Assisting Students Struggling with Reading: Response to Intervention and Multi-Tier Intervention in the Primary Grades

IES PRACTICE GUIDE WHAT WORKS CLEARINGHOUSE

AssistingAssisting StudentsStudents StrugglingStruggling withwith Reading:Reading: ResponseResponse toto InterventionIntervention (RtI)(RtI) andand Multi-TierMulti-Tier InterventionIntervention inin thethe PrimaryPrimary GradesGrades

NCEE 2009-4045 U.S. DEPARTMENT OF EDUCATION The Institute of Education Sciences (IES) publishes practice guides in education to bring the best available evidence and expertise to bear on the types of systemic challenges that cannot currently be addressed by single interventions or programs. Authors of practice guides seldom conduct the types of systematic literature searches that are the backbone of a meta-analysis, although they take advantage of such work when it is already published. Instead, authors use their expertise to identify the most important research with respect to their recommendations, augmented by a search of recent publications to ensure that research citations are up-to-date.

Unique to IES-sponsored practice guides is that they are subjected to rigorous external peer review through the same office that is responsible for independent review of other IES publications. A critical task for peer reviewers of a practice guide is to determine whether the evidence cited in support of particular recommendations is up-to-date and that studies of similar or better quality that point in a different direction have not been ignored. Because practice guides depend on the expertise of their authors and their group decision-making, the content of a practice guide is not and should not be viewed as a set of recommendations that in every case depends on and flows inevitably from scientific research.

The goal of this practice guide is to formulate specific and coherent evidence-based recommendations for use by educators addressing the challenge of reducing the number of children who fail to learn how to read proficiently by using “response to intervention” as a means of both preventing reading difficulty and identifying students who need more help. This is called Response to Intervention (RtI). The guide provides practical, clear information on critical RtI topics and is based on the best available evidence as judged by the panel. Recommendations in this guide should not be construed to imply that no further research is warranted on the effectiveness of particular RtI strategies. IES PRACTICE GUIDE

Assisting Students Struggling with Reading: Response to Intervention and Multi-Tier Intervention in the Primary Grades

February 2009

Panel Russell Gersten (Chair) Instructional Research Group

Donald Compton Vanderbilt University

Carol M. Connor Florida State University

Joseph Dimino Instructional Research Group

Lana Santoro Instructional Research Group

Sylvia Linan-Thompson University of Texas—Austin

W. David Tilly Heartland Area Education Agency

Staff Rebecca Newman-Gonchar Instructional Research Group

Kristin Hallgren Mathematica Policy Research

NCEE 2009-4045 U.S. DEPARTMENT OF EDUCATION This report was prepared for the National Center for Education Evaluation and Regional Assistance, Institute of Education Sciences under Contract ED-07-CO-0062 by the What Works Clearinghouse, which is operated by Mathematica Policy Research, Inc.

Disclaimer

The opinions and positions expressed in this practice guide are the authors’ and do not necessarily represent the opinions and positions of the Institute of Education Sci- ences or the U.S. Department of Education. This practice guide should be reviewed and applied according to the specific needs of the educators and education agency using it, and with full realization that it represents the judgments of the review panel regarding what constitutes sensible practice, based on the research that was available at the time of publication. This practice guide should be used as a tool to assist in decision-making rather than as a “cookbook.” Any references within the document to specific education products are illustrative and do not imply endorse- ment of these products to the exclusion of other products that are not referenced.

U.S. Department of Education Arne Duncan Secretary

Institute of Education Sciences Sue Betka Acting Director

National Center for Education Evaluation and Regional Assistance Phoebe Cottingham Commissioner

February 2009

This report is in the public domain. While permission to reprint this publication is not necessary, the citation should be:

Gersten, R., Compton, D., Connor, C.M., Dimino, J., Santoro, L., Linan-Thompson, S., and Tilly, W.D. (2008). Assisting students struggling with reading: Response to Intervention and multi-tier intervention for reading in the primary grades. A practice guide. (NCEE 2009-4045). Washington, DC: National Center for Edu- cation Evaluation and Regional Assistance, Institute of Education Sciences, U.S. Department of Education. Retrieved from http://ies.ed.gov/ncee/wwc/ publications/practiceguides/.

This report is available on the IES website at http://ies.ed.gov/ncee and http://ies. ed.gov/ncee/wwc/publications/practiceguides/.

Alternative formats On request, this publication can be made available in alternative formats, such as Braille, large print, audiotape, or computer diskette. For more information, call the alternative format center at (202) 205-8113. Assisting Students Struggling with Reading: Response to Intervention and Multi-Tier Intervention in the Primary Grades

Contents

Introduction 1 The What Works Clearinghouse standards and their relevance to this guide 2

Overview 4

Scope of the guide 8

Checklist for carrying out the recommendations 9

Recommendation 1. Screen all students for potential reading problems at the beginning of the year and again in the middle of the year. Regularly monitor the progress of students who are at elevated risk for developing reading disabilities. 11

Recommendation 2. Provide differentiated reading instruction for all students based on assessments of students’ current reading levels (tier 1). 17

Recommendation 3. Provide intensive, systematic instruction on up to three foundational reading skills in small groups to students who score below the benchmark on universal screening. Typically these groups meet between three and five times a week for 20–40 minutes (tier 2). 19

Recommendation 4. Monitor the progress of tier 2 students at least once a month. Use these data to determine whether students still require intervention. For those still making insufficient progress, school-wide teams should design a tier 3 intervention plan. 24

Recommendation 5. Provide intensive instruction daily that promotes the development of various components of reading proficiency to students who show minimal progress after reasonable time in tier 2 small group instruction (tier 3). 26

Appendix A. Postscript from the Institute of Education Sciences 32

Appendix B. About the authors 35

Appendix C. Disclosure of potential conflicts of interest 38

Appendix D. Technical information on the studies 39

References 50

( iii ) ASSISTING STUDENTS STRUGGLING WITH READING: RESPONSE TO INTERVENTION AND MULTI-TIER INTERVENTION IN THE PRIMARY GRADES

List of tables

Table 1. Institute of Education Sciences levels of evidence for practice guides 3

Table 2. Recommendations and corresponding levels of evidence 6

Table 3. Recommended target areas for early screening and progress monitoring 13

Table 4. Foundational reading skills in grades K–2 21

Table 5. Progress monitoring measures in grades K–2 25

Table D1. Studies of tier 2 interventions in grades K–2 reading that met What Works Clearinghouse standards 41

( iv ) Introduction Moderate refers to evidence from studies that allow strong causal conclusions but In the primary grades students with read- cannot be generalized with assurance to ing difficulties may need intervention to the population on which a recommenda- prevent future reading failure. This guide tion is focused (perhaps because the find- offers specific recommendations to help ings have not been widely replicated) or to educators identify students in need of in- evidence from studies that are generaliz- tervention and implement evidence-based able but have more causal ambiguity than interventions to promote their reading offered by experimental designs (such as achievement. It also describes how to carry statistical models of correlational data out each recommendation, including how or group comparison designs for which to address potential roadblocks in imple- equivalence of the groups at pretest is menting them. uncertain).

We, the authors, are a small group with ex- Low refers to expert opinion based on rea- pertise in various dimensions of this topic. sonable extrapolations from research and Several of us are also experts in research theory on other topics and evidence from methodology. The recommendations in studies that do not meet the standards for this guide reflect not only our expertise moderate or strong evidence. and experience but the findings of rigorous studies of interventions to promote Table 1 details the criteria used to deter- reading achievement. mine the level of evidence for each recommendation. For questions about what Each recommendation received a rating works best, high-quality experimental and that describes the strength of the research quasi-experimental studies, such as those evidence that has shown its effectiveness. meeting the criteria of the What Works These ratings—“strong,” “moderate,” or Clearinghouse (www.whatworks.ed.gov), “low”—are defined as: have a privileged position. The evidence considered in developing and rating these Strong refers to consistent and generaliz- recommendations included experimental able evidence that a program causes bet- research on providing differentiated inter outcomes.1 struction in a general education classroom and rigorous evaluations of intensive read- 1. Following WWC guidelines, we consider a posi- ing interventions. We also examined stud- tive, statistically significant effect, or an effect ies on the technical adequacy of batteries size greater than 0.25, as an indicator of posi- of screening measures. tive effects.

( 1 ) Introduction

The What Works Clearinghouse Based on the recommendations and sug- standards and their relevance to gestions for their implementation, ap- this guide pendix D presents more information on the research evidence supporting the The panel relied on WWC Evidence Stan- recommendations. dards to assess the quality of evidence supporting educational programs and The panel would like to thank Kelly Hay- practices and apply a level of evidence mond for her contributions to the analy- rating to each recommendation. The WWC sis, Mary Jo Taylor for her expert editorial addresses evidence for the causal validity assistance, the WWC reviewers for their of instructional programs and practices contribution to the project, and Jo Ellen using WWC Standards. Information about Kerr for her support of the intricate logis- these standards is available at http://ies. tics of the project. We also would like to ed.gov/ncee/wwc/references/standards/. thank Scott Cody for his oversight of the The technical quality of each study is rated analyses and the overall progress of the and placed into one of three categories: practice guide.

• Meets Evidence Standards for random- Dr. Russell Gersten ized controlled trials and regression Dr. Donald Compton discontinuity studies that provide the Dr. Carol M. Connor strongest evidence of causal validity. Dr. Joseph Dimino Dr. Lana Santoro • Meets Evidence Standards with Res- Dr. Sylvia Linan-Thompson ervations for all quasi-experimental Dr. W. David Tilly studies with no design flaws and randomized controlled trials that have problems with randomization, attrition, or disruption.

• Does Not Meet Evidence Screens for studies that do not provide strong evidence of causal validity.

( 2 ) Introduction

Table 1. Institute of Education Sciences levels of evidence for practice guides

In general, characterization of the evidence for a recommendation as strong requires both studies with high internal validity (i.e., studies whose designs can support causal conclusions) and studies with high external validity (i.e., studies that in total include enough of the range of participants and settings on which the recommendation is focused to support the conclusion that the results can be generalized to those participants and settings). Strong evidence for this practice guide is operationalized as: • A systematic review of research that generally meets the What Works Clearinghouse (WWC) standards (see http://ies.ed.gov/ncee/wwc/) and supports the effectiveness of a program, Strong practice, or approach, with no contradictory evidence of similar quality; OR • Several well designed, randomized controlled trials or well designed quasi-experiments that generally meet WWC standards and support the effectiveness of a program, practice, or approach, with no contradictory evidence of similar quality; OR • One large, well designed, randomized controlled, multisite trial that meets WWC standards and supports the effectiveness of a program, practice, or approach, with no contradictory evidence of similar quality; OR • For assessments, evidence of reliability and validity that meets the Standards for Educa- tional and Psychological Testing.a

In general, characterization of the evidence for a recommendation as moderate requires studies with high internal validity but moderate external validity, or studies with high external validity but moderate internal validity. In other words, moderate evidence is derived from studies that support strong causal conclusions, but where generalization is uncertain, or studies that support the generality of a relationship, but where the causality is uncertain. Moderate evidence for this practice guide is operationalized as: • Experiments or quasi-experiments generally meeting WWC standards and supporting the effectiveness of a program, practice, or approach with small sample sizes and/or other conditions of implementation or analysis that limit generalizability and no contrary evidence; OR Moderate • Comparison group studies that do not demonstrate equivalence of groups at pretest and therefore do not meet WWC standards but that (a) consistently show enhanced outcomes for participants experiencing a particular program, practice, or approach and (b) have no major flaws related to internal validity other than lack of demonstrated equivalence at pretest (e.g., only one teacher or one class per condition, unequal amounts of instructional time, highly biased outcome measures); OR • Correlational research with strong statistical controls for selection bias and for discern- ing influence of endogenous factors and no contrary evidence; OR • For assessments, evidence of reliability that meets the Standards for Educational and Psy- chological Testingb but with evidence of validity from samples not adequately representative of the population on which the recommendation is focused.

In general, characterization of the evidence for a recommendation as low means that the recommendation is based on expert opinion derived from strong findings or theories in related Low areas or expert opinion buttressed by direct evidence that does not rise to the moderate or strong levels. Low evidence is operationalized as evidence not meeting the standards for the moderate or high levels. a. American Educational Research Association, American Psychological Association, and National Council on Measurement in Education (1999). b. Ibid.

( 3 ) Assisting Students • Tier 1 instruction is generally defined as reading instruction provided to all Struggling with Reading: students in a class. Beyond this gen- Response to Intervention eral definition, there is no clear con- and Multi-Tier sensus on the meaning of the term tier 1. Instead, it is variously referred to as Intervention for Reading “evidence-based reading instruction,”4 in the Primary Grades “high quality reading instruction,”5 or “an instructional program…with bal- anced, explicit, and systematic reading Overview instruction that fosters both code-based and text-based strategies for word iden- Response to Intervention (RtI) is a compre- tification and comprehension.”6 hensive early detection and prevention strategy that identifies struggling students and • Tier 2 interventions are provided only assists them before they fall behind. RtI sys- to students who demonstrate prob- tems combine universal screening and high- lems based on screening measures or quality instruction for all students with in- weak progress from regular classroom terventions targeted at struggling students. instruction. In addition to general classroom instruction, tier 2 students RtI strategies are used in both reading and receive supplemental, small group math instruction. For reading instruction reading instruction aimed at building in the primary grades (K–2), schools screen foundational reading skills. students at least once a year to identify students at risk for future reading failure.2 • Tier 3 interventions are provided to Students whose screening scores indicate students who do not progress after a potential difficulties with learning to read reasonable amount of time with the are provided with more intensive reading tier 2 intervention and require more interventions. Student responses to the intensive assistance. Tier 3 (or, in dis- interventions are then measured to deter- tricts with more than three tiers, tiers mine whether they have made adequate 3 and above) usually entails one-on- progress and either (1) no longer need the one tutoring with a mix of instruc- intervention, (2) continue to need some tional interventions. Ongoing analysis intervention, or (3) need even more inten- of student performance data is critical sive intervention. in tier 3. Systematically collected data are used to identify successes and In RtI, the levels of interventions are conven- failures in instruction for individual tionally referred to as “tiers.” RtI is typically students. If students still experience thought of as having three tiers, with the difficulty after receiving intensive ser- first tier encompassing general classroom vices, they are evaluated for possible instruction.3 Some states and school dis- special education services. tricts, however, have implemented multi-tier intervention systems with more than three Though a relatively new concept, RtI and tiers. Within a three-tier RtI model, each tier multi-tier interventions are becoming in- is defined by specific characteristics: creasingly common. This is attributed in

2. Johnson, Jenkins, Petscher, and Catts (in 4. Vaughn and Fuchs (2006). press, pp. 3–4). 5. Division for Learning Disabilities (2007). 3. Fuchs, Fuchs, and Vaughn (2008) make the 6. Vellutino, Scanlon, Small, Fanuele, and Sweeney case for a three-tier RtI model. (2007).

( 4 ) Overview

part to the 2004 reauthorization of the In- until they were officially diagnosed with a dividuals with Disabilities Education Act specific learning disability, often not until (IDEA), which encourages states to use RtI to grade 2 or 3.10 This was the practice even help prevent reading difficulties and to iden- though longitudinal research consistently tify students with learning disabilities. showed that students who were weak readers at the early elementary grades tended to RtI’s inclusion in the 2004 reauthorization stay weak readers in the higher grades.11 can be traced to two key reports released in 2002. First, the President’s Commission RtI also urges schools to use evidence- on Excellence in Special Education (2002) based practices in all tiers and to provide report revealed that special education put intensive services only to students who fail too much emphasis on paperwork and too to benefit from a well designed, evidence- little on instruction.7 It recommended that based intervention. This helps to accurately educators put more energy into monitor- determine which students possess learning ing student progress in academic areas disabilities in reading since only students and less into monitoring paperwork and who do not respond to high-quality read- compliance with regulations. ing instruction in their general education classrooms would be considered for special Second, a 2002 report from the National education. Thus, there is the possibility— Academy of Sciences examined the over- and certainly the hope—that RtI will reduce representation of students from minority inappropriate referrals to special educa- subgroups in special education.8 This re- tion, especially of ethnic minority students, port proposed ideas for making the referral low-income students, and students who re- process for learning disabilities more mean- ceived weak reading instruction.12 ingful to classroom teachers, arguing that special education “eligibility ensue when a The panel also believes that RtI holds the student exhibits large differences from typi- most potential for serious ongoing collabo- cal levels of performance in…[reading] and ration between the special education com- with evidence of insufficient response to high- munity and that of general education— quality interventions…in school settings.”9 largely because the collaboration is based This encouraged schools to provide services on objective data and shared understand- to students struggling in reading within ings of the evidence. general education in the early grades before considering special education. Special Summary of the Recommendations education would be considered only for students who failed to respond to evidence- This practice guide offers five concrete based interventions or interventions using recommendations for helping elementary what the field considers best practice. schools implement an RtI framework to ensure that all students in the primary grades There are two potential advantages of RtI learn to read. These recommendations and multi-tier intervention. Struggling students are provided with help in learning 10. Donovan and Cross (2002); Heller, Holtzman, how to read early in their school careers. and Messick (1982). In the past many students were not pro- 11. See Cunningham and Stanovich (1997); Fel- vided with additional assistance in reading ton and Pepper (1995); Phillips, Norris, Osmond, and Maynard (2002); Francis, Shaywitz, Stue- bing, Shaywitz, and Fletcher (1996); Juel (1988); 7. Haager, Klingner, and Vaughn (2007). Torgesen and Burgess (1998); Torgesen, Rashotte, 8. Donovan and Cross (2002). and Alexander (2001). 9. Cited in Haager et al. (2007, p. 5, emphasis 12. Donovan and Cross (2002); Heller, Holtzman, added). and Messick (1982).

( 5 ) Overview

appear in table 2. There are many ways coaches, specialists, and the principal. to orchestrate this process, and imple- This guide provides concrete guidance on menting this system entails involvement how to implement RtI; it does not describe of school personnel at many levels: class- which individuals on the team provide room teachers, special educators, school which services. psychologists, paraprofessionals, reading

Table 2. Recommendations and corresponding levels of evidence

Recommendation Level of evidence

1. Screen all students for potential reading problems at the beginning of the year and again in the middle of the year. Regularly monitor the Moderate progress of students at risk for developing reading disabilities.

Tier 1 intervention/general education

2. Provide time for differentiated reading instruction for all students based Low on assessments of students’ current reading level.

Tier 2 intervention

3. Provide intensive, systematic instruction on up to three foundational reading skills in small groups to students who score below the benchmark Strong score on universal screening. Typically, these groups meet between three and five times a week, for 20 to 40 minutes.

4. Monitor the progress of tier 2 students at least once a month. Use these data to determine whether students still require intervention. For those Low students still making insufficient progress, schoolwide teams should design a tier 3 intervention plan.

Tier 3 intervention

5. Provide intensive instruction on a daily basis that promotes the development of the various components of reading proficiency to students Low who show minimal progress after reasonable time in tier 2 small group instruction (tier 3).

Source: Authors’ compilation based on text.

( 6 ) Overview

We begin with specific methods for setting convergence in findings. It is not impor- up a universal screening system (recom- tant whether a certified teacher or a para- mendation 1). We note the specific read- professional provides the instruction. But ing and reading-related skills that should instruction should be systematic, highly be assessed in screening and progress- explicit, and highly interactive. We note monitoring measures at each grade level. that interventions must not focus only on We assume most educators possess some phonemic awareness, decoding, and fluent knowledge of universal screening. There- reading (depending on student proficiency fore, we provide specific suggestions on level) but should also include vocabulary how to ensure that the screening measures and comprehension components. used are effective. Recommendation 4 addresses using data As part of recommendation 1, we address to monitor progress for students in tier 2 the problem of false positives—students interventions. Although no studies have whose screening scores suggest that they experimentally tested the impact of prog- need additional assistance, but who would ress monitoring on outcomes in reading, do fine without it. This is a particular prob- we still encourage schools to monitor the lem for measures given at the beginning of progress of these students so that person- kindergarten; we explain why and what is nel possess information on how a student recommended. We urge that schools seri- is doing in general reading proficiency ously investigate both the degree to which and improving in specific skills. It is im- a screening measure correctly identifies portant to use progress-monitoring data students at risk for reading difficulties to regroup students after six weeks. Tier and identifies students at low risk for such 2 students who demonstrate improvement difficulties. and return to tier 1 should be carefully monitored to ensure that general class- The second recommendation addresses room instruction is adequate. how educators can use assessment data to differentiate reading instruction in tier Recommendation 5 addresses tier 3 in- 1. For example, classroom teachers can terventions, and we are candid about the use assessment data to determine which paucity of research on effective tier 3 in- students require additional instruction tervention. Tier 3 intervention is the most in decoding and vocabulary and which ambiguous component of RtI, and we did require additional assistance only with not find research on valid programs or decoding instruction. While the concept processes. Based on the content of small- of tier 1 instruction is amorphous, based scale intervention studies and the expert on conventional definitions, differentiated opinion of the panel, we suggest, as Vel- instruction is often mentioned as a critical lutino et al. (2007) suggest, that tier 3 component of tier 1.13 reading instruction be even more intensive than tier 2. Although student reading Recommendations 3 and 4 address tier 2 programs should be individualized, they interventions. In recommendation 3 we should be viewed as more than one-on- suggest that tier 2 students receive small one instruction. In particular, in listening group instruction in homogeneous groups and reading comprehension and vocabu- for 20 to 40 minutes, three to five days a lary development small group instruction week. This recommendation has the most makes sense. We also note that districts research and, most importantly, a clear should carefully monitor the success or failure of tier 3 programs, given the pau- 13. Connor, Morrison, Fishman, Schatschneider, city of available evidence. and Underwood (2007).

( 7 ) Scope of the We limit the focus of the guide to the primary grades because the bulk of the cur- practice guide rent research has focused on these grade levels. The majority of the research on in- Our goal is to provide evidence-based sug- tervention and screening of students with gestions for implementing multi-tier inter- reading difficulties was conducted in early ventions that are feasible and based on grade levels. In addition, for the past 15 evidence from rigorous research. RtI and years, the country has seen a large push multi-tier interventions transgress the bor- for early intervention to prevent reading ders of special and general education and difficulties later.14 demand schoolwide collaboration. Thus, our target audience includes classroom Multi-tier instruction efforts like RtI can teachers in the primary grades, special potentially prevent many struggling begin- educators, school psychologists and coun- ning readers from falling behind in ways selors, as well as administrators. that will harm their future academic success. Some aspects of RtI, however, (such This practice guide provides recommen- as tier 1 instruction) are still poorly de- dations to schools and school districts fined, and there is little evidence that some on using RtI for primary grade students practices of targeted instruction will be struggling with learning how to read. It effective. But a coordinated multi-tier in- is designed to guide educators on how struction program that screens and moni- to identify struggling students using RtI tors students accurately and addresses the and implement interventions to improve core components of reading instruction these students’ reading ability. The guide can prevent struggling beginning read- focuses on screening and interventions ers from becoming struggling adolescent for struggling readers; it does not provide readers and reduce unnecessary referrals recommendations for general classroom to special education. reading instruction.

14. Burns, Snow and Griffin (1996).

( 8 ) Checklist for carrying out the Recommendation 3. recommendations Provide intensive, systematic instruction on up to three Recommendation 1. foundational reading skills in small Screen all students for potential groups to students who score below reading problems at the beginning of the benchmark score on universal the year and again in the middle of the screening. Typically, these groups year. Regularly monitor the progress meet between three and five times a of students who are at elevated risk week for 20 to 40 minutes (tier 2). for developing reading disabilities.  Use a curriculum that addresses the  Create a building-level team to fa- components of reading instruction (com- cilitate the implementation of universal prehension, fluency, phonemic awareness, screening and progress monitoring. phonics, and vocabulary) and relates to students’ needs and developmental levels.  Select a set of efficient screening measures that identify children at risk for  Implement this program three to five poor reading outcomes with reasonable times a week, for approximately 20 to 40 degrees of accuracy. minutes.

 Use benchmarks or growth rates (or  Build skills gradually and provide a combination of the two) to identify chil- a high level of teacher-student interac- dren at low, moderate, or high risk for de- tion with opportunities for practice veloping reading difficulties.15 and feedback.

Recommendation 2. Recommendation 4. Provide differentiated reading Monitor the progress of tier 2 instruction for all students based students at least once a month. Use on assessments of students’ current these data to determine whether reading levels (tier 1). students still require intervention. For those students still making  Provide training for teachers on how insufficient progress, school- to collect and interpret student data on wide teams should design a tier 3 reading efficiently and reliably. intervention plan.

 Develop data-driven decision rules  Monitor progress of tier 2 students for providing differentiated instruction to on a regular basis using grade appropri- students at varied reading proficiency lev- ate measures. Progress monitoring should els for part of the day. occur at least eight times during the school year.  Differentiate instruction—including varying time, content, and degree of sup-  While providing tier 2 instruction, use port and scaffolding—based on students’ progress monitoring data to identify stu- assessed skills. dents needing additional instruction.

 Consider using progress monitoring data to regroup tier 2 students approxi- 15. Schatschneider (2006). mately every six weeks.

( 9 ) Checklist for carrying out the recommendations

Recommendation 5. Provide intensive  Schedule multiple and extended in- instruction on a daily basis that structional sessions daily. promotes the development of the various components of reading  Include opportunities for extensive prof iciency to students who show practice and high-quality feedback with minimal progress after reasonable one-on-one instruction. time in tier 2 small group instruction (tier 3).  Plan and individualize tier 3 instruction using input from a school-based RtI  Implement concentrated instruction team. that is focused on a small but targeted set of reading skills.  Ensure that tier 3 students master a reading skill or strategy before moving on.  Adjust the overall lesson pace.

( 10 ) Recommendation 1. grades 1 and 2 to predict students’ reading performance in subsequent years.16 Screen all students However, it should be cautioned that few of for potential reading the samples used for validation adequately problems at the represent the U.S. population as required by the Standards for Educational and Psy- beginning of the chological Testing.17 The evidence base year and again in the in kindergarten is weaker, especially for measures administered early in the school middle of the year. year.18 Thus, our recommendation for kin- Regularly monitor the dergarten and for grade 1 is to conduct a second screening mid-year when results progress of students tend to be more valid.19 who are at elevated Brief summary of evidence risk for developing reading disabilities. The panel recommends a series of screening measures be employed to assess proficiency in several key areas (see Table 3). Universal screening is a critical first Five correlational studies have demon- step in identifying students who strated that certain types of measures can are at risk for experiencing reading be used to accurately predict future student difficulties and who might need more performance.20 Tests conducted by the As- instruction. Screening should take sessment Committee (2002) demonstrate place at the beginning of each school that these measures meet the standards for year in kindergarten through grade educational and psychological testing21 in 2. Schools should use measures that terms of internal consistency and temporal are efficient, reliable, and reasonably valid. For students who are at risk for reading difficulties, progress in reading and reading related-skills 16. Compton, Fuchs, Fuchs, and Bryant (2006); Mc- should be monitored on a monthly Cardle, Scarborough, and Catts (2001); O’Connor or even a weekly basis to determine and Jenkins (1999); Scarborough (1998a); Fuchs, Fuchs, and Compton (2004); Speece, Mills, Ritchey, whether students are making adequate and Hillman (2003b). progress or need additional support 17. American Education Research Association, (see recommendation 4 for further American Psychological Association, and Na- detail). Because available screening tional Council on Measurement in Education measures, especially in kindergarten (1999). and grade 1, are imperfect, schools 18. Jenkins and O’Connor (2002); O’Connor and Jenkins (1999); Scarborough (1998a); Torgesen are encouraged to conduct a second (2002); Badian (1994); Catts (1991); Felton screening mid-year. (1992). 19. Compton et al. (2006); Jenkins, Hudson, and Level of evidence: Moderate Johnson (2007). 20. Compton et al. (2006); McCardle, Scarbor- The panel judged the level of evidence for ough, and Catts (2001); O’Connor and Jenkins recommendation 1 to be moderate. This rec- (1999); Scarborough (1998a); Fuchs, Fuchs, and Compton (2004); Speece et al. (2003b). ommendation is based on a series of high- 21. American Education Research Association, quality correlational studies with replicated American Psychological Association, and Na- findings that show the ability of measures tional Council on Measurement in Education of reading proficiency administered in (1999).

( 11 ) 1. Screen all students for potential reading problems stability.22 While the panel is not recom- How to carry out this mending which specific measure should be recommendation adopted in each school, the panel does recommend that students are screened with 1. Create a building-level team to facilitate measures that have properties similar to the implementation of universal screening those examined in these studies. and progress monitoring.

In our review of evidence, we detected In the opinion of the panel, a building-level problems with commonly used measures RtI team should focus on the logistics of im- in terms of their ability to correctly iden- plementing school-wide screening and sub- tify children at low risk for experiencing sequent progress monitoring, such as who problems (known as specificity). That is, administers the assessments, scheduling, the measures tend to consistently over- and make-up testing, as well as substantive identify students as needing assistance.23 issues, such as determining the guidelines We also noted a paucity of cross-validation the school will use to determine which studies.24 Nonetheless, the extensive body students require intervention and when of replicated correlational research sup- students have demonstrated a successful ports our conclusion that these are reason- response to tier 2 or tier 3 intervention. able batteries of measures to use for early Although each school can develop its own screening, particularly in grades 1 and 2. benchmarks, it is more feasible, especially during the early phases of implementation, for schools to use guidelines from national databases (often available from publishers, from research literature, or on the Office of Special Education Programs (OSEP) Progress Monitoring and RtI websites25).

2. Select a set of efficient screening measures that identify children at risk for poor reading outcomes with reasonable accuracy.

22. Coefficient alpha estimates are .84 for grade 1 As children develop, different aspects of letter sound knowledge, .80 for grade 1 phoneme blending, and .85 and .83 for grade 1 and 2 word reading or reading-related skills become reading on the Texas Primary Reading Inventory most appropriate to use as screening mea- (1999). Coefficient alpha estimates are .92 and sures. Table 3 highlights the skills most .91 for 6 and 7 year old children on the elision appropriate for each grade level. Some con- measure and .89 and .86 for 6 and 7 year old children on the sound matching measure on the troversy remains about precisely which one Comprehensive Test of Phonological Processing skill is best to assess at each grade level. For (Wagner, Torgeson, and Rashotte 1999). Alternate that reason, we recommend the use of two test-form and stability coefficients exceed .90 in screening measures at each juncture. grade 1 for the word identification fluency task (Compton et al. 2006). For the DIBELS measures alternative-form reliability estimate for grade 1 Table 3 also outlines some commonly letter naming fluency, .86 for grade 1 non-word used screening measures for kindergarten fluency it is .83, and .90 for grade 2 oral reading through grade 2 highlighting their focus, fluency (Good and Kaminski 2003). purpose, and limitations. The limitations 23. Foorman, Fletcher, Francis, Schatschneider, are based on the opinion of the panel.26 and Mehta (1998); O’Connor and Jenkins (1999); Jenkins and O’Connor (2002); McCardle, Scarbor- ough, and Catts (2001). 24. Compton et al. (2006); O’Connor and Jenkins 25. See http://www.rti4success.org/ or http:// (1999); Foorman et al. (1998). www.studentprogress.org/.

( 12 ) 1. Screen all students for potential reading problems

Table 3. Re commended target areas for early screening and progress monitoring

Recommended Proficiencies Measures Purpose Limitations grade levels assessed

Letter naming K–1 Letter name Screening This measure is poor for fluency identification progress monitoring since and the ability students begin to learn to to rapidly associate letters with sounds. retrieve abstract It is not valid for English information learners in kindergarten, but seems valid for grade 1.

Phoneme K-1 Phonemic Screening This measure is problematic Segmentation awareness and progress for measuring progress in monitoring the second semester of grade 1. As students learn to read, they seem to focus less on phonemic skills and more on decoding strategies.

Nonsense word 1 Proficiency and Screening This measure is limited to fluency automaticity and progress only very simple words and with basic monitoring does not tap the ability to phonics rule read irregular words or multisyllabic words.

Word 1–2 Word reading Screening This measure addresses many identification26 and progress of the limitations of nonsense monitoring word fluency by including multisyllabic and irregular words.

Oral reading 1–2 Reading con- Screening Although the measure has fluency nected text and progress moderately strong criterion- (also called accurately and monitoring related validity, it cannot give passage reading fluently a full picture of students’ fluency) reading proficiency. Many students will score close to zero at the beginning of grade 1. The measure still is a reasonable predictor of end of year reading performance.

Source: Authors’ compilation based on Fuchs, Fuchs, Thompson, Al Otaiba, Yen, Yang, Braun, and O’Connor (2001b), Speece et al. (2003b); Schatschneider (2006); O’Connor and Jenkins (1999); and Baker and Baker (2008) for letter naming fluency. For phoneme segmentation, O’Connor and Jenkins (1999). For nonsense word fluency, Speece et al. (2003b); Good, Simmons, and Kame’enui (2001). For word identification, Fuchs, Fuchs, and Compton (2004); Compton et al. (2006). For oral reading fluency, Fuchs, Fuchs, Hosp, and Jenkins (2001a); Fuchs, Fuchs, and Maxwell (1988); Schatschneider (2006); Speece and Case (2001); Gersten, Dimino, and Jayanthi (2008); Baker, Gersten, Haager, and Dingle (2006).

26. Fuchs et al. (2004); Compton et al. (2006)

( 13 ) 1. Screen all students for potential reading problems

Kindergarten screening batteries should on Progress Monitoring and Response to include measures assessing letter knowl- Intervention.31 edge, phonemic awareness, and expressive and receptive vocabulary.27 Unfortu- Predictive validity is an index of how well nately, efficient screening measures for the measure provides accurate informa- expressive and receptive vocabulary are in tion on future reading performance of their infancy. students—and thus is critical. In the opinion of the panel, predictive validity should As children move into grade 1, screening reach an index of 0.60 or higher. batteries should include measures assessing phonemic awareness, decoding, word Reducing the number of false positives identification, and text reading.28 By the identified—students with scores below the second semester of grade 1 the decod- cutoff who would eventually become good ing, word identification, and text reading readers even without any additional help— should include speed as an outcome.29 is a serious concern. False positives lead Grade 2 batteries should include measures to schools providing services to students involving word reading and passage read- who do not need them. In the view of the ing. These measures are typically timed. panel, schools should collect information on the sensitivity of screening measures Despite the importance of vocabulary, lan- and adjust benchmarks that produce too guage, and comprehension development many false positives. There is a tradeoff, in kindergarten through grade 2, very few however, with the specificity of the mea- research-validated measures are available sure and its ability to correctly identify for efficient screening purposes. But -di 90 percent or more of students who re- agnostic measures can be administered ally do require assistance.32 Using at least to students who appear to demonstrate two screening measures can enhance the problems in this area. accuracy of the screening process; however, decision rules then become more Technical characteristics to consider complex.

The panel believes that three characteris- Costs in both time and personnel should tics of screening measures should be ex- also be considered when selecting screen- amined when selecting which measures ing measures. Administering additional (and how many) will be used. measures requires additional staff time and may displace instruction. Moreover, Reliability of screening measures (usually interpreting multiple indices can be a com- reported as internal consistency reliabil- plex and time-consuming task. Schools ity or Cronbach’s alpha) should be at least should consider these factors when se- 0.70.30 This information is available from lecting the number and type of screening the publishers’ manual or website for the measures. measure. Soon this information will be posted on the websites for National Center

27. Jenkins and O’Connor (2002); McCardle, Scar- borough, and Catts (2001); O’Connor and Jenkins (1999); Scarborough (1998a); Torgesen (2002). 28. Foorman et al. (1998). 31. See http://www.rti4success.org/ or http:// 29. Compton et al. (2006); Fuchs et al. (2004). www.studentprogress.org/. 30. Nunnally (1978). 32. Jenkins (2003).

( 14 ) 1. Screen all students for potential reading problems

3. Use benchmarks or growth rates (or a Roadblock 1.2. Universal screening falsely combination of the two) to identify children identifies too many students. at low, moderate, or high risk for developing reading difficulties.33 Suggested Approach. Selecting cut-points that accurately identify 100 percent of the Use cut-points to distinguish between stu- children at risk casts a wide net—also iden- dents likely to obtain satisfactory and un- tifying a sizeable group of children who satisfactory reading proficiency at the end will develop normal reading skills. We rec- of the year without additional assistance. ommend using universal screening mea- Excellent sources for cut-points are any sures to liberally identify a pool of chil- predictive validity studies conducted by dren that, through progress monitoring test developers or researchers based on methods, can be further refined to those normative samples. Although each school most at risk.35 Information on universal district can develop its own benchmarks screening and progress monitoring mea- or cut-points, guidelines from national da- sures can be found at the National Center tabases (often available from publishers, on Student Progress Monitoring or the Iris from research literature, or on the OSEP, Center at Vanderbilt University.36 Progress Monitoring, and RtI websites34) may be easier to adopt, particularly in the Roadblock 1.3. Some students might get early phases of implementation. “stuck” in a particular tier.

As schools become more sophisticated in Suggested Approach. If schools are re- their use of screening measures, many sponding to student performance data will want to go beyond using benchmark using decision rules, students should not assessments two or three times a year and get stuck. A student may stay in one tier use a progress monitoring system. because the instructional match and learning trajectory is appropriate. To ensure Roadblocks and suggested students are receiving the correct amount approaches of instruction, schools should frequently reassess—allowing fluid movement across Roadblock 1.1. It is too hard to establish tiers. Response to each tier of instruction district-specific benchmarks. will vary by student, requiring students to move across tiers as a function of their Suggested Approach. National bench- response to instruction. The tiers are not marks can assist with this process. It often standard, lock-step groupings of students. takes a significant amount of time to estab- Decision rules should allow students show- lish district-specific benchmarks or stan- ing adequate response to instruction at tier dards. By the time district-specific bench- 2 or tier 3 to transition back into lower marks are established, a year could pass tiers with the support they need for con- before at-risk readers are identified and tinued success. appropriate instructional interventions begin. National standards are a reasonable alternative to establishing district-specific benchmarks.

33. Schatschneider (2006). 35. Compton et al. (2006). 34. See http://www.rti4success.org/ or http:// 36. See http://www.studentprogress.org/ or www.studentprogress.org/. http://iris.peabody.vanderbilt.edu/.

( 15 ) 1. Screen all students for potential reading problems

Roadblock 1.4. Some teachers place stu- to determine the confidence interval for dents in tutoring when they are only one each benchmark score. If a students’ score point below the benchmark. falls within the confidence interval, either conduct an additional assessment of those Suggested Approach. No measure is per- students or monitor their progress for a fectly reliable. Keep this in mind when stu- period of six weeks to determine whether dents’ scores fall slightly below or above a the student does, in fact, require addi- cutoff score on a benchmark test. The panel tional assistance.37 recommends that districts and schools review the assessment’s technical manual 37. Francis et al. (2005).

( 16 ) Recommendation 2. skills, readers may wonder where differentiated instruction ends and tier Provide differentiated 2 intervention begins. Differentiated reading instruction instruction applies to all students, while tier 2 instruction applies only to those for all students based at risk in key areas. The panel believes on assessments of that, to be effective, a multi-tier approach can blur the lines between students’ current tier 1 and tier 2, and that sensible data- reading levels (tier 1). driven instruction should permeate all of the tiers of reading instruction.

Ideally, classroom reading instruction Level of evidence: Low would be evidence based. However, research that might provide a clear, The panel judged the level of evidence for comprehensive model of how to teach this recommendation as low. A correla- reading to students in the primary tional study demonstrated that the more grades is lacking.38 The purpose of this teachers used assessment information, the recommendation is to discuss classroom greater their students’ reading skill growth reading instruction as it relates to in grade 1.39 RtI and effective tier 1 instruction. In particular, we focus on the use of Brief summary of evidence assessment data to guide differentiated reading instruction. Tier 1 provides the One descriptive-correlational study exam- foundation for successful RtI overall, ined how student reading growth varied by without which too many students would the degree to which teachers employed a fall below benchmarks. specific differentiation program. This differentiation program relied on assessments to group students. Student reading growth The panel recommends differentiating was higher for teachers who implemented instruction in tier 1. For example, the program with greater fidelity. during independent work time, students weak in vocabulary can How to carry out this practice vocabulary with a partner or recommendation in small groups, while other students form teams to brainstorm character 1. Provide training for teachers on how to traits and motivations for the main collect and interpret student data on read- characters in the story they are reading ing efficiently and reliably. that week. Data from the various screening and progress monitoring Provide training on how to use diagnos- measures in recommendation 1 should tic measures, especially measures for also serve a role in orchestrating those students experiencing difficulty. differentiated instruction. Informal assessments can help educators make better informed decisions. For example, listening to how a student reads a Because differentiated instruction text that is slightly too difficult can yield under tier 1 requires identifying and grouping students to work on targeted 39. Connor, Piasta, Fishman, Glasney, Schat- schneider, Crowe, Underwood, and Morrison 38. National Reading Panel (2000). (2009).

( 17 ) Recommendation 2. Provide differentiated reading instruction for all students useful information and is easily embedded Roadblocks and suggested within lessons. Teachers can ask a student approaches to summarize a story they just read. This exercise will reveal how well the student Roadblock 2.1. It is difficult for teach- comprehends what they read. Listening to ers to interpret assessment results and the student’s summary of the story can subsequently use the information for also reveal other information—for exam- instruction. ple about the student’s own life or what they know of other books.40 Suggested Approach. The panel recommends providing professional develop- 2. Develop data-driven decision rules for pro- ment focused on how to administer as- viding differentiated instruction to students sessments, interpret the results, and use at varied reading proficiency levels for part the information. This should be ongoing. of the day. With proper training, teachers’ instruction may be more effective. According to the panel, independent silent reading activities should be gradu- Roadblock 2.2. Using multiple small ally increased as reading skills improve. groups is difficult when some children have Data on student performance (a measure difficulty paying attention, working inde- of word identification fluency or fluency pendently, and interacting with peers. in reading connected text) should inform this decision. For many grade 1 students, Suggested Approach. Classroom man- independent silent reading time would be agement procedures should be firmly in minimal during the first few months of the place during reading instruction. To facili- year. Student-managed activities should tate effective reading instruction, adminis- be introduced gradually and should focus trators should provide teachers with sup- only on skills students have mastered. portive efforts and motivational strategies, especially in managing independent and 3. Differentiate instruction—including varying small group work. time, content, and degree of support and scaffolding—based on students’ assessed skills.

The panel believes that as students fall below grade expectations, more time in explicit instruction provided by the teacher in small groups is critical to bring their skills to grade level. The panel suggests independent or group work, such as independent silent reading or buddy reading, are more effective when they are gradually increased as student reading skills improve.

40. Snow (2001).

( 18 ) Recommendation 3. vations.41 These studies on supplemental instruction in reading support tier 2 Provide intensive, intervention as a way to improve read- systematic instruction ing performance in decoding. Six studies on up to three showed positive effects on decoding,42 and four showed effects on both decoding foundational reading and reading comprehension.43 Six studies skills in small groups involved one-on-one instruction,44 and the remainder used small groups rang- to students who score ing from two to five students. Given that below the benchmark effect sizes were not significantly higher for the one-on-one approach, small group on universal screening. work could be considered more practical Typically, these for implementation. groups meet between Brief summary of evidence three and five times The 11 studies that met WWC standards or a week for 20 to 40 that met WWC standards with reservations minutes (tier 2). suggest that educators should emphasize the critical reading skills of phonemic awareness, decoding, reading compre- Tier 2 instruction should take place hension, and fluency at appropriate grade in small homogenous groups ranging levels. Two of five studies that measured from three to four students using phonemic awareness demonstrated sig- curricula that address the major nificant effects.45 Five of nine studies that components of reading instruction measured decoding demonstrated signifi- (comprehension, fluency, phonemic cant effects, and students showed positive awareness, phonics, and vocabulary). The areas of instruction are based 41. Ebaugh (2000); Gunn, Biglan, Smolkowski, on the results of students’ scores on and Ary (2000); Mathes, Denton, Fletcher, An- universal screening. Instruction should thony, Francis, and Schatschneider (2005); Jen- be systematic—building skills gradually kins, Peyton, Sanders, and Vadasy (2004); Lennon and introducing skills first in isolation and Slesinski (1999); Vaughn, Mathes, Linan- Thompson, Cirino, Carlson, Pollard-Durodola, and then integrating them with other Cardenas-Hagan, and Francis (2006); Vadasy, skills. Explicit instruction involves more Sanders, and Peyton (2005); Ehri, Dreyer, Flug- teacher-student interaction, including man, and Gross (2007); Gibbs (2001); McMaster, frequent opportunities for student Fuchs, Fuchs, and Compton (2005); Vadasy, Jen- practice and comprehensible and kins, Antil, Wayne, and O’Connor (1997). specific feedback. Intensive instruction 42. Ebaugh (2000); Gunn et al. (2000); Jenkins et al. (2004); Lennon and Slesinski (1999); Va- should occur three to five times per dasy, Sanders, and Peyton (2005); Vaughn et al. week for 20 to 40 minutes. (2006). 43. Gunn et al. (2000); Jenkins et al. (2004); Va- Level of evidence: Strong dasy, Sanders, and Peyton (2005); Vaughn et al. (2006). The panel judged the evidence support- 44. Gunn et al. (2000); McMaster et al. (2005); Va- ing this recommendation as strong based dasy et al. (1997); Vadasy, Sanders, and Peyton (2005); Jenkins et al. (2004); Gibbs (2001). on 11 studies that met WWC standards 45. Ehri et al. (2007); Lennon and Sleskinski or that met WWC standards with reser- (1999).

( 19 ) 3. Provide intensive, systematic instruction on up to three foundational reading skills

effects in five of seven studies46 that mea- students may need further intervention. sured reading comprehension. Only one After five weeks, some students may have study found significant effects in reading caught up. fluency. Vocabulary was the least examined outcome of the 11 studies, with only In choosing an intervention program for 1 study measuring and finding effects on tier 2, administrators should look for vocabulary knowledge.47 programs—either commercially available intervention curricula, commercially Since 7 of the 11 studies that met WWC developed supplemental curricula, or standards or that met standards with res- intervention programs—that are com- ervations produced a significant effect patible with their school’s core reading on at least one reading outcome, and all program and that provide intensive small seven studies used explicit instruction, group instruction in three to four founda- we concluded that explicit instruction tional skills. Ideally, the intervention pro- is an effective approach to use in tier 2 gram has demonstrated its effectiveness intervention.48 through independent evaluations using rigorous experimental or quasi-experi- How to carry out this mental designs. recommendation The intervention curriculum should teach 1. Use a curriculum that addresses the com- and build foundational skills to mastery ponents of reading instruction (phonemic and incorporate some complex reading awareness, phonics, vocabulary, comprehen- skills. Specific components vary by grade sion, and fluency) and relates to students’ level and reflect the changing developmen- needs and developmental level. tal emphasis at different stages in reading. Table 4 highlights the foundational read- Tier 2 intervention curricula are some- ing skills students should develop in kin- times called standard protocols. Standard dergarten through grade 2. Skills validated protocols are tutoring programs taught to by research are indicated by table notes. all students scoring below benchmark.49 The remaining skill areas are considered These “one size fits all” programs address critical by the panel. foundational skills and strategies that are essential to learning to read. The panel The critical skill for kindergarteners to suggests that schools should use interven- master is the ability to segment phonemes, tion programs to provide tier 2 instruction a key indicator of future success or failure for all students scoring below benchmark in reading.50 Also important are letter- for at least five weeks to discern which sound identification, the alphabetic principle (the recognition of the relationship between spoken sounds and letters), and 46. Vadasy, Sanders, and Peyton (2005); Jen- beginning decoding skills (blending writ- kins et al. (2004); Vaughn et al. (2006); Ehri et al. (2007). ten letters into words). Students who can 47. Gunn et al. (2000). perform these tasks understand the pho- 48. Gunn et al. (2000); Jenkins et al. (2004); Ehri nemic elements in words leading to accu- et al. (2007); Ebaugh (2000); Vadasy, Sanders, and rate and fluent decoding.51 Peyton (2005); Vaughn et al. (2006). 49. There are some obvious exceptions, such In general, during the first semester, as students already identified as students with grade 1 students who participate in tier 2 significant cognitive disabilities, students who already have Individualized Education Programs in reading or language involving a much more 50. Lennon and Slesinski (1999). basic curriculum. 51. Gunn et al. (2000).

( 20 ) 3. Provide intensive, systematic instruction on up to three foundational reading skills

Table 4. Foundational reading skills in grades K–2 Grade Skill Kindergarten Phonemic awarenessa Letter soundsb Listening comprehension Vocabulary development Grade 1 Phonemic awarenessc Phonicsd Fluency (high frequency words) Fluency with connected text (second half of the year)e Vocabularyf Comprehensiong Grade 2 Phonicsh Fluency with connected text Vocabularyi Comprehension a. Lennon and Slesinski (1999). b. Lennon and Slesinski (1999). c. Ehri et al. (2007). d. Gunn et al. (2000); Jenkins et al. (2004); Ehri et al. (2007); Mathes et al. (2005); Vadasy, Sanders, and Peyton (2005). e. Ehri et al. (2007). f. Gunn et al. (2000). g. Jenkins et al. (2004); Ehri et al. (2007); Mathes et al. (2005); Vadasy, Sanders, and Peyton (2005); Vaughn et al. (2006). h. Gunn et al. (2000). i. Gunn et al. (2000). Source: Authors’ compilation based on information described in the text. interventions will need instruction in pho- Phonics interventions for grade 2 students nics (decoding one and then two syllable concentrate on learning more difficult words) and fluency. Since these are be- skills, such as digraphs (oa as in boat and ginning readers, fluency instruction dur- ch as in child), diphthongs (ew as in stew, ing the first semester is taught by first oi as in soil), and controlled R (ar as in focusing on fluently and accurately read- car, ur as in fur). These interventions ad- ing short lists of high frequency words. dress structural analysis skills that focus During the second semester, as students on prefixes, suffixes, forming plurals, and move into reading connected text, inter- adding -ed and -ing to form past and pro- ventions focusing on reading accurately, gressive tenses. Students also apply pho- fluently, and with prosody (proper ex- netic skills to words with more than one pression) should be added. Some grade syllable. Fluency should continue to be 1 students will still need intensive and emphasized.53 usually more accelerated instruction in phonemic awareness (blending and seg- Some intervention curricula will include menting sounds) and basic phonics (letter what the panel believes are important ac- sound correspondence) interventions to tivities: literal comprehension (questions increase their understanding of the alpha- whose answers are stated in the text), more betic principle.52 sophisticated comprehension strategies (summarizing a portion of text), listening comprehension strategies, spelling, ex- 52. Gunn et al. (2000); McMaster et al. (2005); Jenkins et al. (2004); Vaughn et al. (2006); Ehri et al. (2007). 53. Gunn et al. (2000).

( 21 ) 3. Provide intensive, systematic instruction on up to three foundational reading skills

pressive writing, and read-alouds. Literal the same intervention—is not effective.56 comprehension and some rudimentary But according to Harn, Linan-Thompson, comprehension instruction occur in many and Roberts (2008), doubling instructional of the successful interventions, and so are time while changing the percentage of recommended.54 Other elements, such as time allotted to each instructional area inferential comprehension and vocabulary in response to students’ changing needs development, may be better developed resulted in better outcomes on timed oral with more heterogeneous groups during reading fluency and word reading mea- the reading language arts block. It is the sures for students. opinion of the panel that an intervention curriculum that covers five to six skills per 3. Build skills gradually and provide a high day may not provide the intensity neces- level of teacher-student interaction with op- sary to improve reading achievement. portunities for practice and feedback.

2. Implement this program three to five Reading instruction should be system- times a week, for approximately 20 to 40 atic—building skills gradually and intro- minutes. ducing skills first in isolation and then by integrating them with other skills to pro- Tier 2 instruction should be implemented vide students practice and to build gen- for 20 to 40 minutes, three to five times eralization.57 Students should be given per week in small groups of three to four clear, corrective feedback, and cumula- students. Student grade level and needs tive review to ensure understanding and should determine the duration. mastery. For example, in phonics, a critical area in grade 1 tier 2 interventions, a sys- An intervention session can range from 20 tematic curriculum might begin by intro- to 30 minutes for kindergarten students ducing a few of the most frequently used to 40 to 50 minutes for grade 2 students, consonants sounds (m, s, t, b) followed by depending on student needs. Providing a vowel, usually the short a. This allows kindergarten students with 20 minutes of students to integrate these newly learned daily instruction has been demonstrated sounds by blending sounds into words. to have a positive impact on their acquisition of early reading skills, such as pho- Reading instruction should also be ex- nemic awareness and letter-sound corre- plicit. Explicit instruction involves a high spondence.55 As students move into grades level of teacher-student interaction that 1 and 2, the time needed for interventions includes frequent opportunities for stu- usually increases as the skills they need dents to practice the skill and clear, spe- to catch up to their peers without reading cific corrective feedback. It begins with difficulties broaden. overt and unambiguous explanations and models. An important feature of explicit A small body of descriptive evidence sug- instruction is making the thinking process gests that the time spent on each area of public. Thinking aloud should occur dur- instruction might be more important than ing all instructional components of tier the total instructional time. How time is 2 interventions ranging from systematic spent and proportioned appears critical. skill building in phonics to teaching more For example, merely doubling instructional time—providing double doses of 56. Wanzek and Vaughn (2007). 57. Gunn et al. (2002); Vadasy, Sanders, and 54. Vaughn et al. (2006); Gunn et al. (2000). Peyton (2005); Vaughn et al. (2006); Mathes et 55. Gunn et al. (2000); Gunn, Smolkowski, Biglan, al. (2005); Jenkins et al. (2004); McMaster et al. and Black (2002); Lennon and Slesinski (1999). (2005).

( 22 ) 3. Provide intensive, systematic instruction on up to three foundational reading skills

complex and intricate comprehension Roadblock 3.2. Finding an additional 15 strategies (such as summarizing or making to 50 minutes a day for additional reading inferences). When thinking aloud, teachers instruction can be a daunting task. should stop, reflect, and formulate an ex- planation of their thinking processes. Suggested Approach. Schools should first determine who will provide the in- Roadblocks and suggested tervention. If the classroom teacher will approaches provide the intervention, then small group instruction could occur when students Roadblock 3.1. Some teachers or read- are working independently at classroom ing specialists might worry about aligning learning centers. In grade 2 classrooms, the tier 2 intervention program with the where there is non-direct instructional core program. time, intervention lessons can occur at times that do not conflict with other criti- Suggested Approach. Since tier 2 in- cal content areas, such as mathematics, struction relies on foundational (and particularly if a person other than the sometimes prerequisite) skills that are classroom teacher is providing the in- determined by the students’ rate of prog- tervention. There may be situations in ress, it is unlikely that the same skill will schools with reading blocks of two to two be addressed in the core reading instruc- and a half hours where it is appropriate tion at the same time. Alignment is not for students to work at learning stations or as critical as ensuring that instruction is complete assignments while the classroom systematic and explicit and focuses on the teacher is conducting tier 2 interventions, high priority reading components. especially if tier 2 students are unable to complete these assignments.

( 23 ) Recommendation 4. Brief summary of evidence

Monitor the progress One study shows that progress monitoring of tier 2 students at in reading (oral reading fluency or word least once a month. Use identification fluency in grades 1 and 2) increases teachers’ awareness of students’ these data to determine current level of reading proficiency and whether students still has a positive effect on the instructional decisions teachers make.60 Collecting and require intervention. using progress monitoring data is some- For those students still times a component of tier 2 instruction. making insufficient How to carry out this progress, school-wide recommendation teams should design a 1. Monitor progress of tier 2 students on a tier 3 intervention plan. regular basis using grade appropriate measures. Monitoring of progress should occur at least eight times during the school year. Schools should establish a schedule to assess tier 2 students at least Some researchers recommend more fre- monthly—reassigning students who quent weekly assessments for monitoring have met benchmarks, graphing student progress.61 However, little evidence students’ progress in reading in a demonstrates that weekly measures are su- reliable fashion, and regrouping perior to monthly ones.62 Many tier 2 inter- students who need continued vention programs (commercially developed, instructional support.58 researcher developed, or district developed) contain weekly mastery tests that educators Level of evidence: Low can use to guide instruction (to know which skills need to be reviewed and re-taught). Of the 11 randomized controlled trials and quasi-experimental design studies that If a tier 2 program does not include mas- evaluated effects of tier 2 interventions tery checks, monitor students’ progress and that met WWC standards or that met weekly, if possible, but no less than once WWC standards with reservations, only 3 a month. The measures should be effi- reported using mastery checks or prog- cient, reliable, and valid. Many progress ress monitoring in instructional decision- monitoring measures are also useful as making.59 None of the studies demonstrate screening measures (see recommenda- that progress monitoring is essential in tion 1). Progress monitoring measures are tier 2 instruction. However, in the opinion the best way to assess students’ retention of the panel, awareness of tier 2 student of material taught and thus their path to progress is essential for understanding reading proficiency. Table 5 indicates ap- whether tier 2 is helping the students and propriate progress monitoring measures whether modifications are needed. for kindergarten through grade 2.

58. Vaughn, Linan-Thompson, and Hickman (2003). 60. Fuchs, Deno, and Mirkin (1984). 59. McMaster et al. (2005); Vaughn et al. (2006); 61. Fuchs, Deno, and Mirkin (1984). Mathes et al. (2005). 62. Johnson et al. (in press).

( 24 ) 4. Monitor the progress of tier 2 students at least once a month

Table 5. Progress monitoring measures in grades K–2

Grade Measure Kindergarten Phonemic awareness measures (especially measures of phoneme segmentation)

Grade 1 Fluent word recognition Nonword (pseudo word reading) Oral reading fluency (connected text)

Grade 2 Fluent word recognition Oral reading fluency

Source: Authors’ compilation based on information described in text.

2. While providing tier 2 instruction, use Roadblocks and suggested progress monitoring data to identify stu- approaches dents needing additional instruction. Roadblock 4.1. Students within classes It is important that tier 2 instruction ad- are at very different levels for tier 2 vances at a good pace. At the same time, intervention. teaching to mastery is paramount since the skills are foundational for future suc- Suggested Approach. If students within cess in reading. If three students are mak- a class are at such diverse levels as to ing progress and one student is lagging necessitate more than two tier 2 groups, behind, an option to consider is to provide consider grouping students across classes. this student with 10 minutes of review, This will facilitate clustering children with practice, and additional focused instruc- similar needs. In such a case a reading spe- tion on material previously taught. If none cialist, paraprofessional, or other school of the students are making progress, take personnel who have received training can a careful look at the tier 2 intervention— conduct the intervention. it may be missing critical components or moving too fast for the students in tier 2 Roadblock 4.2. There is insufficient to master the target skills. time for teachers to implement progress monitoring. 3. Consider using progress monitoring data to regroup tier 2 students approximately Suggested Approach. If teachers are too every six weeks. busy to assess students’ progress with progress monitoring measures, consider Since students’ skill level changes over using paraprofessionals or other school time and in varying degrees, use prog- staff. Train them how to administer such ress monitoring data to regroup students measures. so that the groups are as homogeneous as possible. Ideally, groups may cut across more than one class, if schedules permit.

( 25 ) Recommendation 5. Level of evidence: Low

Provide intensive The level of evidence for this recommen- instruction on a daily dation is low. Although the panel found basis that promotes five studies on this recommendation that met the WWC standards (or met standards the development of the with reservations), no studies reported various components statistically significant impacts on reading outcomes.63 of reading proficiency to students who show Brief summary of evidence minimal progress Despite over 50 years of research on spe- after reasonable time cial education and remedial instruction, major gaps persist in the knowledge of in tier 2 small group how to teach reading to the 3 to 5 percent instruction (tier 3). of students with the most severe reading difficulties.64 The research reveals little about students whose response to typi- Instruction should be intensified by cally effective interventions is low. There- focusing on fewer high priority reading fore, the material below represents the skills during lessons and scheduling opinion of the panel. multiple and extended instructional sessions. One-on-one or small group How to carry out this instruction also provides intensity as recommendation students have more opportunities to practice and respond. One-on-one 1. Implement concentrated instruction that instruction includes giving students is focused on a small but targeted set of feedback based on their individual reading skills. responses, teaching students to mastery based on individual learning Focusing on a small set of reading or read- progress, and planning instruction with ing-related skills is essential to tier 3 in materials and an instructional sequence kindergarten through grade 2 because that meets individual student needs. having too many instructional objectives for struggling readers makes it more difficult to learn the skills well enough for There is no reason to believe that a proficient reading. 65 In the opinion of the tier 3 program should consist primarily panel, too many instructional objectives of one-on-one instruction—though can overwhelm students. Achieving profi- such instruction should be part of ciency is also difficult for students when a student’s daily program. Student instruction is scattered across different progress should be monitored regularly aspects of reading. using progress monitoring measures to assess whether the program is on course and to determine whether a 63. McMaster et al. (2005); Foorman et al. ��(1998); team of professionals needs to refine Blumsack (1996); Gillon (2000); O’Connor and the instructional program to enhance Jenkins (1995). achievement growth. 64. Torgesen, Wagner, and Rashotte (1997). 65. Blumsack (1996); Foorman et al. (1998); Gil- lon (2000).

( 26 ) 5. Provide intensive instruction on a daily basis that promotes reading proficiency

Diagnostic assessments can help determine how to balance competing demands mine why a reading problem is occurring for instructional time. and which reading skills or performance deficits need to be addressed to improve 2. Adjust the overall lesson pace. reading performance. Specifically, educators can ask: what aspects of reading are To provide greater focus to tier 3 instruc- blocking the student from achieving read- tion, teachers can adjust the overall lesson ing proficiency? When these obstacles are pace so that it is slow and deliberate (that determined, high priority skills are identi- is, more intensive). Teachers implement- fied as the focus of tier 3 instruction. ing tier 3 instruction can focus the pace of lessons by focusing on a single com- For example, the panel believes that if a ponent of a lesson. For example, teachers student is struggling with decoding, it does might focus only on introducing the new not make sense to use tier 3 instructional skill rather than implementing a full les- time for summary writing, comprehension son that includes introduction, extended monitoring instruction, or clarification practice, and application. Subsequent tier strategies because the primary reading 3 instruction might review the new skills obstacle for the student is sounding out (with modified or shortened instruction and reading words accurately. Here, de- from the lesson’s introduction) and prac- coding is considered a high priority skill tice the new skills. Instructional pace is because it underlies the student’s overall slowed and focused by implementing a reading difficulty. series of lessons concentrating only on a variety of review and practice activities. Additionally, the panel believes that there Rather than practicing how to identify the should be depth in modeling and practice main idea in one lesson, several lessons with feedback in tier 3 instruction—per- would practice identifying the main idea. haps requiring limited breadth. Such focus provides opportunities to review, practice, 3. Schedule multiple and extended instruc- and reinforce newly learned proficien- tional sessions daily. cies so that students can demonstrate sustained and consistent levels of profi- While research does not suggest a specific ciency across lessons. Often a sustained number of intervention sessions or dura- 90 percent or higher criterion of correct tion of instructional intervention (such as responses on taught material is consid- weeks, months, or years) for tier 3, stud- ered mastery. ies do suggest that students needing tier 3 intervention require more reading in- Tier 3 instruction often focuses on pho- structional time than their peers without nemic awareness and decoding, espe- reading difficulties. On average, students cially for younger students or those with participating in tier 3 interventions receive very limited reading proficiency. However, an additional 75 minutes of instruction per comprehension and vocabulary are also week. Additional instructional time ranges critical.66 For a student receiving tier 3 from about 45 minutes per week67 to 120 instruction, several sessions each week minutes per week.68 might focus on phonemic awareness and decoding in depth. The other sessions In the opinion of the panel, schools could might focus on comprehension and vo- provide an additional 30 minutes of in- cabulary in depth. To date, there are no struction by creating a “double dose” of clear-cut empirical guidelines to deter- 67. Blumsack (1996). 66. National Reading Panel (2000). 68. Gillon (2000).

( 27 ) 5. Provide intensive instruction on a daily basis that promotes reading proficiency

reading time for struggling readers. Rather and practice. When working with small than more of the same, a double dose of groups, educators can increase opportuni- instruction means a teacher might intro- ties to respond and practice by encourag- duce skills during the first session and ing unison group responses. then re-teach with added practice during the second. With one-on-one and small-group instruction, teachers can also provide immedi- Duration, or extended implementation of ate and individualized feedback.70 A key tier 3 intervention, also intensifies instruc- feature of instructional feedback is error tion. Further research is required to exam- correction. By correcting student errors ine the total hours of instruction needed when they are first made, it is much less and relative impact of tier 3 duration. likely that errors will become internalized and therefore repeated. For example, if a 4. Include opportunities for extensive prac- student incorrectly segmented a word, tice and high quality feedback with one-on- the teacher could model the accurate re- one instruction. sponse, give the student another opportunity to segment the word, and return To become proficient in the application of to the missed word later in the lesson to newly acquired skills and strategies, stu- reinforce the correct application of the dents with the most intensive instructional skill. This type of ongoing, guided practice needs will need multiple opportunities provides students with the support and to practice with immediate high-quality feedback they need to become fluent with feedback. According to panel opinion, tier critical reading skills and strategies. 3 students might require 10 or 30 times as many practice opportunities as their 5. Plan and individualize tier 3 instruction peers. An example considered by the panel using input from a school-based RtI team. includes the use of technology for aspects of the reading program. Technology can In the opinion of the panel, tier 3 instruc- be a good means for students to receive tional planning requires an increased level the practice they need, such as practice in of detail because of the individualized letter sound recognition.69 nature of the instruction and particular student reading needs. Students with in- One-on-one instruction is an effective way tensive reading needs require substantial to maximize practice during tier 3 instruc- supports during the initial stages of learn- tion. If scheduling one-on-one instruc- ing. As students progress in their under- tional sessions is not possible, the panel standing and knowledge, these supports suggests students be organized in small are gradually withdrawn so that students groups of homogenous reading needs. can begin to apply skills and strategies One-on-one or small-group instruction independently.71 For students with learn- provides the greatest opportunity for con- ing disabilities, instruction that is care- tinuous and active learning. For example, fully scaffolded is essential to success- in whole-class instruction, individual stu- ful learning.72 Teachers should introduce dents have few opportunities to respond, concepts and skills beginning with easier practice, and interact with the teacher. tasks and progressing to more difficult Meanwhile in one-on-one instruction, a student has many occasions to respond 70. Blumsack (1996); Gillon (2000); McMaster et al. (2005); O’Connor and Jenkins (1995). 69. Barker and Torgesen (1995); Chambless 71. Blumsack (1996); Foorman et al. (1998); Gillon and Chambless (1994); National Reading Panel (2000); O’Connor and Jenkins (1995). (2000). 72. Swanson, Hoskyn, and Lee (1999).

( 28 ) 5. Provide intensive instruction on a daily basis that promotes reading proficiency

tasks.73 When teaching oral segmenting, activities when a student reaches 100 for example, it is easier for students to percent accuracy on all of the items in isolate the first sound than to completely the activity. Teachers can keep notes or segment the word. records about how students perform on different reading tasks. For example, a Material supports also play a role in in- teacher could record the exact words that dividualizing student learning. Graphic a student practices reading, the student’s organizers, procedural facilitators (like word reading accuracy, and the number color-coded question cards representing of times it takes for students to practice a questions to ask before, during, and after word before reading it accurately.76 reading), and concrete manipulatives are all visual prompts or reminders that pro- Roadblocks and suggested vide support to struggling readers as they approaches internalize skills and strategies. For example, a story map can be used to teach stu- Roadblock 5.1. The distinction between dents how to identify a story’s critical com- tier 2 and tier 3 instructional interventions ponents. As students become more adept can often be blurry. at applying segmentation skills or using a story map to aid retelling, these material Suggested Approach. Teachers should prompts are progressively faded out. not be too concerned about tier 2 and tier 3 differences; the tiers are merely a way Teachers can optimize limited instruc- to continually vary resources to match the tional time and instruction by teaching nature and intensity of instructional need. skills or strategies that reinforce each Remember that at present, distinctions be- other. For example, emerging research tween tier 2 and tier 3 are not clear or well suggests that teaching spelling promotes documented. The terms are conveniences reading for struggling readers.74 Students for school personnel. see spellings as maps of phonemic content rather than an arbitrary sequence of Many tier 3 students will also have tier letters. Practice in using the alphabetic 1 and tier 2 instruction as part of their strategy to spell words seems to transfer reading program. A student receiving tier to reading words. 3 instruction focused on decoding and fluency might also participate in a tier 2 het- 6. Ensure that tier 3 students master a read- erogeneous group focused on vocabulary ing skill or strategy before moving on. and comprehension. One limitation with individualized, one-on-one tier 3 instruc- Emerging research on tier 3 instruction tion is that there are few opportunities focuses on individualizing instruction for students to engage in comprehension- by teaching students to mastery. Before building discourse. Increasing comprehen- a student moves to the next lesson, skill, sion through discourse requires different or activity, they must demonstrate that a levels of student language, vocabulary, reading skill or strategy is mastered. When and comprehension skills. Small, hetero- teaching a series of phonemic awareness geneous groups are optimal for building activities,75 teachers should discontinue student language and vocabulary because students have opportunities to hear differ- 73. Blumsack (1996); Foorman et al. (1998); Gil- ent language examples, new vocabulary lon (2000); McMaster et al. (2005); O’Connor and words, and content that helps connect un- Jenkins (1995). derstanding. Discourse-based vocabulary 74. O’Connor and Jenkins (1995). 75. Gillon (2000). 76. O’Connor and Jenkins (1995).

( 29 ) 5. Provide intensive instruction on a daily basis that promotes reading proficiency

and comprehension activities are often planning and individualizing instruction, included in tier 2 interventions. scaffolding skill introduction, enhancing one reading skill or strategy with another Roadblock 5.2. Because most tier 3 stu- (such as adding spelling to reading in- dents have problems with decoding and struction), structuring multiple practice fluently reading connected text, some may opportunities, and providing high quality have tier 3 interventions that only highlight feedback with consistent error corrections, these areas. professional development plans and ongoing mentoring should focus on the details Suggested Approach. Targeting impor- of instructional design and planning. tant comprehension proficiencies (summarizing, use of story grammar elements, Roadblock 5.5. Adding multiple and ex- vocabulary development, listening com- tended instructional sessions to a daily prehension development) need to be part schedule can be overwhelming for some of any solid tier 3 intervention. students and a challenge for schools in terms of scheduling. Roadblock 5.3. School and staff resources are often too limited to support individual- Suggested Approach. If a student re- ized instruction for tier 3 students. quires an additional hour of instruction per day, teachers should consider break- Suggested Approach. Consider creative ing that hour into smaller instructional alternatives for school and staff resources. sessions or using several short activities For example, use community resources, to help maintain student motivation and such as parent or senior citizen volunteers, engagement.77 One session could focus to help reinforce tier 3 instruction. While on decoding, and the follow-up on com- an experienced teacher or interventionist prehension and vocabulary. The morning should teach new skills, volunteers can session could introduce new word reading help reinforce and practice reading in fo- skills, and the afternoon session practice cused one-on-one sessions. Community and review. tutoring programs are also options. Tech- nology is another resource to consider, Early reading provides critical founda- and remember many individualized in- tional skills; such skills and strategies struction activities work well with small, need to be proficient before students enter homogeneous group instruction. the upper elementary grades. Thus, if critical decisions need to be made about add- Roadblock 5.4. Schools tend to give the ing tier 3 instruction to a student’s reading least experienced teachers the toughest-to- program, using time typically allocated teach students. to social studies or science may be necessary. Other less intrusive scheduling op- Suggested Approach. Reevaluate school tions include providing tier 3 instruction schedules to ensure that the more expe- to struggling readers while other students rienced teachers or specialists are teach- are participating in center activities, in- ing tier 3 instruction. This may require dependent projects, or the tier 1 “add-on” some professional development and on- enrichment activities. Tier 3 instruction going mentoring even for skilled veteran could also be provided during whole-class teachers. spelling instruction.

Many teachers do not have the training to teach students with intensive reading difficulties. Given the importance of carefully 77. Gillon (2000).

( 30 ) 5. Provide intensive instruction on a daily basis that promotes reading proficiency

Roadblock 5.6. Some students who require toward reading proficiency should be the tier 3 instruction do not catch-up despite primary goal. Emphasize that the teaching intensive, one-on-one instruction. process should involve more than merely providing students with an opportunity to Suggested Approach. Remind school demonstrate the reading skills that they staff that a school’s goal is to help each already know. It must involve the integra- student reach proficient reading levels if at tion of new knowledge with previously all possible. Obtaining significant progress learned knowledge.

( 31 ) Appendix A. particular types of studies for drawing causal conclusions about what works. Postscript from Thus, one typically finds that a high level the Institute of of evidence is drawn from a body of ran- Education Sciences domized controlled trials, the moderate level from well designed studies that do not involve randomization, and the low What is a practice guide? level from the opinions of respected authorities (see table 1). Levels of evidence The health care professions have em- also can be constructed around the value braced a mechanism for assembling and of particular types of studies for other communicating evidence-based advice to goals, such as the reliability and validity practitioners about care for specific clini- of assessments. cal conditions. Variously called practice guidelines, treatment protocols, critical Practice guides also can be distinguished pathways, best practice guides, or simply from systematic reviews or meta-analy- practice guides, these documents are sys- ses such as What Works Clearinghouse tematically developed recommendations (WWC) intervention reviews or statistical about the course of care for frequently en- meta-analyses, which employ statistical countered problems, ranging from physi- methods to summarize the results of stud- cal conditions, such as foot ulcers, to psy- ies obtained from a rule-based search of chosocial conditions, such as adolescent the literature. Authors of practice guides development.78 seldom conduct the types of systematic literature searches that are the backbone Practice guides are similar to the prod- of a meta-analysis, although they take ad- ucts of typical expert consensus panels vantage of such work when it is already in reflecting the views of those serving published. Instead, authors use their ex- on the panel and the social decisions that pertise to identify the most important come into play as the positions of individ- research with respect to their recommen- ual panel members are forged into state- dations, augmented by a search of recent ments that all panel members are willing publications to ensure that the research to endorse. Practice guides, however, are citations are up-to-date. Furthermore, the generated under three constraints that do characterization of the quality and direc- not typically apply to consensus panels. tion of the evidence underlying a recom- The first is that a practice guide consists mendation in a practice guide relies less of a list of discrete recommendations that on a tight set of rules and statistical algo- are actionable. The second is that those rithms and more on the judgment of the recommendations taken together are in- authors than would be the case in a high- tended to be a coherent approach to a quality meta-analysis. Another distinction multifaceted problem. The third, which is is that a practice guide, because it aims for most important, is that each recommen- a comprehensive and coherent approach, dation is explicitly connected to the level operates with more numerous and more of evidence supporting it, with the level contextualized statements of what works represented by a grade (high, moderate, than does a typical meta-analysis. or low). Thus, practice guides sit somewhere be- The levels of evidence, or grades, are tween consensus reports and meta-anal- usually constructed around the value of yses in the degree to which systematic processes are used for locating relevant 78. Field and Lohr (1990). research and characterizing its meaning.

( 32 ) Appendix A. Postscript from the Institute of Education Sciences

Practice guides are more like consensus These are people the chair believes can panel reports than meta-analyses in the work well together and have the requisite breadth and complexity of the topic that expertise to be a convincing source of is addressed. Practice guides are different recommendations. IES recommends that from both consensus reports and meta- at least one of the panelists be a prac- analyses in providing advice at the level titioner with experience relevant to the of specific action steps along a pathway topic being addressed. The chair and the that represents a more-or-less coherent panelists are provided a general template and comprehensive approach to a multi- for a practice guide along the lines of the faceted problem. information provided in this postscript. They are also provided with examples of Practice guides in education at the practice guides. The practice guide panel Institute of Education Sciences works under a short deadline of six to nine months to produce a draft document. The The Institute of Education Sciences (IES) expert panel interacts with and receives publishes practice guides in education to feedback from staff at IES during the de- bring the best available evidence and ex- velopment of the practice guide, but they pertise to bear on the types of systemic understand that they are the authors and, challenges that cannot currently be ad- thus, responsible for the final product. dressed by single interventions or programs. Although IES has taken advantage One unique feature of IES-sponsored prac- of the history of practice guides in health tice guides is that they are subjected to care to provide models of how to proceed rigorous external peer review through the in education, education is different from same office that is responsible for inde- health care in ways that may require a pendent review of other IES publications. somewhat different design for practice Critical tasks of the peer reviewers of a guides in education have. Even within practice guide are to determine whether health care, where practice guides now the evidence cited in support of particular number in the thousands, there is no sin- recommendations is up-to-date and that gle template in use. Rather, one finds de- studies of similar or better quality that scriptions of general design features that point in a different direction have not been permit substantial variation in the realiza- ignored. Peer reviewers also are asked to tion of practice guides across subspecial- evaluate whether the evidence grade as- ties and panels of experts.79 Accordingly, signed to particular recommendations by the templates for IES practice guides may the practice guide authors is appropriate. vary across practice guides and change A practice guide is revised as necessary to over time and with experience. meet the concerns of external peer reviews and gain the approval of the standards and The steps involved in producing an IES– review staff at IES. The process of external sponsored practice guide are first to select peer review is carried out independent of a topic, which is informed by formal sur- the office and staff within IES that insti- veys of practitioners and requests. Next, a gated the practice guide. panel chair is recruited who has a national reputation and up-to-date expertise in the Because practice guides depend on the topic. Third, the chair, working in collabo- expertise of their authors and their group ration with IES, selects a small number of decision-making, the content of a practice panelists to coauthor the practice guide. guide is not and should not be viewed as a set of recommendations that in every case 79. American Psychological Association depends on and flows inevitably from sci- (2002). entific research. It is not only possible but

( 33 ) Appendix A. Postscript from the Institute of Education Sciences

also likely that two teams of recognized provide substantially better advice than an experts working independently to produce individual school district might obtain on a practice guide on the same topic would its own because the authors are national generate products that differ in important authorities who have to reach agreement respects. Thus, consumers of practice among themselves, justify their recom- guides need to understand that they are, mendations in terms of supporting evi- in effect, getting the advice of consultants. dence, and undergo rigorous independent These consultants should, on average, peer review of their product.

Institute of Education Sciences

( 34 ) Appendix B. Dick Olson to analyze data from the twin sample of the Colorado Learning Disabili- About the authors ties Research Center. Dr. Compton teaches undergraduate and graduate courses in Panel instructional principles and procedures in reading and writing for students with Russell Gersten, Ph.D., is President of disabilities. His research involves model- RG Research Group and Executive Direc- ing individual differences in the develop- tor of Instructional Research Group in ment of reading skills in children. He is Long Beach, California, as well as profes- currently the primary investigator on an sor emeritus in the College for Education Institute of Education Sciences (IES) project at the University of Oregon. Dr. Gersten addressing the key measurement issues is a nationally recognized expert on ef- associated with the Response-to-Interven- fective instructional practices to improve tion (RtI) approach to identifying learning reading comprehension (both for narrative difficulties. and expository text) and has extensive experience with the process of translating Carol McDonald Connor, Ph.D., is an as- research into classroom practice. He has sociate professor at Florida State Univer- led the teams responsible for developing sity and a research faculty member of the observational measures for reading com- Florida Center for Reading Research. She prehension and vocabulary instruction for completed her Ph.D. in Education and was several large-scale randomized control tri- an assistant research scientist in Psychol- als on the impact of observed practices in ogy at University of Michigan prior to com- reading instruction on growth in reading. ing to Florida State. Dr. Connor’s research He is an expert in instructional strategies interests focus on children’s learning in for improving reading comprehension, the classroom from preschool through adaptations of the reading research-base grade 3 and the complex relationships for English language learner students, and between children’s language and literacy longitudinal evaluation of reading pro- skills. She was recently awarded the 2006 grams. He has directed numerous imple- President’s Early Career Award for Sci- mentation studies, large-scale evaluation entists and Engineers and the 2007 APA projects, and randomized trial studies in Richard Snow Award. She is the principal the field of reading, with a focus on low- investigator of two studies funded by the income students and English learners. Institute of Education Sciences and the Na- Additionally, he chaired a panel of ex- tional Institute of Child Health and Human pert researchers for the National Center Development examining the causal effects for Learning Disabilities in June 2005 to of individualizing language arts instruc- synthesize knowledge of best practices in tion for students in grades 1–3 based on early screening and intervention for stu- their language and reading skills. dents with difficulties in mathematics. Joseph A. Dimino, Ph.D., is a research as- Donald L. Compton, Ph.D., is an associ- sociate at the Instructional Research Group ate professor of Special Education at Pea- in Long Beach, California where he is the body College, Vanderbilt University. Before coordinator of a national research project joining the faculty at Vanderbilt, Dr. Comp- investigating the impact of Teacher Study ton taught at the University of Arkansas- Groups as a means to enhance the qual- Fayetteville and spent a year as a post- ity of reading instruction for first graders doctoral research fellow at the Institute in high poverty schools and co-principal for Behavior Genetics at the University of investigator for a study assessing the im- Colorado-Boulder, where he worked with pact of collaborative strategic reading on

( 35 ) Appendix B. About the authors

the comprehension and vocabulary skills to RtI and school improvement, and read- of English language learner and English- ing program evaluation. She has published speaking fifth graders. Dr. Dimino has 36 extensively on the effects of research- years of experience as a general educa- based strategies on student reading. Her tion teacher, special education teacher, research has been recognized with awards administrator, behavior specialist, and from the Council for Exceptional Children researcher. He has extensive experience and the American Educational Research working with teachers, parents, admin- Association. istrators, and instructional assistants in the areas of instruction and early literacy, Sylvia Linan-Thompson, Ph.D., is an reading comprehension strategies, and associate professor at The University of classroom and behavior management in Texas in Austin. Her research interests urban, suburban, and rural communities. include development of reading inter- He has published in numerous scholarly ventions for struggling readers who are journals and coauthored books in reading monolingual English speakers, English comprehension and early reading inter- language learner and bilingual students vention. Dr. Dimino has delivered papers acquiring Spanish literacy. She is co-prin- at various state, national, and interna- cipal investigator of several longitudinal tional conferences, including the Ameri- studies funded by IES and the National can Educational Research Association, Institute of Child Health and Human De- the National Reading Conference, and the velopment examining the language and Council for Exceptional Children and As- literacy development in English and Span- sociation for Supervision and Curriculum ish for Spanish-speaking children and the Development. He consults nationally in the efficacy of a three-tier model of reading areas of early literacy and reading compre- intervention in general education class- hension instruction. rooms and in bilingual classrooms. She has authored curricular programs, book Lana Edwards Santoro, Ph.D., is a re- chapters, journal articles, and a book on search associate with the Instructional reading instruction. Research Group/RG Research Group and the Pacific Institutes for Research. She W. David Tilly III, Ph.D., is the Coordi- is a principal investigator on a series of nator of Assessment Services at Heartland IES–funded research on teaching reading Area Education Agency. He has worked as comprehension to grade 1 students dur- a practicing school psychologist, a univer- ing classroom read-alouds. Of particular sity trainer, a state department of educa- focus is her work to develop supplemental tion consultant and as an administrator interventions for students at risk of early in Iowa. He participated in the leader- reading difficulties, students with vocab- ship of Iowa’s transformation to using RtI ulary and language deficits, and English- practices and has extensive experience language learners. She also serves as prin- working with districts, intermediate agen- cipal investigator on an IES–funded study cies, states, and professional organiza- investigating the impact of enhanced core tions on the implementation of RtI. His reading instruction (tier 1) on the early research interests include implementing literacy achievement of Spanish-speaking system change, instructional interven- English language learners in transitional tions, formative assessment, and translat- bilingual programs. Dr. Santoro consults ing research into practice. He coauthored with state, local, and private agencies on a widely used publication on RtI for the a variety of projects, including training National Association of State Directors of presentations on effective instructional Special Education. strategies, program development related

( 36 ) Appendix B. About the authors

Staff Kristin Hallgren is a research analyst at Mathematica Policy Research and a former Rebecca A. Newman-Gonchar, Ph.D., is classroom educator. She has provided re- a research associate with the Instructional search support for several IES–sponsored Research Group/RG Research Group. She practice guides, including the dropout pre- has experience in project management, vention practice guide, this Response to study design and implementation, and Intervention reading and multi-tier inter- quantitative and qualitative analysis. Dr. ventions practice guide, and other forth- Newman-Gonchar has worked extensively coming practice guides. She has conducted on the development of observational mea- classroom observations and data analy- sures for beginning and expository read- sis for IES–sponsored projects related to ing instruction for two major IES–funded teacher quality and professional develop- studies of reading interventions for Title ment including rigorous evaluations of I students. She currently serves as a re- teacher preparation routes and high-inten- viewer for the What Works Clearinghouse sity teacher induction programs. She has for reading and mathematics interven- also been responsible for communicating tions and Response to Intervention. Her complex research design methodologies to scholarly contributions include concep- district and school-level administrators for tual, descriptive, and quantitative publi- rigorous evaluations of supplemental edu- cations on a range of topics. Her current cational services and mathematics curri- interests include Response to Interven- cula. Ms. Hallgren’s expertise in classroom tion, observation measure development practices and background in research for reading and mathematics instruction, methodology has provided the panel with and Teacher Study Groups. She has served support for translating research principles as the technical editor for several publica- into practitioner friendly text. tions and is a reviewer for Learning Dis- abilities Quarterly.

( 37 ) Appendix C. and state and local education agencies on teaching English language learner students, Disclosure of potential mathematics instruction for struggling stu- conflicts of interest dents, and various issues on bringing research findings into classroom practice. Practice guide panels are composed of individuals who are nationally recognized Joseph Dimino coauthored Interventions experts on the topics about which they are for Reading Success (2007). This is not a rendering recommendations. The Institute published curriculum but a series of sug- of Education Sciences (IES) expects that such gested activities for tier 2 interventions experts will be involved professionally in a for students in the primary grades. He re- variety of matters that relate to their work ceives royalties from Brookes Publishing. as a panel. Panel members are asked to dis- Dr. Dimino excused himself from all dis- close their professional involvements and cussions, reviews, and writing related to to institute deliberative processes that en- this program. This book is not referenced courage critical examination of the views of in this practice guide. panel members as they relate to the content of the practice guide. The potential influence Lana Santoro received royalties as a con- of panel members’ professional engage- sulting author for Scott Foresman Early Read- ments is further muted by the requirement ing Intervention (2003). Dr. Santoro excused that they ground their recommendations in herself from all discussions, reviews, and evidence that is documented in the practice writing related to this program or any other guide. In addition, the practice guide un- Scott Foresman products. The practice guide dergoes independent external peer review does not reference particular intervention or prior to publication, with particular focus on core curricula programs. It merely discusses whether the evidence related to the recom- topics that should be covered in tier 2 inter- mendations in the practice guide has been ventions and uses research (analyzed inde- appropriately presented. pendently by What Works Clearinghouse) as a basis for these decisions. No specific dis- The professional engagements reported cussion of the particular intervention pro- by each panel member that appear most gram took place in panel deliberations. closely associated with the panel recommendations are noted below. Sylvia Linan-Thompson was a co-principal investigator on the research project Russell Gersten has no financial stake in that tested the proactive reading curricu- any program or practice that is mentioned in lum that is published under the name SRA’s the practice guide. He is a royalty author for Early Interventions in Reading. Dr. Linan- what may become the Texas or national edi- Thompson was primarily responsible for tion of the forthcoming (2010/11) Houghton the development of the English language Mifflin reading series,Journeys . At the time of learner component of the curriculum. publication, Houghton Mifflin has not deter- While she is not an author on the program, mined whether or not they will ever release SRA is considering integrating the English this series. Dr. Gersten provided guidance on language learner component into some the product as it relates to struggling and Eng- of their curricula. The proactive reading lish language learner students. He excused intervention was discussed in the context himself from any sessions where Houghton of two research studies, one of which met Mifflin or any of its products were discussed. standards. However, the panel did not ad- The panel never discussed the Houghton dress specific issues in adjusting the cur- Mifflin series. Dr. Gersten also occasionally riculum for English learners, which is the consults with universities, research agencies, focus of Dr. Linan-Thompson’s work.

( 38 ) Appendix D. tify the risk pool of children needing tier 2 intervention. Studies predicting risk in Technical information kindergarten children have reported sensi- on the studies tivity rates approaching minimally acceptable level of 90 percent with specificity Recommendation 1. ranging from 56 percent to 86 percent,81 Screen all students for potential which means that often far too many stu- reading problems at the beginning dents are identified as at-risk for reading of the year and again in the middle difficulties.82 of the year. Regularly monitor the progress of students who are Results are more promising for grades 1 at elevated risk for developing and 2. Several studies have demonstrated reading disabilities. sensitivity in grade 1 above 90 percent with acceptable specificity.83 For example, Level of evidence: Moderate Compton et al. (2006) reports sensitivity rates approaching 100 percent with speci- The panel judged the level of evidence for ficity of 93 percent using a combination of recommendation 1 to be moderate. While a one-time screening battery (containing a growing number of screening studies measures of word identification, phone- are appearing in the research literature, a mic awareness, and rapid naming skill) in majority of studies relies on correlational combination with six weeks on progress designs, lack cross-validation, and fail to monitoring. However, these results have use representative samples. In this appen- not been cross-validated and were not ob- dix, we discuss the limited evidence base tained with a representative sample. Simi- in terms of sensitivity and specificity of lar results have been reported for screen- the measures. ing grade 2 students.84

Sensitivity is the degree to which a mea- Recommendation 2. sure correctly identifies children at risk Provide differentiated reading for experiencing difficulties in learning to instruction for all students based read. In contrast, specificity is the degree on assessments of students’ to which a measure correctly identifies current reading levels (tier 1). children at low risk for experiencing problems. These false positives refer to stu- Level of evidence: Low dents who eventually become good readers but score below the cut-score on the The panel rated the level of evidence for predictive instrument and are thus falsely this recommendation as low based on one identified as at risk. Providing these stu- descriptive-correlational study with first dents with extra tutoring stresses school and second graders that met standards resources, providing intervention to an in- with reservations and the opinion of the flated percentage of the population.80

To date, researchers have placed a pre- 81. Foorman et al. (1998); O’Connor and Jenkins mium on identification and early treatment (1999). of children at risk of future reading failure, 82. See Jenkins and O’Connor (2002) for a dis- cussion of the issue and for designing a manage- and therefore high sensitivity rather than able and acceptable risk pool for use within an specificity is favored. The overall effect of RtI framework. demanding high sensitivity is to over-iden- 83. Compton et al. (2006); O’Connor and Jenkins (1999). 80. Jenkins and O’Connor (2002). 84. Foorman et al. (1998).

( 39 ) Appendix D. Technical information on the studies

panel. The correlational study—Connor et reading domains taught in each tier 2 inter- al. (2008)—examines how student read- vention and any significant outcomes found ing growth varied by the degree to which for each of the five domains. Group size for teachers employed a specific differentia- tier 2 instruction, typical session length, tion program. This differentiation program and duration are also indicated. Note that relied on assessments to group students. many in the field consider frequency and Student reading growth was higher for duration as gauges of intensity of the inter- teachers who implemented the program vention.85 One study is excluded from the with greater fidelity. table but included in the accompanying text because it was a follow-up study of an Recommendation 3. intervention that produced strong effects Provide intensive, systematic in many reading domains.86 reading instruction on up to three foundational reading skills in small Because of the large number of high qual- groups to students who score ity randomized controlled trials and quasi- below the benchmark on universal experimental design studies conducted screening. Typically, these groups using systematic instruction in several of meet between three and five times the critical domains of beginning reading a week for 20 to 40 minutes (tier 2). instruction, the frequency of significant effects, and the fact that numerous research Level of evidence: Strong teams independently produced similar findings, the panel concluded that there is The panel judged the level of evidence sup- strong evidence to support the recommen- porting the recommendation to be strong. dation to provide intensive, explicit, and The panel found 11 studies conducted with systematic instruction in critical reading students in the primary grades that met skills stressed in National Reading Panel WWC standards or met standards with res- for tier 2 interventions.87 ervations. Table D1 provides an overview of each study’s outcomes in each of the 85. National Association of State Directors of five critical aspects of beginning reading Special Education (2005). instruction as articulated in the 11 stud- 86. Gunn et al. (2002). ies. The table provides an overview of the 87. National Reading Panel (2000).

( 40 ) Appendix D. Technical information on the studies

2–3 3–5 5–6 (Some students students Group size 2 students 3 students one-on-one one-on-one one-on-one one-on-one one-on-one one-on-one one-on-one) years) 8 weeks Duration 10 weeks 24 weeks 25 weeks 32 weeks 32 weeks 32 weeks 28 weeks 28 weeks 56 weeks (over two 7 months Intensity week week week Daily Daily Daily Daily Daily Daily Daily Frequency 30 min/day 10 min./day 35 min./day four timesfour a four timesfour a 50 min./day 30 min./day 30 min./day 30 min./day 30 min./day 40 min./day 30 min.30 four times a week three times a times three ≥25 min./day * ^ ^ ns ns ns ns Fluency * ns Vocabulary * * * * ^ ns Reading ^ (proactive intervention) intervention) * (responsive comprehension * * * * * * * Reading domain assessed Reading ns ns ns ns Decoding * * ns ns ns Phonemic awareness - PA F, V D, E C, V V, W > .10). PA, D p PA, D, E D, PA, PA, D, E D, PA, PA, D, C D, PA, sive: E, F, E,sive: F, C Respon PA, D, C, F D, PA, PA, D, E, F, E, D, F, PA, PA, D, E, D, W PA, PA, D, E, D, C,PA, = .05-.10). Both: PA, D, D, Both: PA, Intervention p < .05). p 1 1 1 1 1 1 1 1 1 K K–3 Grade level

Authors’ analysis based on studies in table. Studies in bold statistically showed significant effects at least in domainone of reading instruction. Ebaugh, Ebaugh, 2000 et al., 2005 et Vaughn al., 2006et Study Ehri et al.Ehri et 2007 Gibbs, 2001 et Jenkins al. 2004 Mathes McMaster et al. 2005 et al.et 2005 Lennon and Slesinski, 1999 Gunn et al. et Gunn 2000 Vadasy et al., et Vadasy 1997 Vadasy Table D1. Studies of tier 2 interventions in grades K–2 reading that met What Works Clearinghouse standards Clearinghouse Works What met that reading K–2 grades in 2 interventions tier of Studies D1. Table Note: = phonemic awareness,PA D = decoding, E = encoding (Spelling related phonics to instruction), C = Comprehension, F=Fluency, W = writing V=Vocabulary, ns statistically = not significant ( ^ = approached significance ( Source: * = statistically significant (

( 41 ) Appendix D. Technical information on the studies

Evidence supporting explicit, the core foundational skills for learning systematic instruction as the key how to read. Vaughn et al. found effects instructional delivery method for tier 2 in comprehension as well as decoding, tutoring on foundational reading skills. whereas Ebaugh’s effects were limited to decoding. Ehri included phonemic aware- All 11 studies used programs that system- ness, decoding, reading comprehension, atically taught reading skills,88 with seven and fluency. of these studies demonstrating a positive effect on one or more reading outcomes.89 In summary, this highly explicit, highly For example, Gunn et al. (2000) conducted systematic mode of small group instruc- a randomized controlled trial involving tion consistently produces positive ef- supplementary instruction for students in fects, often significant effects in the area kindergarten through grade 3 in phonemic of decoding and often in comprehension awareness, sound-letter correspondence, and vocabulary as well. What remains un- and decoding. Instruction was highly ex- certain is the balance of “learning to read” plicit, students received many opportuni- skills and comprehension, vocabulary, and ties to practice each skill, and feedback language development in tier 2 interven- was immediate and clear. Reading mate- tions. Most important, the field needs to rial consisted of decodable texts based on systematically study which domain areas current reading levels. Although the em- make the most sense for students at vari- phasis was on decoding and fluency, the ous levels of reading proficiency. Our hy- researchers also found an effect on read- pothesis is that the balance increases to ing vocabulary. more complex reading comprehension activities once students learn to read. How- Jenkins et al. (2004) and Vadasy et al. ever, for those still struggling to learn to (2005) used a virtually identical approach. read, it is unclear how much instruction in Content of the intervention was simi- vocabulary and listening comprehension lar except more time was spent on sight is necessary. words and spelling. Here effects were found not only in decoding but also in In understanding the nature of this body comprehension. The findings suggested of evidence, the reader should keep in that, at least in kindergarten and grade 1, mind that instruction was often one-on- students with strong systematic instruc- one (6 out of 11 of the WWC-rated studies) tion in small groups in phonemic aware- or in very small groups of two to three ness and decoding and fluent reading may students. also show growth in comprehension or vocabulary. In the remainder of the section, we review impacts on specific domains of tier 2 read- Both Ehri et al. (2007) and Vaughn et al. ing instruction. (2006) offered the widest menu of reading domains, including comprehension and vocabulary instruction along with Evidence supporting instruction of critical reading skills 88. Gunn et al. ��(2000);�� McMaster et al. (2005); Va- dasy et al. (1997); Vadasy, Sanders, and Peyton Phonemic awareness. Five studies mea- (2005); Jenkins et al. (2004); Gibbs (2001); Vaughn sured phonemic awareness—a student’s et al. (2006); Ebaugh (2000); Ehri et al. (2007); Ma- thes et al. (2005). understanding that words consist of individual phonemes. Phonemic awareness 89. Gunn et al. (2000); Jenkins et al. (2004); Ehri et al. (2007); Ebaugh (2000); Vadasy, Sanders, and is a potent predictor of future success in Peyton (2005); Vaughn et al. (2006). reading and a critical foundational skill for

( 42 ) Appendix D. Technical information on the studies

becoming a reader.90 Significant outcomes outcome measures in 7 of the 11 studies,94 were found for only two studies although and significant outcomes were reported in most of the tier 2 interventions did have a five studies.95 This also is a sizeable pro- phonemic awareness component.91 portion and indicates that one can expect effects in this domain. This is especially Three of the five studies showed no sig- interesting because of the five studies nificant effects for phonemic awareness. that demonstrated significant effects; only In some cases, ceiling effects may have three had a comprehension component. played a role in the lack of significant find- For example, Vadasy et al. (2005) and Jen- ings. Meanwhile, lack of significant effects kins et al. (2004) included a good deal of in the Gibbs (2001) study may be due to the oral reading of decodable texts96 but no short intensity and duration of the inter- explicit comprehension instruction. Yet ef- vention. In this investigation students re- fects on comprehension were significant. ceived 10 minutes of phonemic awareness The reader should keep in mind that al- instruction five times per week for only though this is an important finding, the eight weeks. In addition, it is common for level of comprehension tapped in most of students’ phonological skills to decrease these measures for grade 1 and 2 students as they begin to understand letter-sound is usually not very complex. correspondence. In other words, by the time students were post-tested their un- Vaughn et al’s (2006) intervention included derstanding of the relationship between a good deal of work with oral reading of letters and the sounds they make may connected text but also small group in- have influenced their performance on the struction in a variety of comprehension phonemic awareness assessments. strategies (using K-W-L, summarization, and retelling). This intervention led to sig- Decoding. Students’ ability to read real nificant effects. words and individual sentences (not connected text), was measured in all nine Vocabulary. Students’ vocabulary knowl- studies.92 Significant effects were reported edge was rarely assessed. Of the three in five of these studies.93 The fact that this studies that assessed this domain,97 signif- finding is replicated frequently indicates icance was reported in only one.98 Reading that the various approaches to systematic vocabulary is thus unlikely to improve un- explicit instruction all seem to produce less the intervention contains a vocabulary growth in this domain. component. But the small number of studies that assessed this phenomenon means Reading comprehension. Reading com- that results are simply inconclusive. prehension assessments were used as

94. ��Gunn et al. (2000); McMaster et al. ��(2005); 90. Vaughn et al. (2006); Gunn et al. (2000); Va- Vadasy, Sanders, and Peyton (2005); Jenkins et dasy, Sanders, and Peyton (2005); Ebaugh (2000); al. (2004); Vaughn et al. (2006); Ehri et al. (2007); Lennon and Slesinski (1999). Mathes et al. (2005). 91. Vadasy, Sanders, and Peyton (2005); Lennon 95. Vadasy, Sanders, and Peyton (2005); Jen- and Slesinksi (1999). kins et al. (2004); Vaughn et al. (2006); Ehri et al. (2007); Mathes et al. (2005). 92. Gunn et al. (2000); McMaster et al. (2005); Va-Va- dasy et al. (1997); Vadasy, Sanders, and Peyton 96. Jenkins et al. (2004) also contained a condi- (2005); Jenkins et al. (2004); Gibbs (2001); Len- tion where students read books that were not non and Slesinski (1999); Ebaugh, (2000); Ehri necessarily decodable. This condition, too, led et al. (2007). to significant effects in comprehension. 93. Ehri et al. (2007); Gunn et al. ��(2000); Jenkins 97. Gunn et al. ��(2000);�� Gunn et al. (2002); McMas- et al. (2004); Lennon and Slesinski (1999); Vadasy, ter et al. (2005). Sanders, and Peyton (2005). 98. Gunn et al. (2000).

( 43 ) Appendix D. Technical information on the studies

Fluency. Students’ ability to read connected of instruction,102 with one study reporting text fluently and accurately was assessed 50 minutes of instruction per session:103 the in 7 of the 11 studies,99 and treatment seven studies that had an effect on decod- students performed significantly better in ing, reading comprehension, or fluency pro- one study and approached significance (p vided instruction for at least 25 minutes,104 was between .5 and .10) in two studies.100 while the three studies that had no signifi- Students’ performance on these measures cant effects varied in the length of sessions resulted in a few intriguing findings. In the from 10 to 35 minutes.105 follow up study conducted a year after the supplemental tier 2 intervention, Gunn It is not possible to determine the role the et al. (2002) found that fluency outcomes number of days of intervention played in were significant, but the original study the studies in which no significant findings (Gunn et al. 2000) did not demonstrate sig- were found despite the intensity of the in- nificant fluency outcomes. In other words, tervention. Although one study provided it may take time before a fluency interven- intervention five times a week, it did so for tion demonstrates impact. only ten minutes a day,106 and one study provided instruction for 35 minutes but As primary grade students practice read- only three times a week.107 Based on the ing fluently, they seem to improve their evidence from these studies, it would be ad- word reading accuracy. When considered visable to provide intervention four to five together, results suggest that fluency inter- times a week and for at least 30 minutes. ventions are a promising practice, as opposed to a clear evidence-based practice for In 6 of the 11 studies students were in- tier 2 interventions at this point in time. structed on one-on-one.108 Configurations for the remaining studies109 consisted of small groups ranging from two to six stu- Research supporting intensity: dents. The panel suggests that the combi- frequency and duration of sessions nation of intensity (the amount of time per and group size session) and duration (number of weeks) rather than the grouping configuration Tier 2 instruction varied from three to may be the critical variable contributing to five times a week. Six of the studies with significant outcomes on decoding, reading comprehension, or fluency provided 102. ��Ebaugh (2000); Gibbs (2001); Gunn et al. daily instruction.101 But data suggesting (2000); Mathes et al. (2005); Jenkins et al. (2004); Lennon and Slesinski (1999); McMaster et al. that daily interventions lead to better ef- (2005); Vadasy et al. (1997); Vadasy, Sanders, and fects than those administered four days a Peyton (2005). week or even three is insufficient. 103. Vaughn et al. (2006). 104. Ebaugh (2000); Gunn et al. (2000); Gunn et In terms of length of intervention sessions, al. (2002); Jenkins et al. (2004); Lennon and Sle- nine studies provided at least 25 minutes sinski (1999); Vadasy, Sanders, and Peyton (2005); Vaughn et al. (2006). 105. McMaster et al. (2005); Vadasy et al. (1997); 99. Gunn et al. (2000); Mathes et al. (2005); Jen-Jen- Gibbs (2001). kins et al. (2004); Ehri et al. (2007); McMaster et al. (2005); Vadasy, Sanders, and Peyton (2005); 106. Gibbs (2001). Vaughn et al. (2006). 107. McMaster et al. (2005). 100. ��Gunn et al. (2002); Vadasy, Sanders, and 108. McMaster et al. (2005); Vadasy et al. (1997);�� Peyton (2005); Ehri et al. (2007). Vadasy, Sanders, and Peyton (2005); Jenkins et al. 101. ��Ebaugh (2000); Gibbs (2001); Gunn et al. (2004); Gibbs (2001); Erhi et al. (2007). (2000); Lennon and Slesinski (1999); Vaughn et 109. Lennon and Slesinski (1999); Ebaugh (2000); al. (2006); Mathes et al. (2005). Gunn et al. (2000); Vaughn et al. (2006).

( 44 ) Appendix D. Technical information on the studies

positive outcomes for students. However, bilingual students were selected using a this is only speculative at this point. priori criteria: schools were providing Eng- lish intervention for reading to at least two The only inference that can be clearly classes of grade 1 English language learner drawn is that the 10-minute phonemic students, at least 60 percent of the student awareness lessons conducted daily for population was Latino, and schools’ state- eight weeks were not intense enough to level reading achievement tests at grade 3 produce significant effects in reading- indicated that 80 percent or more of stu- related skills. The one-on-one sessions dents passed the test. tended to be reasonably lengthy (30 minutes) and of long duration. Three of the The research team screened all students four produced significant effects.110 in 14 bilingual, grade 1 classrooms in the four schools. Criteria for selecting In the four investigations where students students for the intervention were deter- were taught in small groups111 significant mined as being those who scored below outcomes were reported for interventions the 25th percentile in grade 1 on the Letter that ranged between 10 weeks and 1.5 years Word Identification subtest in both Span- and were conducted for 25 to 50 minutes ish and English, and who were unable to daily. Only Mathes et al. (2005) and Vaughn read more than one word from the simple et al. (2006) reported significant effects in word list. Two-hundred sixteen students reading comprehension. Significant out- were administered both the Spanish and comes in decoding and fluency were re- English screen at the four target schools. ported by Gunn (2000), while Lennon and One-hundred eleven students (51 percent) Slesinski (1999) reported significant effects met the Spanish intervention inclusion in phonemic awareness and decoding. De- criteria, 69 students (32 percent) met the coding was the only outcome measure in English intervention inclusion criteria, and the Ebaugh (2000). Unfortunately, after 58 students (27 percent) met both criteria. 30 minutes of instruction per day for 32 Eleven students met the English cutoff but weeks, there were no significant effects. not the Spanish cutoff, and these students were not eligible for the intervention.

A study of intensive, explicit, and The study was initiated with 24 interven- systematic small group instruction— tion students and 24 contrast students Vaughn, Mathes, Linan-Thompson, and due to ordinary attrition (students’ Cirino, Carlson, Pollard-Durodola, families moving or students transferring et al. 2006 to other schools), the study ended with 22 intervention and 19 contrast students (8 This intervention study was conducted in percent attrition for intervention and 21 two sites in Texas that were selected be- percent attrition for contrast); data were cause they were representative of the pop- not obtainable on one student (contrast) ulation areas where large numbers of bi- at either testing time point. The mean age lingual students go to school and because of the 47 students with pretest data was students were receiving reading instruc- 6.59 years (SD = 0.54). All students were tion in English. Four schools within these Hispanic, and female students comprised districts that were considered effective for 50 percent of the sample (n = 23).

110. Vadasy, Sanders, and Peyton (2005); Jenkins Eligible students received daily supple- et al. (2004); Ehri et al. (2007). mental instruction from October to April. 111. Lennon and Slesinski (1999); Ebaugh (2000); Each session was 50 minutes long. Forty Gunn et al. (2000); Vaughn et al. (2006). minutes were spent on literacy instruction.

( 45 ) Appendix D. Technical information on the studies

The literacy strands varied in time from to 3 standard deviations below normative 5 to 12 minutes. More time was dedicated levels for both groups, with performances to a strand when new elements were intro- nearing the average range only for English duced, and less time when it was review. Word Attack scores (for both groups). The read aloud was always 10 minutes. Intervention students’ performance on Eng- Daily lesson plans were comprised of six lish measures indicate that they outper- to ten short activities representing five formed control students on measures that content strands: phonemic awareness, ranged from rapid letter naming to reading letter knowledge, word recognition, con- comprehension as measured by WLPB-R pas- nected text fluency, and comprehension sage comprehension subtest. Intervention strategies. Daily lessons were fully speci- students’ were able to match sounds, blend fied and provided exact wording to ensure sounds to form words, segment words into teachers’ language was clear and kept to a phonemes, and delete sounds better than minimum. To ensure student engagement, control students. They also outperformed there was constant interaction between intervention students on the WLPB-R Word the instructor and students. The lesson Attack subtest, indicating that intervention cycle included modeling, group response, students demonstrated a greater ability to and individual turns. Pacing was another apply phonic and structural analysis skills characteristic of the intervention. A rapid to pronounce phonetically regular nonsense pace was maintained both in the exchange words in English. in each strand and in moving from activity to activity within each lesson. Tutors Recommendation 4. Monitor the also consistently monitored students’ re- progress of tier 2 students at least sponses, provided positive praise for cor- once a month. Use these data to rect responses, and scaffolded errors as determine whether students still they occurred. Finally, mastery checks require intervention. For those were conducted after every five lessons. students still making insufficient progress, school-wide teams should Given that students were assigned to in- design a tier 3 intervention plan. tervention and contrast groups randomly, there were no significant group mean dif- Level of evidence: Low ferences in performance on either of the skills (Woodcock Language Proficiency The panel rated the level of evidence as Battery-Revised (WLPB-R), letter word iden- low. Only three studies114 of tier 2 inter- tification, and experimental word read- ventions that met WWC standards or that ing list) used in the intervention screen, met standards with reservations included in English112 or Spanish.113 Furthermore, a weekly progress monitoring or unit mas- mean comparison of skill performance tery component. However, neither of the on the larger battery administered prior studies evaluated progress monitoring as to the onset of treatment indicated that an independent variable. Thus, no infer- students in the intervention and contrast ences can be drawn about its effectiveness groups performed at comparable levels based on the research reviewed. on all English and Spanish skills assessed, with no significant differences between In the Mathes et al. (2005) study, teachers students on any measure. Reading and lan- used data from student assessments to guage performances were approximately 1 identify needs and strengths, and planned

112. Woodcock (1991). 114. Mathes et al. (2005); McMaster et al. (2005); 113. Woodcock and Muñoz-Sandoval (1995). Gibbs (2001).

( 46 ) Appendix D. Technical information on the studies

instruction from that analysis. In the Gibbs need tier 2 instruction so that tier 2 groups (2001) study, tutors collected data weekly remain homogenous.121 An advantage of using mastery tests. After each mastery using progress monitoring measures for test, tutors were directed to proceed to the these decisions (as opposed to daily or next lesson or to repeat lessons based on weekly mastery tests) is that they provide number of correct responses.115 a more valid picture of overall growth in reading proficiency. A few studies of tier 2 interventions that met WWC standards or that met standards Recommendation 5. with reservations reported using data at Provide intensive instruction certain points during the intervention, on a daily basis that promotes but did not report on how often data was the development of the various collected or if students were regrouped components of reading proficiency based on progress. Two studies used data to students who show minimal to inform student pairings for practice progress after reasonable time in within tier 2 instruction.116 Pairs were re- tier 2 small group instruction (tier 3). arranged when one participant was making more progress than the other.117 Three Level of evidence: Low additional studies used data on student progress to determine the type of instruc- The level of evidence for this recommendation that students received, such as echo, tion is rated as low. Although the panel found partner, or independent reading.118 five studies that met the What Works Clear- inghouse standards (or met standards with Despite the lack of evidence supporting reservations) relating to this recommenda- use of progress monitoring, the panel de- tion, no studies reported statistically signifi- cided to recommend this practice for stu- cant impacts on reading outcomes.122 dents in tier 2 interventions. In our experience, progress monitoring data are used For the purposes of this document, tier 3 to determine students’ response to tier 2 is defined as a layer of instructional inter- instruction and to inform instructional de- vention for any student in kindergarten cisions in two ways.119 First, data are used through grade 2 who requires something to identify students who need additional substantially more intensive or substan- instruction to benefit from tier 2 instruc- tially different than tier 2 instruction. tion. This additional support, usually an However, tier 2 instruction is not the same additional 10 to 15 minutes, is provided to as special education for students with accelerate the learning of a student who is learning disabilities. Thus, we did not in- lagging behind the other students in the clude the literature base for special educa- group.120 It also identifies students who no tion for students with learning disabilities longer require tier 2 instruction. It can be in our review, though undoubtedly some used to regroup students who continue to ideas about promising practice might be gleaned from this body of research.

115. Gibbs (2001). Distinctions between tier 2 and tier 3 inter- 116. Lennon and Slesinski (1999); Gunn et al. (2000). ventions are far from clear. In our search 117. Lennon and Slesinski (1999). 118. Jenkins et al. (2004); Vadasy, Sanders, and 121. Vaughn, Linan-Thompson, and Hickman Peyton (2005); Ehri et al. (2007). (2003). 119. McMaster et al. (2005). 122. McMaster et al. (2005); Foorman et al. (1998); 120. Vaughn, Linan-Thompson, and Hickman Blumsack (1996); Gillon (2000); O’Connor and (2003). Jenkins (1995).

( 47 ) Appendix D. Technical information on the studies

of the literature, we found two studies123 Blumsack (1996) required that students on interventions that simultaneously tar- master segmenting three phoneme items geted tier 2 and tier 3 students.124 We, and letter-sound associations before mov- therefore, included these two studies since ing forward to the next activity level. they provided adequate information to draw inferences about the impact on tier All five studies included instruction that 3 students; three studies clearly addressed was designed to systematically move from a tier 3 population.125 easy to more difficult skills. Three of the five studies included specific material Although we found no evidence of signifi- supports and manipulatives to make stu- cant effects, we believe that several of the dent learning more concrete.129 In sum- studies suggest promising practices for mary, all involved systematic instruction tier 3 intervention. The reader should keep with extensive practice, clear feedback, in mind, though, that these are merely po- teaching to mastery, and carefully thought tentially promising practices. out progression from easy to hard learning activities—all elements of direct Although all five studies focused on a small instruction.130 number of high priority reading-related skills, only one included actual work on reading of words or pseudowords.126 A A study of carefully planned trend across interventions was the use of individualized instruction— multiple and extended instructional ses- O’Connor and Jenkins, 1995 sions ranging from a month127 to a full school year.128 O’Connor and Jenkins (1995) wanted to know whether teaching spelling to kinder- A key trait of all five studies was the use garteners experiencing extreme difficulty of extensive practice on the targeted read- learning to read would accelerate their reading-related skills. In all but one (Foorman ing growth. The intervention included ten et al. 1998), if a student made an error, the students who had been identified as devel- teaching guide or script provided explicit opmentally delayed and eligible for special procedures for teachers to correct student education services while in kindergarten. responses. Another key trait of all but Foor- The students had previously participated in man et al. (1998) was to use mastery crite- 60 hours of code-emphasis, decoding-based ria. Before a student could progress to the reading instruction as part of SRA’s Reading next skill level or new activity, they had Mastery I series,131 which explicitly teaches to perform a task correctly. For example, phonics and blending of phonemes.

123. Foorman et al. (1998); McMaster et al. Ten students were paired and then ran- (2005). domly assigned to experimental and con- 124. Both tier 2 and tier 3 studies were control conditions. Students in the experi- ducted with students in the primary grades mental group received 20 minutes of daily with reading difficulties or significant delays in individual spelling instruction during May reading (for example, students were considered of their kindergarten year in addition “nonresponders” due to reading performance and growth rates substantially below average to their daily small group code-empha- achieving peers). sis reading instruction provided in small 125. Blumsack (1996); Gillon (2000); O’Connor and Jenkins (1995). 129. Blumsack (1996); Gillon (2000); O’Connor and 126. Foorman et al. (1998). Jenkins (1996). 127. O’Connor and Jenkins (1995). 130. Engelmann and Carnine (1981). 128. McMaster et al. (2005); Foorman et al. (1998). 131. Engelmann and Bruner (1988).

( 48 ) Appendix D. Technical information on the studies

groups. In their spelling lessons, students the word) and then present earlier tasks pointed to and wrote letters that made as prompts to help guide the student’s re- a particular sound, started a particular sponse. Students in the control group re- word, or ended a particular word. Lessons ceived no spelling instruction at all. They at the end of the instructional sequence spent their time practicing reading words. required students to use magnetic letters to spell words from a selected word list, Results from O’Connor and Jenkins indi- as well as write two or three of the same cate that the intensive spelling instruc- words on paper. As students mastered tion component resulted in promising, words on a particular word list, new words although non-significant, effects in many were introduced. The teacher tracked the aspects of reading and spelling. A mea- exact words presented during a session, sure of decoding approached significance the student’s accuracy, and the number of with a p level of .09. Despite outcomes on times the student practiced the word be- spelling and word reading measures, there fore mastering the word’s spelling. were no differences between groups on a phonemic segmentation task. Spelling instruction included systematic routines. For example, a teacher would In addition to careful instructional plan- ask students to show (point to) a letter that ning that included individualized student makes a particular sound, then write the feedback and error correction, mastery letter that makes that particular sound. criteria, and lessons that moved system- Next, students would show a letter that atically from easier tasks for more difficult starts a word and then write the letter tasks, O’Connor and Jenkins’ results may that starts the word. These routines were suggest a promising practice for students repeated across lessons. Instruction was who require tier 3 intervention. Specifi- also scaffolded from easier tasks to more cally, the students who received spelling difficult tasks. Instruction began with an had a clearer and more direct presenta- individual letter and sound, then moved tion of how the alphabetic principle (words to first sounds, and then last sounds. Stu- include letters and letters are linked to dent feedback was also individualized. If a sounds) works in reading. Spelling may be student had difficulty with a word, teachers a more accessible way to apply phonologi- would first ask the child to orally segment cal skills to reading. Potentially, spelling the word (a scaffold or support strategy could help demonstrate how word read- to help the student identify the sounds in ing works.

( 49 ) References Catts, H. (1991). Early identification of dyslexia: Evidence from a follow-up study American Educational Research Associa- of speech-language impaired children. tion, American Psychological Associa- Annals of Dyslexia, 41(1), 163 –177. tion, and National Council on Measure- Chambless, J., & Chambless, M. (1994). ment in Education. (1999). Standards for The impact of instructional technology educational and psychological testing. on reading/writing skills of 2nd grade Washington, DC: AERA Publications. students. Reading Improvement, 31(3), American Psychological Association. 151–155. (2002). Criteria for practice guideline Compton, D. L., Fuchs, D., Fuchs, L. S., & development and evaluation. American Bryant, J. D. (2006). Selecting at-risk Psychologist, 57(12), 1048–1051. readers in first grade for early inter- Assessment Committee. (2002). Analysis of vention: a two-year longitudinal study reading assessment measures, coding of decision rules and procedures. Jour- form for Dynamic Indicators of Basic nal of Educational Psychology, 98(2), Early Literacy Skills. Retrieved from 394–409. the University of Oregon, DIBELS data Connor, C. M., Morrison, F. J., Fishman, system website: https://dibels.uoregon. B. J., Schatschneider, C., & Underwood, edu/techreports/dibels_5th_ed.pdf. P. (2007). The early years: Algorithm- Badian, N. A. (1994). Preschool prediction: guided individualized reading instruc- orthographic and phonological skills, tion. Science, 315(5811), 464–465. and reading. Annals of Dyslexia, 44(1), Connor, C. M., Piasta, S. B., Fishman, B., 3–25. Glasney, S., Schatschneider, C., Crowe, E., Baker, S., Gersten R., Haager, D., & Dingle, Underwood, P., & Morrison, F. J. (2009). M. (2006). Teaching practice and the Individualizing student instruction pre- reading growth of first-grade English cisely: Effects of child by instruction learners: Validation of an observation interactions on first graders’ literacy instrument. Elementary School Journal, development. Child Development, 80(1), 107(2), 199–219. 77–100. Baker, S. K., & Baker, D. L. (2008). English Cunningham, A. E., & Stanovich, K. E. learners and response to intervention: (1997). Early reading acquisition and Improving quality of instruction in gen- its relation to reading experience and eral and special education. In E. L. Grig- ability ten years later. Developmental orenko (Ed.), Educating individuals with Psychology, 33(6), 934–945. disabilities: IDEA 2004 and beyond. New Division for Learning Disabilities. (2007). York: Springer. Thinking about response to intervention Barker, A. B., & Torgesen, J. K. (1995). An and learning disabilities: A teacher’s evaluation of computer-assisted in- guide. Arlington, VA: Author. struction in phonological awareness Donovan, S., & Cross, C. T. (Eds.). (2002). with below average readers. Journal of Minority students in special and gifted Educational Computing Research, 13(1), education. Washington, DC: National 89–103. Academies Press. Blumsack, J. B. (1996). Teaching phonologi- Ebaugh, J. C. (2000). The effects of fluency cal awareness to children with language instruction on the literacy development impairments. (Doctoral dissertation, of at-risk first graders. (Doctoral dis- Syracuse University, 1996). Disserta- sertation, Fordham University, 2000). tion Abstracts International, 58(07A), Dissertation Abstracts International, 74 –2587. 61(06A), 0072.

( 50 ) R efERENCES

Ehri, L. C., Dreyer, L. G., Flugman, B., & of learning. American Educational Re- Gross, A. (2007). Reading rescue: An search Journal, 21(2), 449–450 effective tutoring intervention model Fuchs, L. S., Fuchs, D., & Compton, D. L. for language-minority students who (2004). Monitoring early reading devel- are struggling readers in first grade. opment in first grade: Word identifica- American Educational Research Journal, tion fluency versus nonsense word flu- 44(2), 414– 48. ency. Exceptional Children, 71(1), 7–21. Englemann, S., & Bruner, E. (1988). Read- Fuchs, L S., Fuchs, D., & Hamlett, C. L. (1989a). ing Mastery. Chicago: Science Research Effects of alternative goal structures Associates. within curriculum-based measurement. Engelmann, S., & Carnine, D. (1982) Theory Exceptional Children, 55(5), 429–438. of instruction: principles and practice. Fuchs, L. S., Fuchs, D., & Maxwell, L. (1988). New York: Irvington. The validity of informal reading com- Felton, R. H. (1992). Early identification of prehension measures. Remedial and children at risk of reading disabilities. Special Education, 9(2), 20–29. Topics in Early Childhood Special Educa- Fuchs, L. S., Fuchs, D., Hosp, M., & Jen- tion, 12(2), 212–229. kins, J. R. (2001a). Oral reading fluency Felton, R. H., & Pepper, P. P. (1995). Early as an indicator of reading competence: identification and intervention of pho- A theoretical, empirical, and historical nological deficits in kindergarten and analysis. Scientific Studies of Reading, early elementary children at risk for 5(3), 239–256. reading disability. School Psychology Fuchs, D., Fuchs, L. S., Thompson, A., Al Review, 24(3), 405– 414. Otaiba, S., Yen, L., Yang, N., Braun, M., & Field, M. J. & Lohr, K. N. (Eds). (1990). Clini- O’Connor, R. (2001b). Is reading impor- cal practice guidelines: Directions for a tant in reading-readiness programs? A new program. Washington, DC: National randomized field trial with teachers as Academy Press. program implementers. Journal of Edu- Foorman, B. R., Fletcher, J. M., Francis, D. cational Psychology, 93(2), 251–267. J., Schatschneider, C., & Mehta, P. (1998). Fuchs, D., Fuchs, L. S., & Vaughn, S. (Eds.). The role of instruction in learning to (2008). Response to intervention. Newark, read: Preventing reading failure in at- DE: International Reading Association. risk children. Journal of Educational Gersten, R., Dimino, J., & Jayanthi, M. (2008). Psychology, 90(1), 37–55. Reading comprehension and vocabulary Francis, D. J., Fletcher, J. M., Stuebing, K. K., instruction: Results of an observation Lyon, G. R., Shaywitz, B. A., & Shaywitz, study of first grade classrooms.Paper S. E. (2005). Psychometric approaches to presented at the annual meeting of the the identification of LD: IQ and achieve- Society for the Scientific Study of Read- ment scores are not sufficient. Journal ing, Asheville, NC, July 10–12, 2008. of Learning Disabilities, 38, 98–108. Gibbs, S. E. L. (2001). Effects of a one-to-one Francis, D. J., Shaywitz, S. E., Stuebing, phonological awareness intervention on K. K., Shaywitz, B. A., & Fletcher, J. M. first grade students identified as at risk (1996). Developmental lag versus deficit for the acquisition of beginning read- models of reading disability: A longitu- ing. (Doctoral dissertation, University dinal, individual growth curve analy- of South Carolina, 2001). Dissertation sis. Journal of Educational Psychology, Abstracts International, 62(07A), 0202. 88(1), 3 –17. Gillon, G. T. (2000). The efficacy of phono- Fuchs, L. S., Deno, S. L. & Mirkin, P. K. (1984). logical awareness intervention for chil- Effects of frequent curriculum-based dren with spoken language impairment. measurement on pedagogy, student Language, Speech and Hearing Services achievement, and student awareness in Schools, 31(2), 126–141.

( 51 ) R efERENCES

Good, R. H., & Kaminski, R. (2003). Dy- Jenkins, J. R., & O’Connor, R. E. (2002). Early namic indicators of basic early literacy identification and intervention for young skills. Longmont, CO: Sopris West Edu- children with reading/learning disabili- cational Services. ties. In R. Bradley, L. Danielson, and D. Good, R. H., Simmons, D. C., & Kame’enui, P. Hallahan (Eds.), Identification of learn- E. J. (2001). The importance of deci- ing disabilities: Research to practice (pp. sion-making utility of a continuum of 99–149). Mahwah, NJ: Erlbaum. fluency-based indicators of founda- Jenkins, J. R., Peyton, J. A., Sanders, E. A., tional reading skills for third grade & Vadasy, P. F. (2004). Effects of read- high-stakes outcomes. Scientific Studies ing decodable texts in supplemental of Reading, 5(3), 257–288. first-grade tutoring. Scientific Studies of Gunn, B., Biglan, A., Smolkowski, K., & Ary, Reading, 8(1), 53–85. D. (2000). The efficacy of supplemen- Johnson, E., Jenkins, J., Petscher, Y., & Catts, tal instruction in decoding skills for H. (in press). How can we improve the ac- Hispanic and non-Hispanic students in curacy of screening instruments? Learn- early elementary school. The Journal of ing Disabilities Research & Practice. Special Education, 34(2), 90–103. Juel, C. (1988). Learning to read and write: Gunn, B., Smolkowski, K., Biglan, A., & A longitudinal study of 54 children Black, C. (2002). Supplemental instruc- from first through fourth grades. Jour- tion in decoding skills for Hispanic and nal of Educational Psychology, 80(4), non-Hispanic students in early elemen- 437– 4 47. tary school: a follow-up. The Journal of Lennon, J. E., & Slesinski, C. (1999). Early Special Education, 36(2), 69–79. intervention in reading: Results of a Haager, D., Klingner, J., & Vaughn, S. (Eds.). screening and intervention program for (2007). Evidence-based reading practices kindergarten students. School Psychol- for response to intervention. Baltimore, ogy Review, 28(3), 353–364. MD: Paul Brooks Publishing Co. Mathes, P. G., Denton, C., Fletcher, J., An- Harn, B. A., Linan-Thompson, S., & Rob- thony, J., Francis, D., & Schatschneider, erts, G. (2008). Intensifying instruction. C. (2005). The effects of theoretically Journal of Learning Disabilities, 41(2), different instruction and student char- 115–125. acteristics on the skills of struggling Heller, K. A., Holtzman, W. H., & Messick, S. readers. Reading Research Quarterly, (Eds.). (1982). Placing children in special 40(2), 148–182. education: A strategy for equity. Wash- McCardle, P., Scarborough, H. S., & Catts, ington, DC: National Academy Press. H. W. (2001). Predicting, explaining, and Individuals with Disabilities Education preventing children’s reading difficul- Improvement Act, Pub. L. No. 108–446 ties. Learning Disabilities Research & (2004). Practice, 16(4), 230–239. Jenkins, J. R. (2003, December). Candidate McMaster, K. L., Fuchs, D., Fuchs, L. S., & measures for screening at-risk students. Compton, D. L. (2005). Responding to Paper presented at the Conference on nonresponders: An experimental field Response to Intervention as Learning trial of identification and intervention Disabilities Identification, sponsored by methods. Exceptional Children, 71(4), the National Research Center on Learn- 445–463. ing Disabilities, Kansas City, MO. National Association of State Directors Jenkins, J. R., Hudson, R. F., & Johnson, E. of Special Education. (2005). Response S. (2007). Screening for at-risk readers to intervention: Policy considerations in a response to intervention frame- and implementation. Alexandria, VA: work. School Psychology Review, 36(4), Author. 582–600.

( 52 ) R efERENCES

National Reading Panel. (2000). Teaching children. Washington, DC: National children to read: An evidence-based as- Academy Press. sessment of the scientific research litera- Speece, D., & Case, L. (2001). Classifica- ture on reading and its implications for tion in context: an alternative approach reading instruction (National Institute of to identifying early reading disability. Health Pub. No. 00-4769). Washington, Journal of Educational Psychology, 93(4), DC: National Institute of Child Health 735–749. and Human Development. Speece, D., Mills, C., Ritchey, K., & Hillman, Nunnally, J. (1978). Psychometric theory. E. (2003b). Initial evidence that letter New York, NY: McGraw–Hill. fluency tasks are valid indicators of O’Connor, R. E., & Jenkins, J. R. (1995). Im- early reading skill. Journal of Special proving the generalization of sound/ Education, 36(4), 223–233. symbol knowledge: Teaching spelling Swanson, H. L., Hoskyn, M., & Lee, C. to kindergarten children with disabili- (1999). Interventions for students with ties. Journal of Special Education, 29(3), learning disabilities: A meta-analysis 255–275. of treatment outcomes. New York, NY: O’Connor, R. E., & Jenkins, J. R. (1999). The Guilford Press. prediction of reading disabilities in Technical report: Texas primary read- kindergarten and first grade. Scientific ing inventory (1999 Edition). Retrieved Studies of Reading, 3(2), 159 –197. from: http://www.tpri.org/Documents/ Phillips, L. M., Norris, S. P., Osmond, W. C., 19981999TechnicalReport.pdf. & Maynard, A. M. (2002). Relative read- Torgesen, J. K. (2002). The prevention of ing achievement: A longitudinal study reading difficulties. Journal of School of 187 children from first through sixth Psychology, 40(1), 7–26. grades. Journal of Educational Psychol- Torgesen, J. K., & Burgess, S. R. (1998). ogy, 94(1), 3–13. Consistency of reading-related pho- President’s Commission on Excellence in nological processes throughout early Special Education. (2002). A new era: childhood: Evidence from longitudinal- Revitalizing special education for chil- correlational and instructional stud- dren and their families. Washington, ies. In J. Metsala, & L. Ehri (Eds.), Word DC: Author. recognition in beginning reading (pp. Scarborough, H. S. (1998a). Early identifica- 161–188). Hillsdale, NJ: Lawrence Erl- tion of children at risk for reading dis- baum Associates. abilities: Phonological awareness and Torgesen, J. K., Rashotte, C. A., & Alexander, some other promising predictors. In B. A. (2001). Principles of fluency instruc- K. Shapiro, P. J. Accardo, & A. J. Capute tion in reading: Relationships with es- (Eds.), Specific reading disability: A view tablished empirical outcomes. In M. Wolf of the spectrum (pp. 75–119). Timonium, (Ed.), Time, fluency, and developmental MD: York Press. dyslexia. Parkton, MD: York Press. Schatschneider, C. (2006). Reading difficul- Torgesen, J. K., Wagner, R. K., & Rashotte, ties: Classification and issues of predic- C. A. (1997). Prevention and remedia- tion. Paper presented at the Pacific Coast tion of severe reading disabilities: Keep- Regional Conference, San Diego, CA. ing the end in mind. Scientific Studies of Snow, C. E. (2001). Reading for understand- Reading, 1(3), 217–234. ing. Santa Monica, CA: RAND Education Vadasy, P. F., Jenkins, J. R., Antil, L. R., and the Science and Technology Policy Wayne, S. K., & O’Connor, R. E. (1997). Institute. The effectiveness of one-to-one tutoring Snow, C. S., Burns, S. M., & Griffin, P. (1998). by community tutors for at-risk begin- Preventing reading difficulties in young ning readers. Learning Disability Quar- terly, 20(2), 126–139.

( 53 ) R efERENCES

Vadasy, P. F., Sanders, E. A., & Peyton, J. A. Vellutino, F. R., Scanlon, D. M., Small, S. (2005). Relative effectiveness of read- G., Fanuele, D. P., & Sweeney, J. (2007). ing practice or word-level instruction Preventing early reading difficulties in supplemental tutoring: How text mat- through kindergarten and first grade ters. Journal of Learning Disabilities, intervention: A variant of the three-tier 38(4), 364–380. model. In D. Haager, S. Vaughn, & J. K. Vaughn, S., & Fuchs, L.S. (2006). A re- Klinger (Eds.), Validated practices for sponse to “Competing views: A dialogue three tiers of reading intervention (pp. on response to intervention.” Assess- 186). Baltimore, MD: Paul H. Brookes ment for Effective Intervention, 32(1), Publishing Co. 58–61. Wagner, R. K., Torgesen, J. K., & Rashotte, C. Vaughn, S., Linan-Thompson, S., & Hick- A. (1999). Comprehensive test of phono- man, P. (2003). Response to instruction logical processing. Austin, TX: PRO-ED. as a means of identifying students with Wanzek, J., & Vaughn, S. (2007). Research- reading/learning disabilities. Excep- based implications from extensive early tional Children, 69(4), 391–409. reading interventions. School Psychol- Vaughn, S., Mathes, P., Linan-Thompson, ogy Review, 36(4), 541–562. S., Cirino, P., Carlson, C., Pollard-Du- Woodcock, R. W. (1991). Woodcock lan- rodola, S., Cardenas-Hagan, E., & Fran- guage proficiency battery—revised. Chi- cis, D. (2006). Effectiveness of an Eng- cago, IL: Riverside. lish intervention for first-grade English Woodcock, R. W., & Muñoz-Sandoval, A. F. language learners at risk for reading (1995). Woodcock language proficiency problems. Elementary School Journal, battery—revised: Spanish form. Itasca, 107(2), 153–180. IL: Riverside.

( 54 )