Article Review Structure Quantitative Introductio

LSHSS

Language and Literacy Curriculum Supplement for Preschoolers Who

Are Academically At Risk: A Feasibility Study

Laura M. Justice The Ohio State University, Columbus

Anita S. McGinty Sonia Q. Cabell Carolyn R. Kilday

University of Virginia, Charlottesville

Kathy Knighton Ginger Huffman

West Virginia Department of Education, Charleston

A n unprecedented number of 3- to 5-year-old children currently attend publicly funded preschool and pre-kindergarten (pre-K) programs (National Institute for Early Education Research, 2004). Many of these children exhibit elevated risks for later reading problems due to poverty and its well- established impact on developmental precursors to reading achieve- ment. Consequently, increased attention is being directed toward ensuring that children who attend publicly funded preschool programs

ABSTRACT: Purpose: The potential benefit that a low-cost scripted language and literacy curriculum supplement titled “Read It Again!” (RIA; L. M. Justice, A. S. McGinty, A. R. Beckman, & C. R. Kilday, 2006) may have on preschool-age children’s skills was explored. RIA was developed to meet the needs of preschool educators who may not have access to current commercially available high-cost language and literacy curricula, which often require ongoing intensive professional development. RIA involves implementing 60 large-group lessons over a 30-week period that feature repeated use of 15 commercial storybooks. Method: Using a quasi-experimental pre–post research design, 11 preschool teachers implemented RIA in their classrooms for an academic year, and 9 teachers working in comparable preschool

have adequate opportunities to develop such critical reading pre- cursors as vocabulary knowledge, narrative ability, phonological awareness, and print knowledge (see Barnett, 2001; Dickinson & Brady, 2006; Justice & Kaderavek, 2004). Early Reading First pro- vides an apt example: This federally funded program was designed to create preschool “centers of excellence” in language and emer- gent literacy throughout the United States to serve as model instruc- tional programs (Institute of Education Sciences, 2007). The potential

programs served as comparisons. Language and literacy measures were collected in the fall and spring of the year. Results: Children whose teachers implemented RIA had higher scores in the spring on measures of language (i.e., grammar and vocabulary) and measures of literacy (i.e., rhyme, alliteration, and print). Effect- size estimates were consistent with medium- to large-size effects. Conclusions: RIA may be a viable means of enhancing the language and literacy instruction that is delivered within preschool classrooms and, therefore, a means of enhancing children’s language and literacy learning. Future directions for continued evaluation of RIA are discussed.

KEY WORDS: emergent literacy, preschool programs, at-risk children, curriculum

LANGUAGE, SPEECH, AND HEARING SERVICES IN SCHOOLS � Vol. 41 � 161–178 � April 2010 * American Speech-Language-Hearing Association 161

for such large-scale prevention-focused initiatives to reduce dispa- rities in reading achievement between children with economic dis- advantages and their more advantaged peers is the key catalyst behind such initiatives. Toward this goal, the current study investigated the pilot effects of a classroom-based, prevention-focused language and emergent literacy curriculum supplement on the language and emer- gent literacy skills of preschoolers who qualify for targeted publicly funded preschool programs based on family income and, in some cases, additional demographic risk factors (e.g., family instability).

Designing early childhood programs that are centers of excel- lence for promoting young children’s language and literacy develop- ment is a complex endeavor that likely involves efforts at many layers within the system of early childhood education. There is a wide range of options currently being considered for how to best build classroom-based support for children’s language and literacy development, ranging from preservice teacher training to ongoing professional development and support (see Justice & Vukelich, 2008; Zaslow & Martinez-Beck, 2006). Specifically, the use of scientifically validated language and literacy curricular tools is regarded as an important mechanism for ensuring that children receive high-quality language and emergent literacy instruction in the classroom (Christie, Vukelich, & Enz, 2007). In fact, the federal government recently made a significant investment in evaluating 14 promising preschool curricula involving language and literacy skills (as well as mathe- matical skills) through the Preschool Curriculum Evaluation Research Initiative (Preschool Curriculum Evaluation Research Consortium, 2008). Further, the federal government is actively compiling reviews of the evidentiary basis for specific early childhood language and literacy curricula.1 The goal of these efforts is to enhance the ability of early childhood practitioners to use evidence-based curricular tools to enhance their classroom instructional practices.

Findings regarding the benefits of preschool language and liter- acy curricula, however, are mixed and raise questions regarding the effectiveness of existing curricular tools for promoting high-quality classroom language and literacy instruction. For example, many targeted interventions conducted within preschool classrooms have had notably large effects on young children’s language and emergent literacy skills (e.g., Justice & Ezell, 2004; van Kleeck, Vander Woude, & Hammett, 2006; Whitehurst et al., 1994). Further, there are a number of efficacy and effectiveness studies providing empirical support for the positive impacts of language and literacy curricula within preschool classroom settings (e.g., Assel, Landry, Swank, & Gunnewig, 2007; DeBaryshe & Gorecki, 2007; Fischel et al., 2007). DeBaryshe and Gorecki (2007), for instance, evaluated child out- comes for the “Learning Connections” (LC) curriculum, a curric- ulum that was designed by the authors to improve teacher instruction in oral language, phonological and phonemic awareness, alphabet knowledge and print conventions, and emergent writing. Children who received LC showed greater end-of-year outcomes on measures of phonemic awareness and emergent writing relative to children who received the prevailing curriculum, although LC did not accel- erate children’s growth in vocabulary or emergent reading. Simi- larly, Assel et al. (2007) evaluated child outcomes for two commercial curricula, both of which provide teachers with an explicit scope, sequence, and instructional activities focused on alphabet knowl- edge, phonological awareness, and oral language. In general,

1For descriptions and reviews of many of the mentioned curricula, see the What Works Clearinghouse at http://ies.ed.gov/ncee/wwc/.

children who received either of the experimental curricula showed improved outcomes on measures of prereading, vocabulary, and listening comprehension relative to children in control classrooms in which no specified curricular scope and sequence was used. How- ever, it is important to note that effects of these curricula were also moderated by program type, where the strongest effects were seen in Head Start classrooms, as opposed to Title I and state-funded pre-K.

In contrast to these positive findings, however, the federal pre- school curriculum evaluation showed that only one of the 14 cur- ricula demonstrated significant effects on both children’s language and emergent literacy skills (effect sizes ranged from –.38 to .68 across all curricula). Interestingly, the federal initiative failed to find significant effects for the two curricula that were found to benefit children’s skill development in an independent evaluation by Assel et al. (2007). The variability in findings regarding the benefit of preschool language and literacy curricula suggests that curricula ef- fects, even when evident, may not be stable across samples or imple- mentation efforts. The divergent findings regarding language and literacy curriculum effects suggest that more needs to be learned regarding how to best design and implement curricula so they can effectively support young children’s skills development.

One possible drawback to many available language and literacy curricula is that they do not appear to be easily implemented well by teachers, which may reduce the benefit of such tools when they are considered at scale (Assel et al., 2007; Justice, Mashburn, Hamre, & Pianta, 2008; Pence, Justice, & Wiggins, 2008; Wasik, Bond, & Hindman, 2006). For example, teachers implementing structured language and literacy curricula may show immediate fidelity to structural aspects of the curricula (e.g., using specified materials, following lesson plans), but their implementation of more process- oriented elements of the curricula is typically much more variable (e.g., using recasts and expansions during conversations; see Justice et al., 2008; Pence et al., 2008). In fact, evidence suggests that many teachers require sustained, distributed support if they are to use many of the prevailing language and literacy curricula (Assel et al., 2007; Wasik et al., 2006; for a discussion, see Dickinson & Brady, 2006). As such, research-based implementation efforts often re- quire intensive and ongoing professional development provided by university-led experts and teams to ensure efficacious implementa- tion of language and literacy programs (e.g., DeBaryshe & Gorecki, 2007; Jackson et al., 2006; Wasik et al., 2006). Drawing from an earlier example, DeBaryshe and Gorecki (2007) reported providing teachers with two 1.5-hr in-service workshops across the year, weekly coaching by a research assistant, and meetings every third week with the lead researcher to assist teachers in planning and imple- mentation. Similarly, Assel et al. (2007) reported providing teachers with a 4-day institute focused on the curricula to which they were assigned and twice-monthly in-class mentorship over the course of the year to support implementation. Other studies report similarly intensive models of professional development, including specially de- signed credit-bearing college or university courses and distributed learn- ing models using distance technologies (Dickinson & Brady, 2006).

Although sustained professional development efforts can be effec- tive in supporting teacher implementation of language and emergent literacy curricula (Fukkink & Lont, 2007), these coaching models can be expensive, time consuming, and difficult to access, if not prohibitive, for many preschool programs. Thus, a gap exists be- tween the conditions under which many language and literacy curricula are shown to be efficacious or effective in research studies and the conditions under which preschool programs can readily

162 LANGUAGE, SPEECH, AND HEARING SERVICES IN SCHOOLS � Vol. 41 � 161–178 � April 2010

adopt language and literacy curricula in “business-as-usual” condi- tions. Indeed, outside of a research study context, some preschool programs do not have the resources—both money and personnel— to provide preschool educators with the necessary intensive supports they might need to implement evidence-based language and literacy techniques and curricula. Plagued with systemic challenges such as teacher turnover rates five times that of the K–12 school system (Barnett, 2003; Bellam, Burton, Whitebook, Broatch, & Young, 2002) and state-per-pupil spending falling below that of the K–12 system in 80% of states (Bryant et al., 2002), preschools need access to empirically validated tools and programs that can be easily used by large numbers of professionals at relatively low costs.

The present study examined the feasibility of implementing a language and literacy curriculum supplement, “Read It Again!”2

(RIA; Justice, McGinty, Beckman, & Kilday, 2006), that can be used to effectively enhance young children’s language and emergent liter- acy skills but requires minimal material costs or ongoing profes- sional development supports. RIA was developed in collaboration with the West Virginia Department of Education for the purpose of providing early childhood programs across the state with a scien- tifically based supplement that (a) could be used effectively by edu- cators with little training, (b) was compatible with a variety of program structures and curricula, and (c) could be implemented relatively inexpensively. As is occurring in many states across the nation, West Virginia is moving toward universal access to pre- school education; specifically, Policy 2525 (Universal Access to a Quality Early Education System, 2002) stipulates that all 4-year- olds in the state will have access to preschool by 2012–2013. The policy emphasizes not only access, but also quality and consistency of the education offered to children within a variety of program typologies (e.g., Head Start, public pre-K).

RIA was designed to provide early childhood educators with a curriculum supplement that may be layered into existing classroom instruction to explicitly address language and literacy goals within their programs. RIA was developed in a 3-year collaborative process that featured an iterative, interactive design process whereby researchers designing the tool actively sought practice-oriented feedback from end users, including teachers, speech-language pathologists, teaching assistants, and administrators. The 3-year collaborative process featured (a) a series of planning and design meetings to identify key characteristics of an effective supplement, to include an initial draft of the tool (Year 1); (b) a 15-week pilot implementation in- volving 18 early childhood educators and their school-based col- leagues in two counties to study feasibility of use by professionals working in a variety of settings (Year 2); and (c) a 30-week effec- tiveness study involving 20 teachers, 11 of whom implemented RIA for an entire academic year (Year 3). This “bidirectional model of reciprocal influence (i.e., research 6 practice)” provides a potentially important solution to closing the research-to-practice gap (Schaughency & Ervin, 2006, p. 159).

A particularly important feature of RIA’s development was its adherence to a priori design principles dually focused on using avail- able scientific knowledge of language and literacy instruction for young children and maximizing ease of implementation for teachers. The first design principle was systematicity, whereby RIA system- atically addresses a predefined scope and sequence of language and literacy learning targets that are critically linked to early and later

2The most current version of this supplement, “Read It Again—PreK!,” is available as a free download at www.myreaditagain.com.

reading achievement. Although research findings have supported the efficacy or effectiveness of a range of intervention techniques and targets to bring about developmental change, a commonality among approaches is explicit goal setting within a systematic scope and sequence of language and literacy instruction (e.g., DeBaryshe & Gorecki, 2007; O’Connor, Notari-Syverson, & Vadasy, 1996; van Kleeck, Gillam, & McFadden, 1998). The development team researched a range of materials to develop a scope and sequence, to include the meta-analyses of the National Early Literacy Panel (2004), Hammill (2004), and Scarborough (1998); a comprehen- sive synthesis of states’ early reading and writing standards (Bodrova, Leong, Paynter, & Semenov, 2000); prevailing preschool language and literacy curricula (e.g., Bunce, 1995; O’Connor, Notari-Syverson, & Vadasy, 2005); and research articles examining predictive relations between early achievements in language and literacy and children’s later reading outcomes (e.g., Connor, Morrison, & Slominski, 2006; Lomax & McGee, 1987; Storch & Whitehurst, 2002). The eventual scope encompassed two domains of language (vocabulary and narrative) and two domains of literacy (print knowledge and pho- nological awareness); a total of 24 objectives aligned to these four domains comprised the sequence of instruction, with approximately six specific objectives per domain, as presented in Appendix A.

The second design principle was applicability, whereby RIAwas designed for widespread use (a) in a variety of programs (e.g., Head Start, private preschools, public pre-K, home-based child care) and (b) by a wide range of professionals with diverse levels of back- ground knowledge and experience. In early childhood settings, program structure and workforce heterogeneity is substantial. Pro- grams may be full or part time and may be located in widely different spaces (e.g., homes, elementary schools; Brandon & Martinez-Beck, 2006). Layered on this setting-level variability are early childhood professionals whose qualifications can range from a high school diploma to a master’s degree in early childhood education (Adminis- tration for Children and Families, 2006; Bellam et al., 2002). In the design of the RIA curriculum, we needed to contend with the heterogeneity at the setting and user levels, as these can have marked implications for the effectiveness of instruction (see Schaughency & Ervin, 2006).

For the setting-level variability, we organized the curriculum to provide two brief lessons each week, requiring an estimated 40 min of instructional time per week; thus, we made RIA accessible to programs serving children only 2 days a week and those that operate on a half-day schedule. Given user-level (e.g., teacher-level) hete- rogeneity and our recognition that some users may have little if any formal knowledge of language and literacy instruction, RIA features a relatively scripted design, consisting of 60 lesson plans with ex- plicit objectives, materials, activity sequences, and even suggested language for instruction (see Appendix B). Curricula can range from codified frameworks that specify philosophies of learning to highly scripted manuals that specify session-by-session activities (Lonigan, Elbert, & Johnson, 1998). In scripted approaches, the pace, se- quence, and activities are governed by the curriculum itself rather than by the professional. The use of scripted versus individualized lessons is controversial (e.g., MacGillivray, Ardell, Curwen, & Palma, 2004), and some researchers have suggested that scripted instruction is less effective than nonscripted instruction (Moustafa & Land, 2002). Yet, studies have shown comprehensive scripted instruction to increase children’s academic achievement (Borman et al., 2005a, 2005b), and manualized lesson plans have been asso- ciated with efficacious preschool language and literacy instruction

Justice et al.: A Feasibility Study of a Preschool Curriculum Supplement 163

(e.g., Byrne & Fielding-Barnsley, 1993; Justice, Chow, Capellini, Flanigan, & Colton, 2003; O’Connor, Jenkins, Leicester, & Slocum, 1993; van Kleeck et al., 2006). Further, sequencing of instructional objectives is aligned with best practice, as it serves to makes the instructional path transparent to both children and teachers (Scott- Little, Kagan, & Frelow, 2003; Wixson & Dutro, 1999) and provides teachers with a level of specificity regarding what children should know and be able to do (Roskos, Rosemary, & Varner, 2006; Scott- Little et al., 2003).

The RIA program was also designed to be applicable to children in early childhood education classrooms regardless of their hetero- geneous skill levels and/or potentially wide range of risk factors (e.g., poverty, developmental difficulties). One particular area of child heterogeneity, namely, incoming language skills, may be an espe- cially important consideration in relation to RIA effects. Research has suggested that children’s developmental language character- istics may moderate the benefit that a particular language or literacy experience or intervention offers (e.g., Justice et al., 2003, Penno, Wilkinson, & Moore, 2002; Torgesen et al., 1999; van Kleeck et al., 1998; Vellutino et al., 1996). Thus, a particular intervention may show group mean benefits but may also be highly variable in its effect on individual children or groups of children based on their skill in language (as well as other relevant developmental areas; e.g., Bierman, Nix, Greenberg, Blair, & Domitrovich, 2008). RIA is de- signed to provide a balance of instructional approaches to address the potentially varied learning needs of young children with a range of language skills. For example, RIA instructional techniques such as teacher-directed instruction and explicit skills focus have dem- onstrated importance for increasing language and emergent literacy learning for children with low language or literacy abilities (e.g., Connor et al., 2006; Penno et al., 2002; Ukrainetz, Cooney, Dyer, Kysar, & Harris, 2000). Additionally, RIA allows for significant skills practice, repetition of material, and integration across skills, which are also considered important components of learning for children with low language abilities (Coyne, Kame’enui, & Carnine, 2007; Fazio, 1997). Yet, a central component of RIA is an interac- tive book reading instructional approach, where teachers actively involve children through questions. This type of interactive book reading session has been shown to benefit the language and literacy skills of a variety of children, including children whose language abilities are already in the average range (e.g., van Kleeck, Gillam, Hamilton, & McGrath, 1997; Whitehurst et al., 1994). Understanding how a curricular tool such as RIA may vary in its benefit as a function of children’s language skill is informative in considering the appli- cability of this program to meet children’s heterogeneous learning needs. This study directly addresses this point by considering not only whether RIA appears to benefit children’s language and liter- acy development, but also whether RIA is more or less beneficial to children based on their initial language skill.

The third design principle was feasibility, whereby RIA was de- signed for ready implementation with relatively few material re- sources, costs, or ongoing professional development required of its users. Although a number of well-designed and empirically vali- dated early language and literacy curricula are available for use by the early childhood community, their use may be prohibitive to educa- tors or administrators with limited resources. The material costs associated with one popular early language and literacy curriculum involve not only the cost of the curriculum kit, but also the cost of a boxed set of several hundred manipulatives and materials that need to be purchased separately. A center seeking to implement this

curriculum in six classrooms would need to spend approximately $20,000 to do so, which does not include professional development of its teachers. RIAwas designed as an alternative to those programs or curricula. Consequently, in the design of RIA, the development team worked within a set of constraints requiring the final product to be a self-contained manual that could be placed on the Internet free of charge. Additionally, the team’s intent was for required supple- mentary materials to cost no more than $100 to $200, to be easily accessible (i.e., not obscure items), and to be high yield (i.e., to be functional for other purposes). With these goals in mind, the devel- opment team opted for the use of storybooks as a relatively in- expensive, accessible, and high-yield material that should be heavily featured in the curriculum. A set of 20 storybooks (estimated cost $100) was selected for repeated use in the 60 RIA lessons; the story- books are used as a way to both organize instruction using a before reading, during reading, and after reading sequence, and embed code- and meaning-based instruction in a context that is familiar to both teachers and children. The sample lesson presented in Appen- dix B illustrates the lesson format.

In terms of professional development, RIAwas created to be min- imally dependent on ongoing professional development or coaching. In this study, participating teachers had had a semester of practice with some of the RIA lessons before this study and also received training in language and literacy (1 day) and training on the RIA program (4 hr) before the study. At the study’s inception, teachers received an additional half-day of refresher training on the RIA program, which included a discussion of the changes that had been made to the program. In total, professional development for this curriculum supplement involved approximately 16 hr of workshop training delivered at two points during teachers’ participation, as well as weekly use of the materials for 45 weeks (30 weeks during this study and 15 weeks of “practice” before implementation). It is important to note, however, that during the teachers’ use of the program, there was a minimal amount of discussion with the research team and almost no feedback on implementation. Thus, the RIA model is specifically designed for feasibility, where professional development is in the form of workshops (a format that is familiar and is widely used by many districts and early childhood programs) and where materials are designed to be easily and independently accessible to teachers, requiring minimal ongoing feedback or coach- ing regarding their use.

In the present article, we report findings from a study that was designed to explore the feasibility of implementing a language and literacy supplement that is beneficial to children and easily imple- mented by teachers. Campbell et al. (2000) illustrated a model of progressive, cyclical research as a means of determining how to design and implement effective interventions, especially “complex” interventions involving many pieces and personnel. In such a cy- clical approach, parameters of an intervention (e.g., intensity) may be explored and rigorously tested. Yet, additional exploratory and efficacy studies may be needed to determine which design features and model of implementation may be most beneficial to the most people. Although the field of early education has explored instruc- tional parameters of language and literacy curricula in relation to children’s skills (e.g., different teaching techniques, different instruc- tional targets), less attention has been given to designing interventions that are cost effective and accessible to the end user (e.g., teachers). This feasibility study directly explores this research-to-practice pa- rameter of early language and literacy curricular interventions by considering the potential benefit that RIA may provide to accelerating

164 LANGUAGE, SPEECH, AND HEARING SERVICES IN SCHOOLS � Vol. 41 � 161–178 � April 2010

children’s language and literacy development during their pre-K year when it is implemented with minimal coaching or ongoing profes- sional support.

The current study was exploratory in nature and, consistent with pilot- or feasibility-study designs, represents a precursor to more rigorously designed, randomized controlled trials that seek to estab- lish the causal effects of a given intervention. Feasibility studies are an especially important initial step when conducting classroom- based research (van Teijlingen & Hundley, 2001). That is, when conducting research on interventions that are implemented to groups of children clustered in classrooms, randomization should occur at the level of the classroom or teacher, resulting in a signif- icant investment of resources to recruit and retain the number of classrooms needed for a sufficiently powerful design. Feasibility research outcomes, such as that reported here, inform the planning of later studies as they can provide details on the number of clusters required in larger scale trials, including presumed effect sizes and intraclass correlations (Campbell, Mollison, & Grimshaw, 2001). As a feasibility study, the proposed study has several key design limi- tations, including use of a quasi-experimental research design, that preclude us from making definitive statements regarding causal impacts of the intervention. Nonetheless, our intent was to consider whether RIA appears to exhibit promise as a curricular tool, thus suggesting the need to further consider RIA as a potentially effective, low-cost, and accessible language and literacy curricular tool. The study involved two research questions:

& Do children who receive RIA for a 30-week period in their preschool classrooms exhibit better language and literacy skills in the spring of the year relative to children in comparison classrooms? The careful design of RIA focuses on targets and teaching techniques supported by empirical …