Research Examples: Centers for Implementing K-3 Behavior and Reading Intervention Models

Preventing Reading Difficulties: A Three-Tiered Intervention Model

Sharon Vaughn, Ph.D., and Jeanne Wanzek, Ph.D.
University of Texas Center for Reading and Language Arts

Goals

The overall goals of this five-year project (2002-2006) were

to develop, evaluate, and disseminate a school-based model for the prevention of reading disabilities
to develop a three-tiered intervention model to support students at risk for developing reading disabilities
to reduce the number of students identified for special education based on reading disabilities.

Six elementary schools in a district near Austin, Texas, participated in the research study. Within these schools, more than 80 percent of the students were minority students, and more than 80 percent received free or reduced lunch.

With this research, Vaughn and her colleagues examined the effectiveness of the implementation and sustainability of the three-tier model through observations, interviews, and field notes. They anticipated that this aspect of the evaluation would provide valuable information about barriers to and facilitators of effective implementation.

One of the three practical outcomes that researchers anticipated to be of highest importance to special education and general education teachers was to determine the effectiveness of relatively brief interventions (for example, Tier 2 interventions that comprised approximately 50 sessions for 20 to 30 minutes per day) compared with the effectiveness of more intensive interventions in Tier 3 (100 sessions for 50 to 60 minutes/day).

Overview

For this research study, three intervention levels were implemented across kindergarten through third grade to prevent reading difficulties. All levels included screening, systematic progress monitoring, and the use of scientifically based reading instruction. Students who did not make adequate progress in Tier 1 (general education) received Tier 2 reading intervention. Students who caught up to their peers left Tier 2 but their progress continued to be monitored in Tier 1. A similar process was followed for Tier 3. All students remained in Tier 1 even when they were also being served in Tier 2 or Tier 3.

Principal involvement. The role of the principal cannot be overemphasized. There was a strong link between principal leadership and teacher interest, motivation, and effective implementation of the program. The researchers and project directors encouraged principals' involvement in the project by meeting with them each month, presenting at a district-wide principal meeting once each semester, and co-presenting with principals at a conference of the Texas Elementary Principals and Supervisors Association. The project directors also assisted schools in sustaining the practices implemented in the research project.

Researchers shared standardized test data and progress monitoring data with school and district leaders to inform their decision making regarding student, teacher, and school progress toward successful reading outcomes. Researchers provided graphs illustrating DIBELS class- and school-wide data for principals and provided school- and district-level data to district administrators.

Professional development. Professional development was extensive and assisted teachers and grade-level teams in their understanding and use of progress monitoring information. Professional development also assisted individual teachers and principals in interpretation, grouping, and instructional practices related to students most at risk for reading difficulties.

Session topics that related to Tier 1 included DIBELS for progress monitoring, phonological awareness, classroom behavior management, using assessment information to group students for instruction, differentiated instruction, using data to make instructional decisions, implementation of K-PALS (Peer Assisted Learning Strategies), teacher partnering, focus group discussions, collaborative strategic reading, vocabulary instruction, advanced word study, effective instructional practices, and organizing and designing the core reading block.

Professional development sessions that related specifically to Tier 2 included phonological/phonemic awareness, letter and sound identification, phonics and word recognition, fluency, word reading, sentence/story reading, passage reading, and comprehension.

Sessions related to Tier 3 included sound review, phonics and word recognition, vocabulary, fluency, passage reading, and comprehension.

Focus groups and coaching. All participating teachers engaged once a year in focus groups to provide feedback about what was helpful and what was difficult with regard to the three-tier model. Researchers also wanted to determine the effectiveness of coaching and in-classroom support for enhancing implementation, progress monitoring, and ultimately student outcomes. Rigorous training and reliability procedures were used to prepare three-tier project testers.

Tier 1

Tier 1 instruction took place in the general education classroom with the general education teacher. Reading instruction took place for at least 90 minutes each day, was scientifically based, and emphasized the five critical elements of reading. Curriculum and instruction in kindergarten through second grade included a variety of strategies, and ideas based on scientifically based reading research and content previously developed by the Vaughn Gross Center for Reading and Language Arts (VGCRLA) were shared during professional development sessions. In addition, in kindergarten, the curriculum included Phonemic Awareness in Young Children and K-PALS; in first grade, PALS; and in second grade, partner reading.

Benchmark assessment data (DIBELS) were collected at the beginning, middle, and end of the year to identify students who needed intervention. Teachers used DIBELS to gather progress-monitoring data to inform and adjust their reading instruction. After DIBELS assessments, teachers were given a bar graph that indicated DIBELS scores for individual students. Tier 1 coordinators met with the teachers and provided instructional recommendations to increase student progress. Teachers also completed the Social Skills Rating System (SSRS) for students.

Decision Rules for Tier 2 and Tier 3 Instruction

For purposes of research, the kindergartners and first graders whose response to general education instruction was not adequate received additional instruction in Tier 2 from researchers. No student in kindergarten or first grade received Tier 3 instruction. Students in second and third grade who did not respond adequately to general education received Tier 3 instruction from researchers. The school provided Tier 2 instruction for second- and third-graders needing that level of instruction. Tables 1 and 2 describe qualification and exit criteria for Tiers 2 and 3.

Table 1: Tier 2 Qualification and Exit Criteria

Table 2: Tier 3 Qualification and Exit Criteria

Tier 2

Instruction and interventions. Tier 2 interventions, when needed, began immediately after identification with benchmark testing and were coordinated with the general education teacher. The interventions emphasized the five critical elements of beginning reading and were systematic, explicit, and included modeling, multiple examples, and corrective feedback. Reading interventions included phonological/phonemic awareness, letter and sound identification, phonics and word recognition, fluency, word reading, sentence/story reading, passage reading, and comprehension.

Personnel. For this research, the personnel providing Tier 2 instruction were graduate research assistants and full-time staff hired for tutoring; all had a college degree, some were certified teachers, and all were trained before beginning teaching. During hiring, Vaughn's team looked for tutors with previous experience teaching and working in schools, knowledge of reading instruction, and a willingness to implement a standard protocol intervention. To be considered "qualified," personnel had to be trained to 100 percent implementation fidelity. (Tier 2 interventionists implementing the research treatment did not attend the Tier 1 professional development sessions.)

Setting/time/pacing. Tier 2 treatment/tutoring sessions were always held outside the general education classroom (in pod areas or in a separate classroom, for example). One "round" of Tier 2 instruction lasted for 10 to 12 weeks (about 50 sessions), with each session lasting at least 30 minutes. This was in addition to the 90 minutes of core reading instruction provided in the general education classroom. Teacher-to-student ratios were either one-to-four or one-to-five. After one round of Tier 2, students could either exit Tier 2, repeat Tier 2, enter Tier 3, or be referred to special services. Pacing was matched to each student's skill level, and each student had multiple opportunities to participate and respond.

Progress monitoring. Teachers used DIBELS benchmarks and progress monitoring (assessment data collected three times each year) to inform decisions about grouping and to adjust their reading instruction. In addition, the teachers monitored the progress of students in Tier 2 every week to ensure adequate progress on the targeted skill.

All teachers were trained in the administration and interpretation of DIBELS. (Researchers assisted in the collection and interpretation of DIBELS.) Teachers also completed the Social Skills Rating System (SSRS) on students.

Qualification and exit criteria. A student was selected for Tier 3 instruction in one of three ways:

If progress was not sufficient after two rounds of Tier 2 instruction even after adjustments to instruction
If a marked lack of progress was evident after only one round of Tier 2 and further Tier 2 instruction was deemed insufficient
If the student required re-entry into Tier 3 after a previous exit

A student exited Tier 3 when she or he reached the benchmark on the targeted skills. A student who had previously exited Tier 3 re-entered as needed.

Instruction and interventions. The program for Tier 3 was scientifically based and emphasized the critical elements of reading for students with reading difficulties or disabilities. Tier 3 instruction was sustained, intensive, and strategic; it was specifically designed and customized for small group (one-to-three) reading instruction. Interventions for Tier 3 included sound review, phonics and word recognition, vocabulary, fluency, passage reading, and comprehension.

When a student "repeated" or got a "second dose" of a Tier 2 or Tier 3 intervention, the intervention was continued with the same intensity, at the student's skill level, with the speed of progression being determined by student mastery, as was the case for the first "dose."

Interventions were not scheduled during core reading instruction, and teachers selected the times for interventions according to times in the classroom that they thought students could make up work.

Personnel. Qualified personnel for Tier 3 were classroom teachers, reading specialists, or outside interventionists. In the research treatment, the qualified personnel were graduate research assistants and full-time staff hired for tutoring. All had college degrees, some were certified teachers, and all were trained to 100 percent implementation fidelity before beginning teaching. Researchers looked for tutors with previous experience teaching and working in the schools, knowledge of reading instruction, and a willingness to implement a standard treatment protocol.

Setting/time/pacing. The setting for Tier 3 instruction was always outside the general education classroom (in pod areas or separate classrooms, for example). Each group consisted of two or three students. Tier 3 instruction was nearly always one 50- or 60-minute session each day for 100 days (across the school year). Students could exit after 50 sessions. For a couple of groups of students, Tier 3 comprised two sessions each day (30 minutes per session) for 100 days across the school year.

Number of cycles. Students could have a number of Tier 3 intervention cycles. Students in the second grade design might have experienced Tier 1 plus Tier 3 in consecutive semesters if the assessments at the beginning of each semester determined that they were eligible.

A student who had received previous Tier 3 instruction and had exited could re-enter Tier 3 as needed. Students could exit from Tier 3 intervention during the middle of the school year only if they demonstrated grade-level performance on oral reading fluency measures.

Progress monitoring. Progress monitoring occurred twice a month on the targeted skill to ensure adequate progress and learning and was based on the grade level of the students. Phonemic segmentation fluency and nonsense word fluency were used in the fall of first grade, and nonsense word fluency and oral reading fluency were used in the spring of first grade. Oral reading fluency was used in second and third grade. At each level of the three-tier model, there was documentation of the individual characteristics, background, school experiences, and outcomes of students who did, and did not, make adequate progress. (See Table 2 for qualification and exit criteria for Tier 3.)

Fidelity of Implementation

Fidelity measures/observations - Tier 1. Fidelity of implementation was monitored for all interventions. To monitor implementation, the researchers used classroom observations (a strictly passive activity for the researcher) and student data. The researchers reviewed several tools for collecting data on teachers' delivery of reading instruction and selected the revised Instructional Content Emphasis (ICE-R) instrument. This is a valid and reliable observation instrument used to systematically categorize and code the content of reading and language arts instruction and can be used to collect data helpful in answering the following questions: What is being taught? How is it being taught? How well is it being taught? What is being used to teach? Reliability checks were done before instructional methods were used in the schools, and frequent discussions between observers ensured that instruction was coded reliably across observers. During observations, Tier 1 coordinators used a checklist of key features (the ICE-R). Data from the ICE-R determined fidelity.

Informal classroom observations and visits. School site coordinators visited the teachers in their classrooms on a regular basis and provided follow-up to the professional development sessions (for example, modeling strategies). During informal visits, school site coordinators did not complete the ICE-R and were able to be actively involved (for example, modeling a lesson). Informal visits usually lasted only 15 to 30 minutes. Field notes or short observation checklists were completed after each informal observation. Data collected during these visits were then compiled with the ICE-R data to create a complete picture of the instruction each teacher provided at Tier I.

Classroom observation data were collected three times for current intervention teachers and once for other participating teachers to document reading instruction and the accurate implementation of strategies addressed in professional development activities.

Fidelity measures for Tier 2 and Tier 3. For Tier 2 and Tier 3 interventions, two observers were trained on specific fidelity measures using videos. The observers then used live observations for reliability. Inter-rater reliability for both intervention measures was above 90 percent. Reliability was also discussed at length in training. Fidelity checklists were completed three times per semester for each tutor.

RTI as an Effective Prevention System

For this research, the reading skills of all students were assessed. Comparison of control and experimental groups indicated that the three-tier model improved the reading outcomes of students participating in Tier 1 interventions and decreased the number of students in need of Tier 2 interventions. Implementation of Tier 2 intervention for struggling readers was also shown to improve student reading outcomes and allow most students to exit intervention.

Parent Involvement

Parents were provided information and training to facilitate active involvement in student reading development. Researchers planned to inform and train parents by using an enhanced web site, literacy-related articles in school and district newsletters, and informational workshops at individual schools. Researchers also considered giving a parent survey.

Disability Determination

This model was not used for specific learning disability determination and special education eligibility. The three-tier project focused on effectiveness for early identification and remediation of students at risk for reading problems and students with disabilities. It allowed stakeholders to examine the reading profiles of students later identified for special education, including the amount and effectiveness of Tier 2 and Tier 3 interventions provided. Although this model is not used to determine SLD eligibility, the researchers established a well-organized and sophisticated data management system that allowed ready access to academic information about specific students.

Due Process Procedures

The project was not involved with due process procedures.