidelity checklist (with a Likert scale rating of 1 = poor, 2 = good, 3 = excellent). Interrater reliability of 98.1% was established. In addition to recording whether the lesson/script for each activity was followed accurately and for the appropriate time, the checklist asked the following: "Was the timing accurate? Was the pacing appropriate? Did students master the lesson? Were errors corrected? Were students attentive?" Fidelity ratings for the tutors ranged from .77 to .98 (M = .89).
Data Collection and Measures
All students were individually assessed by trained research staff in quiet areas near their classrooms. Prior to testing, we required staff to reach 98% accuracy on a checklist evaluating their accuracy at following administration and scoring procedures. Staff checked protocols to ensure accurate scoring and for relevant subtests and derived standard and W scores (Rasch ability scores that provide equal-interval measurement characteristics centered at 500). Due to the size and complexity of the project, all staff members were aware of assignment to condition. Under more ideal circumstances, assessors would be blind to condition; therefore, we took steps throughout the project to remind staff that the purpose of the study was to learn which condition was more effective and to caution them that experimenter bias could undermine an otherwise very carefully planned study (e.g., Rosenthal & Rosnow, 1984).
Screening and progress-monitoring measures. Five assessments were used to screen students' skills and to monitor their progress, which included the Teacher Rating of Reading Problems (Speece & Case, 2001; Speece et al., 2011) and four screening assessments that were individually administered in less than 5 min per student before each 8-week intervention session and at posttest. In September, teachers ranked the overall reading ability of students, using a 5-point Likert scale on the Teacher Rating of Reading Problems (Speece & Case, 2001; Speece et ah, 2011): 5 = well above grade level, 4 = above grade level, 3 = on grade level--no additional support needed, 2 = below grade level-additional support needed, 1 = well below grade level--intensive support needed. Concurrent validity with standardized timed and untimed measures of reading ranged from .61 to .69. Students' letter sound fluency was assessed with the 1-min AIMSWeb Letter Sound Fluency (Shinn & Shinn, 2004). In this task, children are presented an array of 10 rows of 10 lowercase letters per line and are asked to name as quickly as they can as many of the sounds that the letters make. Testing is discontinued if no correct sounds occur in the first row. Raw scores are reported, and alternate form reliability is .90. Next, we used a 1-min criterion-referenced curriculum-based measure: the Word Identification Fluency task (D. Fuchs et ah, 2004). In this task, students read from an array of 50 first-grade sight words (randomly selected from the Dolch word list of 100 frequent words). Raw scores are reported, and alternate form reliability was reported to be .97 (D. Fuchs et ah, 2004). Then, students' word reading fluency was assessed with the two 45-s subtests of the standardized Test of Word Reading Efficiency (Torgesen, Wagner, & Rashotte, 1999). The Sight Word Efficiency and Phoneme Decoding Efficiency subtests require students to read sight words and nonsense words, respectively, on a list of increasing difficulty. Test-retest reliability is .90.
Additional reading assessments. We selected three norm-referenced subtests of the Woodcock Johnson-III (Woodcock, McGrew, & Mather, 2001) to assess word reading, decoding and comprehension; the test manual reports reliability of .91, .94, and .83, respectively. The Letter Word Identification subtest requires students to identify letters in large type and then to read increasingly difficult words. The Word Attack subtest initially requires students to identify the sounds of a few single letters and progress to decoding items of increasingly complex letter combinations that follow regular patterns in English orthography but are nonwords. Reading comprehension was assessed with the norm-referenced Passage Comprehension subtest, a cloze task that requires students to identify a missing key word that is consistent with the context of a written passage. Initially, examiners read the sentence, and the items have pictures, but later, students must read the sentence or passage and identify the missing key word.
Finally, we assessed oral reading fluency as an outcome posttest using the Dynamic Indicators of Basic Early Literacy Skills (Good & Kaminski, 2002) Oral Reading Fluency measure. Students read a previously unseen grade-level passage, and the score is the number of correct words per min. Test-retest reliability for elementary students is .92.
Research Question I
What are the effects of dynamic RTI and typical RTI on student reading outcomes? The first research question compared the effects of the two RTI conditions on student reading outcomes. We first examined the multiple outcome measures using principal component analysis to avoid multiple analyses. Results revealed that the assessments loaded on one factor, and so factor scores were created: one for the fall assessments and one for the spring assessments (see Table 2). These variables were used in subsequent analyses (fall reading and spring reading, respectively). There were no significant differences in fall reading factor scores for students in dynamic RTI compared to the typical RTI groups, t(553) = .981, p = .327, 95% confidence interval [-.083, .249], Next, we examined the overall intent-to-treat effect using hierarchical linear modeling (HLM), because children were nested in classrooms and HLM accommodates such data (Raudenbush & Bryk, 2002). The unconditional model with the spring reading factor score revealed an intraclass correlation of .250 (i.e., proportion of between-classroom variance). Results of the analyses (see Table 3) revealed that students in the dynamic RTI group had statistically significantly higher spring reading scores than did students in the typical RTI group, with an effect size of .314 (Hedges's g; http://www.polyu.edu.hk/mm/ effectsizefaqs/calculator/calculator.html), which is a moderate effect (C. J. Hill, Bloom, Black, & Lipsey, 2008) with practical importance (What Works Clearinghouse; http://ies .ed.gov/ncee/wwc/).
Research Question 2
Does assignment to specific tiers predict gains on standardized assessments, and does this differ when comparing dynamic and typical RTI groups? The second research question explored whether assignment to tiers was related to reading growth on a standardized assessment. We used the spring Woodcock Johnson-III Brief Reading W score as the outcome so that we could examine growth in scores over the school year, which was not possible with the reading factor scores. There were no significant between-group differences in fall Brief Reading W scores, t(554) = 1.387, p = .166, confidence interval [-1.391, 8.081], We then examined spring Brief Reading W scores as a function of the tier that students were initially eligible for, utilizing latent growth curve HLM (Raudenbush & Bryk, 2002). We entered the variable tier (1, 2, or 3) as a time-varying covariate at Level 1, where time was in months and centered at the end of the year. We did this because students moved from tier to tier across the three sessions during first grade (see Table 4 and Figure 1). Condition was entered at Level 2: the student level. Results revealed no significant differences by condition in growth over time for students who were eligible for Tier 1-only instruction (blue solid and dotted lines). This finding is expected given that children were randomly assigned to condition within classrooms; thus, students in both conditions who were eligible for Tier 1 only received instruction by their classroom teacher. Students who received Tier 2 interventions had lower scores than did students who were eligible for only Tier 1 instruction. Tier 2 students in the dynamic RTI condition had significantly higher reading outcome scores when compared to students initially eligible for Tier 2 in the typical RTI condition, who by design received Tier 2 if they did not respond to Tier 1 over the first or second session. Not surprising, students initially eligible for Tier 3 had the weakest scores over the school year, but students initially eligible for Tier 3 who received immediately it because they were in dynamic RTI achieved higher Brief Reading scores when compared to Tier 3 students in typical RTI who had to wait until the beginning of Session 3.
To understand the effects of fast-tracking students with the weakest initial skills to the most intensive treatment, we conducted a randomized controlled experiment with matched pairs of students within classrooms assigned to either dynamic or typical RTI, which required students to begin in Tier 1 and which waited for their response. By design, both conditions were identical in that teachers used the same core reading program, which was implemented effectively, and interventionists provided the same Tier 2 and 3 interventions, which were implemented with fidelity; the only difference across conditions was when the supplemental Tier 2 and 3 interventions began. The study was conducted for a full school year and was unique relative to prior RTI investigations in allowing movement across tiers every 8 weeks in tandem with report card periods and in allowing fast-tracking to Tier 3 for the most needy students. As some districts may use versions of RTI that are similar to both conditions, our study may inform the controversy in the field of special education about ensuring that RTI not become another "wait to fail" model (e.g., Denton et al., 2006; D. Fuchs, LS Fuchs & Compton, 2012; Vaughn et al., 2010).
When addressing the first research question--regarding the effects of dynamic RTI and typical RTI on student reading outcomes by the end of first grade--we used an intent-to-treat analysis that accounted for nesting by using HLM. Findings revealed that the dynamic condition was more effective than the typical RTI. The effect size was moderate (ES = .36), which exceeds the .25 recommended by the What Works Clearing House (http://ies.ed.gov/ncee/wwc/) for a substantively important finding that meets study standards for evidence-based practices. We cannot directly compare this effect size with prior studies including Tier 3 interventions, because of differences in initial criteria, definitions of responsiveness, and limited study of movement within a study year. However, this finding extends the prior research (Compton et al., 2012; Gilbert et al., 2013; Vaughn et al., 2009; Vaughn et al, 2010) indicating that multitier models have potential to improve reading growth and that not all students need to go through Tier 1 or Tier 2.
Given that students' performance on standardized reading scores is a hallmark of their reading skills, our second research question examined whether there were interactions between assignment to tiers and condition that were related to performance on these measures. As expected, given screening for tiers, at the onset, students eligible for Tier 3 performed significantly lower than students eligible for Tier 2. It was possible to identify students who needed the most intensive intervention using brief screeners (approximately 3-5 min of assessment) and a Teacher Rating of Reading Severity (Speece & Case, 2001). Also as expected, due to random assignment of students within classrooms to condition, there were no significant differences in outcomes for students within Tier 1 across the two conditions. Further, we carefully documented that Tier 1 was effective, which has not routinely been established in prior research (D. R. Hill, King, Lemons, & Partanen, 2012).
There was a significant interaction between Tier 2 and Tier 3 and condition, indicating that students in the dynamic condition achieved higher Brief Reading skills than students in the typical condition. That students were equal at pretest due to random assignment, that they received the same Tier 1, and that they were given the same interventions in both conditions all add confidence that fast-tracking through the dynamic condition led to significantly higher reading outcomes. It is encouraging that only a small percentage of students ended first grade with standard scores below 91 (the 25th percentile) for word reading (4.30% on Word Attack and 7.20% on Letter Word Identification); however, in terms of their comprehension, 19.40% had standard scores below 91 on passage comprehension. Such variability across measures is consistent with prior research (e.g., Gilbert et al., 2013; O'Connor et ah, 2005; Vaughn et ah, 2009; Vellutino et ah, 2008).
Although the purpose of this article was not to specifically analyze our classification accuracy or to statistically analyze sensitivity and specificity, several indicators suggest that across the year, children who needed intervention received it and that the impact of unnecessary Tier 2 and 3 appeared to be negligible. First, had there been false positives, it would be likely that tiers might not be significantly different, but they were clearly distinguishable on all measures. Furthermore, Table 1 shows that students initially eligible for Tier 3 had standard scores about 1.5 standard deviations lower than peers in Tier 1 and that students in Tier 2 had mean standard scores between a half to a whole standard deviation lower than students in Tier 1. In addition, if false positives were an issue in the dynamic condition, then they would not have screened eligible by the second session, but Figure 2 shows that of 23 students were fasted-tracked to Tier 3, only four responded well enough to be eligible to return to Tier 2 (only by the third intervention session). Of the 58 students fast-tracked to Tier 2 in the dynamic condition, 51 remained eligible for Tier 2 at the second session screening, and the remaining seven were eligible for Tier 3 (through the end of the study); by the third screening session, it was encouraging that 25 of the original 58 students in Tier 2 responded well enough to return to Tier 1. Thus, taken together, it seems unlikely that the students were false positives and that intervention was warranted.
Second, with regard to false negatives (students being missed for Tier 2 or Tier 3 who really need it), as shown in Figure 2, in the dynamic condition, 180 students initially screened as eligible for Tier 1, and of these, only 19 of them scored as needing Tier 2 in the second session, and only three scored as needing Tier 3 in the third session. By contrast, in the typical condition, whereby all students began in Tier 1, by the second intervention session, 68 (26%) screened eligible for Tier 2, and by the third intervention session, 16 screened eligible for Tier 3. Arguably, these 16 children may have waited too long for intervention. Thus, with regard to students waiting too long in Tier 2 in the dynamic condition, of the 58 students who began in Tier 2, 51 again screened eligible for Tier 2 in the second session (and 7 screened eligible for Tier 3); by the third session, an additional nine screened eligible for Tier 3. Thus, our findings extend work by O' Connor et al. (2005) supporting the merit of movement within RTI systems; that is, 8-week periods were feasible to track progress, and fast-tracking is promising, as opposed to forcing students to wait for more intense intervention. We contend that fast-tracking may also be more in line with the very concept of a free and appropriate public education under the Individuals with Disabilities Education Act. However, even fast tracking was not able to consistently close the reading gap; that cautionary note is consistent with all other RTI investigations to date that have included Tier 3 (Denton et al., 2006; Gilbert et al., 2013; O'Connor et al., 2005; Vaughn et al., 2009; Vellutino et al., 2008; Wanzek & Vaughn, 2010).
Implications for Practice and Challenges That Arose
We hope that our lessons learned, including the challenges that we faced, can guide future RTI research and practice. Based on our findings and prior research, there appears converging evidence indicating that it is possible to identify, at the start of first grade, which children will need the most intensive intervention. Further, the present study used an experimental design to show that when students with the weakest initial skills received the most intensive intervention (through the dynamic condition), their reading performance was significantly stronger at the end of the year than students in the typical condition, where students waited for 8 weeks before moving into Tier 2 or 3. The 8-week sessions were feasible for change and related to report card periods. In retrospect, we feel that it was a good decision to add the meaning-focused modules to both Tier 2 and Tier 3, beginning with dialogic reading, reading decodable texts, and reading and responding to decodable texts that we wrote to target the sequencing text structure. Similarly, our decision to use local norms, along with the Teacher Rating of Reading Severity (Speece & Case, 2001; Speece et al., 2011), rather than a standard score to assign children to tiers led to good buy-in from schools and teachers. Indeed, our data were used by schools in their initial RTI meetings, which was helpful to principals in their first implementation year of RTI in the district.
We did face several challenges that schools would likely face as well in conducting this study. Most important, students in Tier 3 did not catch up to students in Tier 1. This finding is consistent with the research including Tier 3 reading interventions; thus, it is likely that they will need continued help that may include special education (Denton et al., 2006; Gilbert et al., 2013; O'Connor et al., 2005; Vaughn et al., 2009; Vellutino et al., 2008). O'Connor et al. (2005) reported that 60% of K-3 students who received intervention and qualified for Tier 3 could not read on grade level at the end of third grade. In addition, a sobering finding from longitudinal research is that when interventions have stopped at the end of first grade, students' standard scores slid such that by the end of third grade, nearly a third of students read below the 30th percentile (Vellutino et al., 2008) or worse (60% and 46% of students who received Tier 2 and Tier 3, respectively; Gilbert et al., 2013).
Second, we believe that schools will need to be prepared to be flexible in RTI implementation. Although we initially conceptualized the interventions as standard protocols so that we could determine that they were completed with fidelity, across the year it was necessary to individualize pacing, to regroup some students due to behavioral or personality issues, to regroup to minimize heterogeneity of skills, and to tailor readability of text to the group need. Further, when it was time to move students from Tier 3 to Tier 2, we wished that we had had a Tier 2.5 because, by then, the Tier 2 groups were so much further ahead in terms of the amount of text that they could independently read as well as the phonic rules and sight words that they had mastered. An important implication is that researchers and schools will likely grapple with using data to ensure that interventions are at the appropriate instructional level, or "sweet spot," for individual students. Another research group has used data-based regrouping but for Tier 2 in kindergarten. Coyne and colleagues (in press; Lentini & Coyne, 2013) used curriculum mastery data to keep groups more homogeneous, to adjust the pace of instruction, to release strong responders to Tier 1, and to enroll weaker students into Tier 2, and they found that students in the regrouping condition outperformed peers in the standard intervention group.
In addition, we had initially thought that Tier 2 would be easier for interventionists than Tier 3. However, given the larger sizes of the group in Tier 2 (with subsequently more heterogeneity of skills) and given that the code-focused aspect of Tier 2 intervention was less scripted than Tier 3, it took more careful planning. Finally, although we worked closely with school leaders during the project, it was very challenging to schedule intervention, particularly moving across sessions.
Limitations and Directions for Future Research
As with most school-based research, there are some potential limitations to our study findings, and some lead to directions for future research. First, as mentioned previously, assessors were not blind to students' conditions, due to the size and complexity of the project. Although we took care to explain to staff that this was a true experiment to learn which condition was most efficacious, future research should use stronger design and engage a separate team of assessors. Second, we did not test for specificity or sensitivity of classification, and additional research is needed using methods such as receiver operator curves. Future research is needed to analyze profiles of these inadequate responders, to track their reading trajectories and special education classification longitudinally, and to learn more about school-delivered interventions.
Third, our findings might differ had we selected students using national rather than local norms or if the movement rules across tiers differed. Clearly, there is a need for future research to establish criteria for responsiveness and movement, even though our criteria appeared feasible and efficient. Fourth, our participants included only beginning first graders, and so findings may not generalize to older students. Additional research is needed for persistently poor readers beyond primary grades. Fifth, our findings may not generalize to schools with weaker or different Tier 1 instruction. All of our teachers used the same systematic core, which we observed to be effectively implemented through observations and as indicated by standard scores for students who participated in Tier 1 only. Thus, research is warranted with different populations and instructional core reading programs.
Sixth, although we considered it a strength of our study that we chose to use a standard treatment protocol and selected Tier 2 and Tier 3 intervention that aligned with the core and was supported by the local school district, future research is needed to examine other interventions and to compare conditions like our dynamic condition to problem-solving protocols, particularly in Tier 3. For example, there has been some promising work with brief experimental analysis (e.g., Bums & Wagner, 2008) to identify optimal interventions for individual children.
In summary, the results of this randomized controlled trial revealed that immediately providing Tier 2 and 3 interventions to students who qualify led to generally stronger reading outcomes by the end of first grade, in contrast to typical RTI, which waited for students to respond to Tier 1 for 8 weeks before providing intervention, thus resulting in the most intensive interventions being delayed. RTI protocols have shown promise in preventing reading difficulties related to inadequate instruction. Nevertheless, there is marked variation in how and when students receive supplemental intervention. Dynamic RTI protocols, such as the one used in this study, suggest that there is no reason to delay intervention, that any effect of false negatives is negligible, and that, broadly implemented, dynamic RTI, including a foundation of effective Tier 1 instruction, can improve reading outcomes for all children.
Al Otaiba, S., Connor, C. M., Folsom, J. S., Greulich, L., Meadows, J., & Li, Z. (2011). Assessment data-informed guidance to individualize kindergarten reading instruction: Findings from a cluster-randomized control field trial. The Elementary School Journal, 111, 535-560. doi:http://dx.doi.org/10.1086/659031
Al Otaiba, S., Connor, C. M., Foorman, B. R., Greulich, L., & Folsom, J. S. (2009). Implementing response to intervention: The synergy of beginning reading instruction and early intervening services. In T. E. Scruggs & M. A. Mastropieri (Eds.), Advances in learning and behavioral difficulties: Vol. 22. Policy and Practice (pp. 291-316). Bingley, England: Emerald.
Al Otaiba, S., Folsom, J. S., Schatschneider, C., Wanzek, J., Greulich, L., Meadows, J., ... Connor, C. M. (2011). Predicting first-grade reading performance from kindergarten response to Tier 1 instruction. Exceptional Children, 77, 453-170.
Al Otaiba, S., & Fuchs, D. (2002). Characteristics of children who are unresponsive to early literacy intervention. A review of the literature. Remedial and Special Education, 23, 300316. doi:http://dx.doi.org/10.1177/074193250 20230050501
Al Otaiba, S., & Fuchs, D. (2006). Who are the young children for whom best practices in reading are ineffective? an experimental and longitudinal study. Journal of Learning Disabilities, 39, 414-431.
Bereiter, C., Brown, A., Campione, J., Carruthers, I., Case, R., Hirshberg, J., ... Treadway, G. H., Jr. (2002). Open court reading. Columbus, OH: SRA/McGraw-Hill.
Bums, M. K., & Wagner, D. (2008). Determining an effective intervention within a brief experimental analysis for reading: A meta-analytic review. School Psychology Review, 37, 126-136.
Compton, D. L., Fuchs, D., Fuchs, L. S., Bouton, B., Gilbert, J. K., Barquero, L. A., ... Crouch, R. C. (2010). Selecting at-risk first-grade readers for early intervention: Eliminating false positives and exploring the promise of a two-stage gated screening process. Journal of Educational Psychology, 102, 327-340. doi:http://dx.doi.org/10.1037/a0018448
Compton, D. L., Gilbert, J. K., Jenkins, J. R., Fuchs, D., Fuchs, L. S., Cho, E., ... Bouton, B. (2012). Accelerating chronically unresponsive children to tier 3 instruction: What level of data is necessary to ensure selection accuracy? Journal of Learning Disabilities, 45, 204-216. doi:http:// dx.doi.org/10.1177/0022219412442151
Connor, C. M., Morrison, F. J., Fishman, B., Crowe, E. C., Al Otaiba, S., & Schatschneider, C. (in press). A longitudinal cluster-randomized control study on the accumulating effects of individualized literacy instruction on students' reading from 1st through 3rd grade. Psychological Science.
Coyne, M. D., Simmons, D. C., Simmons, L. E., Hagan-Burke, S., Kwok, O., Kim, M, ... Rawlinson, D. M. (in press). Adjusting beginning reading intervention based on student performance: An experimental evaluation. Exceptional Children.
Denton, C. A., Fletcher, J. M., Anthony, J. L., & Francis, D. J. (2006). An evaluation of intensive intervention for students with persistent reading difficulties. Journal of Learning Disabilities, 39, 447-466. doi:http://dx.doi.org /10.1177/00222194060390050601
Fuchs, D., Fuchs, L. S., & Compton, D. L. (2004). Identifying reading disabilities by responsiveness-to-instruction: Specifying measures and criteria. Learning Disability Quarterly, 27, 216-227. doi:http://dx.doi. org/10.2307/1593674
Fuchs, D., Fuchs, L. S., & Compton, D. L. (2012). Smart RTI: A next-generation approach to multilevel prevention. Exceptional Children, 78, 263-279.
Fuchs, L. S., & Vaughn, S. (2012). Responsiveness-to-intervention: A decade later. Journal of Learning Disabilities, 45, 195-203. doi:http://dx.doi.org/10.1177/002221941244 2150
Gersten, R., Compton, D., Connor, C. M., Dimino, J., Santoro, L., Linan-Thompson, S., & Tilly, W. D. (2009). Assisting students struggling with reading: Response to intervention (RtI) and multi-tier intervention in the primary grades. Retrieved from http://ies.ed.gov/ncee/ wwc/PracticeGuide.aspx?sid=3
Gilbert, J. K., Compton, D. L., Fuchs, D., Fuchs, L. S., Bouton, B., Barquero, L. A., & Choo, E. (2013). Efficacy of a first-grade responsiveness-to-intervention prevention model for struggling readers. Reading Research Quarterly, 48, 135-154. doi:http://dx.doi .org/10.1002/rrq.45
Good, R. H., & Kaminski, R. A. (Eds.). (2002). Dynamic indicators of basic early literacy skills (6th ed.). Eugene, OR: Institute for Development of Educational Achievement.
Gough, P. B., & Tunmer, W. E. (1986). Decoding, reading, and reading disability. Remedial and Special Education (RASE), 7, 6-10.
Haager, D., Gersten, R., Baker, S., & Graves, A. (2003). The English language learner observation instrument for beginning readers. In S. V. & K. L. Briggs (Eds.), Reading in the classroom: Systems for the observation of teaching and learning (pp. 111-144). Baltimore, MD: Brookes.
Hill, C. J., Bloom, H. S., Black, A. R., & Lipsey, M. W. (2008). Empirical benchmarks for interpreting effect sizes in research. Child Development Perspectives, 2, 172-177. doi:http://dx.doi.org/10.1111/j.1750-8606.2008 .00061.x
Hill, D. R., King, S. A., Lemons, C. J., & Partanen, J. N. (2012). Fidelity of implementation and instructional alignment in response to intervention research. Learning Disabilities Research & Practice, 27, 116-124. doi: 10.1111/j.15405826.2012.00357.x
Lentini, A. R., & Coyne, M. D. (2013). Addressing false positives in early reading assessment using intervention response data. Manuscript submitted for publication.
Lonigan, C. J., & Whitehurst, G J. (1998). Relative efficacy of parent and teacher involvement in a shared-reading intervention for preschool children from low-income backgrounds. Early Childhood Research Quarterly, 13, 263-290.
Mathes, P. G., Denton, C. A., Fletcher, J. M., Anthony, J. L., Francis, D. J., & Schatschneider, C. (2005). The effects of theoretically different instruction and student characteristics on the skills of struggling readers. Reading Research Quarterly, 40, 148-182. doi:http://dx.doi .org/10.1598/RRQ.40.2.2
Mathes, P. G., Torgesen, J. K., Wahl, M., Menchetti, J. C., & Grek, M. L. (1999). Proactive beginning reading: Intensive small group instruction for struggling readers. Dallas TX: Southern Methodist University.
Mellard, D., McKnight, M., & Jordan, J. (2010). RtI tier structures and instructional intensity. Learning Disabilities Research & Practice, 25, 217-225. doi:10.1111/j.15405826.2010.00319.x
Nelson, J. R., Benner, G. J., & Gonzalez, J. (2003). Learner characteristics that influence the treatment effectiveness of early literacy interventions: A meta-analytic review. Learning Disabilities Research & Practice, 18, 255-267. doi:http://dx.doi.org/10.1111/15405826.00080
O'Connor, R. E., Harty, K. R., & Fulmer, D. (2005). Tiers of intervention in kindergarten through third grade. Journal of Learning Disabilities, 38, 532-538. doi: 10.1177/00222 194050380060901
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage.
Rosenthal, R., & Rosnow, R. L. (1984). Applying Hamlet's question to the ethical conduct of research: A conceptual addendum. American Psychologist, 39, 561-563. doi:http://dx.doi. org/10.1037/0003-066X.39.5.561
Shinn, M. M., & Shinn, M. R. (2004). AIMSweb. Eden Prairie, MN: Edformation.
Speece, D. L., & Case, L. P. (2001). Classification in context: An alternative approach to identifying early reading disability. Journal of Educational Psychology, 93, 735-749. doi:http://dx. doi.org/10.1037/0022-0663.93 .4.735
Speece, D. L., Schatschneider, C., Silverman, R., Case, L. P., Cooper, D. H., & Jacobs, D. M. (2011). Identification of reading problems in first grade within a response-to-intervention framework. Elementary School Journal, 111, 585-607.
Torgesen, J. K., Wagner, R., & Rashotte, C. A. (1999). Test of word reading efficiency. Austin, TX: PRO-ED.
Vaughn, S., Denton, C. A., & Fletcher, J. M. (2010). Why intensive interventions are necessary for students with severe reading difficulties. Psychology in the Schools, 47, 432-144.
Vaughn, S., Wanzek, J., Linan-Thompson, S., & Murray, C. S. (2007). Monitoring response to supplemental services for students at risk for reading difficulties: High and low responders. In S. R. Jimerson, M. K. Bums, & A. M. VanDerHeyden (Eds.), Handbook of response to intervention: The science and practice of assessment and intervention (pp. 234-243). New York, NY: Springer, doi: 10.1007/978-0387-49053-3_17
Vaughn, S., Wanzek, J., Murray, C. S., Scammacca, N., Linan-Thompson, S., & Woodruff, A. L. (2009). Response to early reading intervention examining higher and lower responders. Exceptional Children, 75, 165-183.
Vellutino, F. R., Scanlon, D. M., Zhang, H., & Schatschneider, C. (2008). Using response to kindergarten and first grade intervention to identify children at-risk for long term reading disability. Reading and Writing: An Interdisciplinary Journal, 21, 437-180. doi: 10.1007/s 11145-007-9098-2.
Wanzek, J., & Vaughn, S. (2007). Research-based implications from extensive early reading interventions. School Psychology Review, 36, 541-561.
Wanzek, J., & Vaughn, S. (2008). Response to varying amounts of time in reading intervention for students with low response to intervention. Journal of Learning Disabilities, 41, 126-142.
Wanzek, J., & Vaughn, S. (2010). Tier 3 interventions for students with significant reading problems. Theory Into Practice, 49, 305-314.
Woodcock, R. W., McGrew, K. S., & Mather, N. (2001). Examiner's manual: Woodcock-Johnson III Tests of Cognitive Ability. Itasca, IL: Riverside.
Zirkel, P. A., & Thomas, L. B. (2010). State laws for RTF An updated snapshot. TEACHING Exceptional Children, 42, 56-63.
Stephanie Al Otaiba (1), Carol M. Connor (2), Jessica S. Folsom (3), Jeanne Wanzek (3), Luana Greulich (4), Christopher Schatschneider (4), and Richard K. Wagner (5)
(1) Southern Methodist University
(2) Arizona State University
(3) Florida Center for Reading Research, Florida State University
(4) Andrews University
(5) Florida Center for Reading Research, Florida State University
Stephanie Al Otaiba, Department of Teaching and Learning, Annette Caldwell Simmons School of Education and Human Development, Southern Methodist University, PO Box 750455, Dallas, TX 75275-0455.
Stephanie Al Otaiba, Southern Methodist University-Department of Teaching and Learning; Carol M. Connor, Arizona State University; Jessica S. Folsom, Florida State University; Jeanne Wanzek, Florida State University; Luana Greulich, Andrews University; Christopher Schatschneider, Florida State University; and Richard K. Wagner, Florida State University.
This work was supported by a multidisciplinary Learning Disabilities Center grant (P50HD052120) from the Eunice Kennedy Shriver National Institute of Child Health and Human Development. The content is solely the responsibility of the authors and does not necessarily represent the official views of the Eunice Kennedy Shriver National Institute of Child Health and Human Development or the National Institutes of Health.
Manuscript received March 2013; accepted August 2013.
Table 1. Demographics by Initial Eligibility to Tier and by Condition. Tier 1 Typical Dynamic (n = 178) (n = 180) Age, y 6.16 (0.33) 6.17 (0.34) Male, % 46.1 52.5 Free/reduced-price lunch, % 50.0 48.0 Race, % Caucasian 40.4 44.1 African American 47.2 43.0 Other 12.4 12.8 Special education, n Speech impairment 6 5 Language impairment 3 2 Deaf or hard of hearing 0 0 Mild ID/DD 0 1 Learning disability 0 1 Emotional handicapped 0 1 Visual impairment 3 1 ASD 1 1 IQ Verbal N/A N/A Nonverbal N/A N/A Initial screen (Sept) TRRP 0.36 (1.13) 0.38 (1.18) LSF 43.94 (12.41) 43.93 (12.8) WIF 35.43 (23.68) 39.4 (23.76) SWE (a) 94.96 (13.29) 97.20 (13.44) PDE (a) 95.41 (12.09) 95.56 (13.56) Second screen (Dec) LSF 52.52 (13.28) 51.2 (12.52) WIF 54.10 (27.47) 56.53 (25.23) SWE (a) 104.29 (14.27) 105.84 (14.64) PDE (a) 100.37 (12.59) 100.37 (13.26) Third screen (Mar) TRRP, M 0.25 0.14 LSF 58.3 (13.16) 56.29 (13.68) WIF 66.48 (25.05) 70.4 (23.00) SWE (a) 104.29 (14.27) 105.84 (14.64) PDE (a) 103.01 (12.82) 104.87 (13.45) Growth, Z score Fall to winter 0.03 (0.62) -0.04 (0.57) Winter to spring -0.00 (0.64) 0.07 (0.67) Tier 2 Typical Dynamic (n = 54) (n = 58) Age, y 6.33 (0.47) 6.21 (0.40) Male, % 20.4 60.3 Free/reduced-price lunch, % 70.4 67.2 Race, % Caucasian 61.1 53.4 African American 33.3 36.2 Other 5.6 10.3 Special education, n Speech impairment 7 9 Language impairment 3 7 Deaf or hard of hearing 1 1 Mild ID/DD 1 2 Learning disability 3 0 Emotional handicapped 0 0 Visual impairment 2 1 ASD 1 0 IQ Verbal 90.78 (11.99) 93.19 (13.44) Nonverbal 90 (12.16) 90.31 (11.58) Initial screen (Sept) TRRP 2.43 (1.81) 2.40 (1.8) LSF 33.24 (10.70) 34.21 (10.3) WIF 9.98 (6.24) 8.41 (5.79) SWE (a) 77.74 (7.54) 78.49 (7.01) PDE (a) 80.54 (7.99) 83.48 (7.41) Second screen (Dec) LSF 45.91 (13.00) 49.45 (13.91) WIF 23.96 (18.3) 22.09 (16.13) SWE (a) 87.93 (12.99) 89.81 (14.35) PDE (a) 86.83 (9.72) 87.43 (9.25) Third screen (Mar) TRRP, M 1.37 1.43 LSF 51.35 (13.22) 61.07 (14.39) WIF 41.78 (21.55) 39.26 (26.03) SWE (a) 87.93 (12.997) 89.81 (14.35) PDE (a) 90.24 (1 1.51) 92.83 (10.73) Growth, Z score Fall to winter -0.05 (0.61) -0.00 (0.67) Winter to spring 0.12 (0.54) 0.45 (0.66) Tier 3 Typical Dynamic (n = 29) (n = 23) Age, y 6.3 (0.49) 6.36 (0.46) Male, % 20.7 52.2 Free/reduced-price lunch, % 69.0 91.3 Race, % Caucasian 51.7 69.6 African American 27.6 21.7 Other 20.7 8.7 Special education, n Speech impairment 3 4 Language impairment 4 4 Deaf or hard of hearing 0 0 Mild ID/DD 2 2 Learning disability 4 0 Emotional handicapped 1 0 Visual impairment 3 0 ASD 0 0 IQ Verbal 89.04 (13.85) 87.35 (11.44) Nonverbal 85.86 (10.49) 84.61 (10.02) Initial screen (Sept) TRRP 4.55 (0.51) 4.61 (0.5) LSF 19.41 (9.97) 17.43 (9.3) WIF 4.17 (4.15) 4.17 (4.38) SWE (a) 72.03 (8.09) 69.17 (7.42) PDE (a) 75.93 (5.35) 75.35 (4.53) Second screen (Dec) LSF 36.76 (12.83) 42.57 (15.15) WIF 10.31 (9.21) 8.00 (6.94) SWE (a) 76.28 (9.80) 78.23 (11.09) PDE (a) 81.69 (8.94) 81.40 (8.25) Third screen (Mar) TRRP, M 3.655 3.3 LSF 46.17 (12.01) 54.91 (1 1.97) WIF 18.55 (14.92) 16.13 (13.93) SWE (a) 76.28 (8.96) 78.22 (1 1.09) PDE (a) 83.00 (8.98) 88.04 (9.51) Growth, Z score Fall to winter -0.28 (0.53) -0.06 (0.49) Winter to spring -0.09 (0.50) 0.19 (0.72) Note. Values In M (SD) unless specified otherwise. By design, no students in traditional response to intervention received Tier 2 or 3 during Session I even if they were eligible; similarly, no students in traditional response to intervention received Tier 3 until Session 3. ID/DD = intellectual disability or developmental disability; ASD = autism spectrum disorder; N/A = not assessed; TRRP = Teacher Rating of Reading Problems (Speece et al., 201 I); LSF = AIMSWeb Letter Sound Fluency (Shinn & Shinn, 2004); WIF = Word Identification Fluency (Fuchs, Fuchs, & Compton, 2004); SWE = Sight Word Efficiency; PDE = Phonemic Decoding Efficiency. (a) Standard scores on the Test of Word Reading Efficiency (Torgesen, Wagner, & Rashotte, 1999) Table 2. Results of Principal Component Analyses for Reading Outcomes: Fall (Pretreatment) and Spring (Posttreatment). Variable Component I Fall component matrix Letter Word Identification (a) 0.963 Word Attack (a) 0.849 Passage Comprehension (a) 0.896 Word Identification Fluency (b) 0.95 Letter Sound Fluency (c) 0.63 Phonemic Decoding Efficiency (d) 0.913 Oral Reading Fluency (e) 0.93 Spring component matrix Letter Word Identification (a) 0.816 Word Attack (a) 0.826 Passage Comprehension (a) 0.864 Word Identification Fluency (b) 0.882 Letter Sound Fluency (c) 0.302 Phonemic Decoding Efficiency (d) 0.887 Sight Word Efficiency (d) 0.94 Oral Reading Fluency (e) 0.922 TRRP -0.688 Note. For both primary component analsyes, only one component was extracted. TRRP = Teacher Rating of Reading Problems (Speece et al., 2011). (a) Woodcock-Johnson III Test of Achievement (Woodcock, McGrew, & Mather, 2001). (b) Word Identification Fluency (Fuchs, Fuchs, & Compton, 2004). (c) AIMSWeb (Shinn & Shinn, 2004). (d) Test of Word Reading Efficiency (Torgesen, Wagner, & Rashotte, 1999). (e) Dynamic Indicators of Basic Early Literacy Skills (Good & Kaminski, 2002). Table 3. Hierarchical Linear Modeling Results of Intent-to-Treat Analysis. Fixed Effect Coefficient SE T Ratio df P Intercept -0.10 0.10 -1.07 33 .294 Dynamic RTI 0.17 0.05 3.19 527 .002 [chi Random Effect SD Variance df square] P Intercept 0.50 0.25 33 201.73 < .001 Level I 0.86 0.74 Note. Deviance = 1,342.56. RTI = response to intervention. Table 4. Latent Growth Curve Analysis of Brief Reading Scores Across First Grade as a Function of Tier and Status. Fixed Effect Coefficient SE T Ratio df p Intercept 489.22 2.49 196.18 33 <.001 Dynamic -2.43 2.58 -0.94 518 .347 Tier -15.48 1.15 -13.42 470 <.001 Tier x 3.81 1.52 2.51 470 .012 Dynamic Slope 4.66 0.12 37.36 518 <.001 Dynamic -0.03 0.18 -0.16 518 .872 [chi Random Effect SD Variance df square] P Level 2 Intercept 11.47 131.57 503 1193.84 <.001 Slope 0.75 0.57 536 620.37 .007 Level 1 10.83 117.32 Level 3 Intercept 9.32 86.791 33 253.07 <.001 Note. Deviance = 13,378.06. Latent growth curve analysis of Brief Reading scores across the school year as a function of tier (1, 2, or 3) as a time-varying covariate and dynamic versus traditional response to intervention. Time is in months centered at the end of the year.
|Printer friendly Cite/link Email Feedback|
|Author:||Otaiba, Stephanie Al; Connor, Carol M.; Folsom, Jessica S.; Wanzek, Jeanne; Greulich, Luana; Schatsc|
|Date:||Oct 1, 2014|
|Previous Article:||Report to the readership: Exceptional Children, volume 80.|
|Next Article:||Self-management interventions on students with autism: a meta-analysis of single-subject research.|