EARLY LEARNING FOR CHILDREN WITH SPECIAL NEEDS

EARLY LEARNING FOR CHILDREN WITH SPECIAL NEEDS This is course work. The only tool to be used is the uploaded article and there should be absolutely no outside sourcing, answers must be solely made from the article provided. Section 1. ? Summarize the results/findings of the article ? Critique/Evaluate the Results issues: 1. What are the major results/findings of the study? 2. How do these results answer the original research question(s)? 3. How confident are you with the findings? Section 2. ? Summarize the Discussion/Conclusion of the article ? Critique/Evaluate the Implications of the Findings: 1. What conclusions did the authors reach? 2. Are these conclusion(s) appropriate? 3. What did the researchers report as the implications of the study? 4. In your opinion, what is the significance of the findings for your area of interest? Vol.78. No. 4, pp. 471-490. ©20¡2 Councilor Exceptional Children. High-Quality School-Based Pre-K Can Boost Early Learning for Children With Special Needs DEBORAH A. PHILLIPS MARY E. MELOY Georgetown University ABSTRACTr:: Tbis article assesses tbe eficts of Tuba, Oklaboma's scbool-based prekindergarten program on tbe scbool readiness ofcbildren witb special needs using a regression discontinuity design. Participation in tbe pre-Kprogram was associated witb significant gains for cbildren witb special needs in early literacy scores, but not in matb scores. Tbese gains were not statistically different fiom tbose exbibited by tbeir classmates witbout special needs. Findings are interpreted as indicating tbat bigb-quality state pre-K programs can serve as effective early intervention programs for cbildren tuitb special needs. hildren with special needs have participated in public education alongside their typically developing peers for more than 30 years. Passage of the Education for All Handicapped Children Act of 1975 (EHA; now the Individuals With Disabilities Education Act, IDEA) assured children with disabilities the right to a free and appropriate education. The legislation placed a priority on serving children in the least restrictive environment, fueling a nationwide expansion of inclusive public school classrooms. More recently, the explicit inclusion of students with special needs in the accountability provisions of the No Child Left Behind Act of 2001 has reinforced the value placed on inclusive education and added to both the pressures and opportunities that confront school systems as they attempt to provide these children with appropriate and effective instruction (Lordeman oí Jones, 2010; Wakemaii, Browder, Meier, & McColl, 2007). The 1986 reauthorization of EHA provided strong financial incentives to states to provide public education for all eligible 3- to 5-year-old children who met criteria for developmental delay by 1991-1992, signaling an important shift towards preventive approaches to special education (Farran, 2000; Krauss & Hauser-Cram, 1992). The hope was that early intervention would result in lower special education costs over the school years. During the ensuing decade, the number of 3- to 5-year-olds receiving special education services grew by almost 50% (U.S. Department of Education, Office of Special Education and Rehabilitative Services, 2005), reaching 706,242 children (5.82% of the preschool population) in 2006 (Blackorby et al., 2010). One third of these preschoolers received . all of their special education in early childhood Exceptional Children 4 7 1 environments with peers without disabilities, including Head Start, child care, and pte-K settings (U.S. Department of Education, 2006). These trends have been accompanied by a very active debate regarding the most appropriate settings, activities, and focus and intensity of setvices for advancing the development of this young cohort (Administration for Children and Families, 1995; Bailey, McWilliam, Buysse, & Wesley, 1998; Curalnick, 2001; Kochanek & Buka, 1999; Wolery & Bailey, 2002). The growth in reliance on inclusive early education settings, in particular, has directed empirical attention to examining the role that high-quality early education programs can play in fostering school readiness for these children (Carlson et al., 2009; Hebbeler et al., 2007; Holahan & Costenbader, 2000). In fact, early childhood education has long been viewed as an important strategy for enhancing the later academic success of young childten, especially those who are vulnerable as a result of environmental or biological circumstances (Shonkoff & The growth in reliance on inclusive early education settings, in particular, has directed empirical attention to examining the role that high-quality early education programs can play in fostering school readiness. Meisels, 2000). Extensive reviews of the relevant literature for children with disabilities have consistently concluded that early education programs can positively influence school participation and outcomes for this population (Dunst, Snyder, & Mankinen, 1989; Farran, 2000; Guralnick, 1998), although strong research designs are rare and efforts to identify more and less effective approaches to eatly intervention have remained elusive. It is, nevertheless, well documented that low-income children who participate in highquality eatly education programs experience reductions in special education placements once they enter school (Campbell & Ramey, 1995; Conyets, Reynolds, & Ou, 2003; Redden, Ramey, Ramey, Fotness & Btezausek, 2003; Schweinhart, Barnes, & Weikatt, 1993). However, the landscape of eatly childhood education has undergone dramatic changes in the last decade as a result of widespread expansions in state pte-K education (Barnett et al., 2010). These programs have added a prominent option to the array of early intervention programs for young children with special needs, yet their potential to foster the eatly learning and development of children with special needs remains an unaddressed question. Batnett and colleagues (2010) estimated that in the 2009-2010 school yeat, 425,388 3- and 4-year-olds with special needs wete in state pre-K classrooms (both school-based and mixed delivery systems) with typically developing peets—comprising approximately 5.1% of the total preschool-age population. Of particular interest are the growing number of school-based pre-K classrooms, in light of prior evidence that effective "transition" or kindergarten entry strategies significantly improve the school success of children with special needs and that childten are more likely to receive transition services if they attend pre-K and kindergarten in the same school (Carlson et al., 2009; Schulting, Malone, & Dodge, 2005). Wolery and colleagues (Wolery, Holcombe, Brookfield et al., 1993; Wolery, Holcombe, Venn et al., 1993) have reported that close to three quarters of public school pre-K programs include children with disabilities. The pressing question is whether schoolbased pre-K offers these children a head start towards successful elementary school performance, as it does for other children (Gormley, Phillips, & Gayer, 2008; Gormley, Phillips, Newmark, Welti, & Adelstein, 2011), ot whether it constitutes an early start to learning disparities that distinguish many of these childten ftom theit typically developing classmates as they move through school (U.S. Department of Education, National Center for Education Statistics, 2007). This study was designed to address this question. Specifically, we utilized a quasi-expetimental design (regression discontinuity) to compare the kindergarten achievement test scores of children with individualized education programs (IEPs) who had attended the school-based pre-K in Tulsa, Oklahoma (as 4-year-olds) to the pre-K achievement test scores of children with IEPs who wete about to begin the ptogtam the following 4 7 2 Summer 2012 school year. We further compared the kindergarten test scores of the pre-K enrollees with special needs to those of their classmates without IEPs. The Oklahoma pre-K program has received national attention because, as one of a handful of programs with universal eligibility, it reaches a higher percentage of 4-year-olds (71%) than any other pro[;ram in the nation (Barnett et al., 2010). The pre-K program is not only universal in its eligibility; it is fully inclusive in its approach to special education. It also offers relatively highquality educational opportunities to the enrolled children compared to other pre-K programs around the country (Phillips, Gormley, & Lowenstein, 2009), and thus offers an important opportunity to examine whether high-quality, school-based pre-K can serve as an effective early intervention program in fostering school readiness among children with special needs. There are reasons to be both hopeful and concerned about the role of school-based pre-K programs in the education of children with special needs. On the one hand, pre-K programs share features with early intervention programs that appear to be effective in supporting the development of young children with special needs. For example, the developmental benefits of highquality early intervention services appear to derive, in part, from their child-focus and reliance on structured, carefully sequenced curricula (Graham & Bryant, 1993; Guralnick, 1998; Odom & Diamond, 1998; Shonkoff & Hauser-Cram, 1987), features that are central to school-based pre-K education. In their comprehensive analysis of services provided for young children with special needs in Montgomery County, Maryland, Markowitz, Hebbeler, Larson, Cooper, and Edmisten (1991) found that children who received services in classroom settings made larger gains on the Battelle Developmental Inventory (Newborg, 2004) than did children who received services in home-based or therapeutic settings—a result that has been replicated (Schwartz, Carta, & Grant, 1996). Moreover, although research findings on inclusion are mixed (see Guralnick, 1997; Odom & Diamond, 1998), the weight of the evidence seems to suggest that inclusion is beneficial for children with mild disabilities, and may support social development to an even greater extent than academic skills (Bailey et al., 1998; Holahan & Costenbader, 2000; Odom et al., 2004). Many school-based pre-K programs (including Tulsa's program) rely heavily, if not exclusively, on inclusive classrooms to serve young children with special needs (Smith, Kleiner, Parsad, & Farris, 2003). As a case in point, evidence on Head Start, for which federal guidelines require inclusion of children with special needs, has documented the program's effectiveness in promoting the development of these children, albeit modestly (Conyers et al., 2003; Gietzen & Vermeersch, 1980). Indeed, recent evidence from the National Head Start Impact Study revealed that children with special needs who enrolled in Head Start as 3- year-olds had made significant gains in math and social-emotional development at the end of first grade relative to their peers who did not attend Head Start (U.S. Department of Health and Human Services, Administration for Children and Families, 2010). Children without special needs who attended the program did not realize these gains. Whether these findings generalize to school-based pre-K programs, however, remains to be seen. On the other hand, it is widely recognized that inclusion does not guarantee quality services (Buysse, Wesley, Bryant, & Gardner, 1999; Division for Early Childhood, 2007), and that, although necessary, high quality alone does not appear to maximize the learning and development of children with special needs in early education settings (Guralnick, 2001; Odom et al., 2004). In addition, teachers in early childhood inclusive programs, including state pre-K programs, are typically not trained in special education, nor are special education professionals routinely available to these programs (Buysse, Wesley, Keyes, & Bailey, 1996; McDonnell, Brownell, &C Wolery, 1997; Whitebook et al., 2004; Wolery & Wilbers, 1994). Finally, close family involvement and support, along with deliberate orchestration of the array of supports and services needed by children with disabilities, are core tenets of service delivery for these children (Bruder, 2005; Dunst &¿ Trivette, 1997; Guralnick, 1997, 1998, 2005a, 2005b). Yet school-based teachers and administrators do not typically view these kinds of activities as central responsibilities. Exceptional Children 4 7 3 This study assessed the impact of Tulsa's high-quality, inclusive school-based pre-K program on the school readiness of children with special needs. We addressed two questions: 1. Does enrollment in Tulsa's pre-K program foster school readiness among children with special needs? We hypothesized that children with special needs who attended the pre-K program would perform at significantly higher levels on tests of preliteracy and premath skills at kindergarten entry when compared to children with special needs who had not yet experienced the pre-K program. 2. Does the Tulsa pre-K program have comparable impacts on children with and without special needs? Given the documented quality of the program, and its emphasis on structured instruction and inclusion—factors that have been linked to successful programs for children with special needs—we hypothesized that the effect of the program would be comparable for children with and without special needs. METHODS PARTICIPANTS Participants in this study were kindergarten and pre-K children enrolled in the Tulsa, Oklahoma, school-based pre-K program during the 2005-2006 and 2006-2007 school years, respectively. After obtaining institutional review board approval, those who enrolled during the 2005—2006 school year (the treatment group, referred to as alumni) were tested as they were entering kindergarten. Those who enrolled during 2006—2007 (the comparison group, referred to as entrants) were tested as they were entering pre-K. The combined sample of 3,048 participants included 312 children with special needs (129 entrants and 183 alumni) and 2,752 typically developing children (1,367 entrants and 1,385 alumni). We included typically developing children in the study sample to compare the effect of pre-K participation for children with and without special needs. The majority of both groups were in full-day programs (88.4% and 83.1%, respectively) and there were no differences between the two groups on full day status. All pre-K classrooms in Tulsa maintain a child: teacher ratio of 10:1 and a maximum classroom size of 20. Two teachers are assigned to each classroom. All lead teachers must have a bachelor's degree and an early childhood teaching certification. These teachers are paid the same wage scale as K-12 teachers in the Tulsa Public School system. Assistant teachers do not have any specific education or training requirements. Special needs status was determined using administrative data records that identified children who had an IEP in place. The administrative data records also provided the date of the initial IEP meeting, disability code(s), and the level of classroom inclusion the child experienced (full inclusion, inclusion with pull-out for special services, and noninclusive). Children were considered to have special needs for the purpose of this study if the date of their initial IEP meeting occurred prior to the end of their kindergarten year. Selection of the end of kindergarten as our cut-off was guided by feedback from the Tulsa Public School system indicating that children who are granted IEPs during kindergarten are not substantively different (in terms of the presence or severity of disability) from children who are given IEPs during the pre-K year (A. McKenzie, personal communication, September 8, 2008), perhaps because of the similar school-based setting and personnel of pre-K and kindergarten in Tulsa. Nevertheless, in light of evidence that type, severity, and complexity of special needs predict timing of identification (Palfrey, Singer, Walker, & Butler, 1987) and that disabilities such as speech or language impairments or autism tend to be identified earlier than others (Guarino, Buddin, Pham, & Cho, 2007), we examined whether the children identified as having special needs before or during pre-K differed from those identified in kindergarten. The results revealed that children who received their IEPs in kindergarten and those who received them earlier did not differ in their disability code(s) or achievement test scores, and sensitivity checks confirmed that our substantive results do not differ when children who received IEPs in kindergarten are removed from the sample. 474 Summer 2012 A total of 312 childten from the larger Tulsa Public Schools (TPS) pre-K sample had been designated special needs (had IEPs in place) by the end of kindergarten. Of those children, 250 children or 80% were in full-inclusion pre-K classrooms, 46 children or 14.5% were in full inclusion classrooms with periodic removal for special services (e.g. speech therapy sessions), and the remaining 16 children or 5.5% were in other, noninclusive, pre-K classroom settings. Only children in full inclusion and full inclusion with removal for special services classrooms—124 entrants and 172 alumni, for a total sample of 296 children—were included in the analyses. Children in other classroom settings were dropped to ensure that estimates reflected the effects of the "typical pre-K experience" on children with special needs and could be compared to the experience of typically developing children in the program. There were no major differences in the demographic characteristics or disability codes of the 16 children who were dropped from the sample and those who remained. Our final sample (A'^ = 296) consisted of 94 children (31.8%) whose special needs were identified prior to enteting the pre-K program, 147 children (42.9%) who were identified during pre- K, and 55 children (18.6%) who were identified during kindergarten. The majority of the children in this sample (289 or 97.5%) were primarily categorized as having a developmental delay. In Tulsa, this category is utilized as a catch-all for young children with mild to moderate needs who are not achieving at the level of their peers, but for whom future developmental status is uncertain (A. McKenzie, personal communication, September 8, 2008). Of the children who were categorized as having a developmental delay, 121 were assigned no other disability code. The majority of these children (168) were, however, assigned a secondary disability code as follows: 156 with speech impairments, 14 with learning disabilities, one with autism, one with other health impairments, and one with hearing impairments. The remaining seven children were not coded as experiencing developmental delays—but were categorized as having speech impairments only (six) or speech impairments and mental retardation (one). Thus, the sample of children with special needs in this study, although heterogeneous, consisted predominantly of those with mild and moderate delays who may or may not continue to receive services in elementary school and beyond. Of the 296 children with special needs included in the sample, 68.6% were male, 38.5% were white, 40.5% were black, 8.9% were Hispanic, and 11.8% were Native American. In terms of mother's education, 16.8% of mothers of children with special needs had not completed high school, 24.9% had graduated high school, 47.4% had completed an associate's degree or attended some college, and 11.0% had completed a bachelor's degree or higher. Finally, 64.5% of children with special needs qualified for free lunch, 10.5% qualified for reduced-price lunch, and 23% paid full-price lunch; 5.1% were English language learners; 56.1% lived with their biological father; and 50.5% had internet access in the home. The sample of children with special needs in this study, although heterogeneous, consisted predominantly of those with mild and moderate delays who may or may not continue to receive services in elementary school and beyond. The 2,752 typically developing children in this study had not been identified as developmentally delayed or otherwise in need of special services (IEP status) by the end of their kindergarten year. Of the typically developing children (both entrants and alumni), 49.6% were male, 32.8% were White, 34.2% were Black, 22.2% were Hispanic, 9.4% were Native American, and the remaining were Asian. In addition, 19.1% of these children's mothers had not completed high school, 27.4% had graduated high school, 39.7% had completed an associate's degree or attended some college, and 13.9% had completed a bachelor's degree or higher. Finally, 63.6% of typically developing children qualified for free lunch, 12.4% qualified for reduced-price lunch, and 24% paid full-price lunch; 15.5% were English language learners; 62.5% lived with their biological father; and 51.9% had Internet access in the home. Exceptional Children MEASURES AND PROCEDURE The data used in this study are from student tests and parent surveys administered in August 2006 in Tulsa, Oklahoma, as well as administrative data records from the TPS system accessed in June 2008. Children were tested on their pre-academic skills using the Woodcock-Johnson Tests of Achievement III (WJ III; Woodcock, McCrew, & Mather, 2001), a nationally normed, widely used assessment tool that has been used extensively with racially and socioeconomically mixed samples, and with children with special needs (Chase- Lansdale et al., 2003; Henry, Cordon, St Rickman, 2006; Puma et al., 2005). Three subtests of the WJ III were selected to reflect age-appropriate preacademic skills: Letter-Word Identification, Spelling, and Applied Problems. The Letter-Word Identification subtest measures prereading and reading skills. It requires children to identify letters that appear in large type and to pronounce words correctly (the child is not required to know the meaning of any particular word). The Spelling subtest measures prewriting and spelling skills. It measures skills such as drawing lines and tracing letters and requires the child to produce uppercase and lowercase letters and to speir simple words correctly. The Applied Problems subtest measures early math reasoning and problem-solving abilities. It requires the child to analyze and solve math problems, performing relatively simple calculations. These subtests are appropriate for relatively young children, including preschoolers (Mather & Woodcock, 2001), and have been used in other studies with this age group (Chase-Lansdale et al., 2003; Henry et al., 2006; Puma et al., 2005). Barbara Wendling, a nationally recognized expert on the WJ III and a highly experienced trainer, trained teachers to administer the three tests at one of two training sessions held in Tulsa in late August 2006. Teachers administered the WJ III subtests during the first week of school (designated as a testing week for TPS, prior to the start of classes). Teachers administered all tests in English unless the child being tested was designated a bilingual student, in which case the child was also given a Spanish version of the test, the Batería III Woodcock-Mufioz (Woodcock, Muñoz-Sandoval, McGrew, & Mather, 2005). Only the English-language test scores are analyzed in this article. We collected data on individual child and parent characteristics via surveys that were completed by the parents while their child was being tested. Parents were asked the child's race and gender, the mother's highest level of education, if the father lived at home with the child, and whether the family had Internet access in the home. We measured family socioeconomic status via an income proxy. Schools reported the lunch status of all children, which provided three levels (i.e., free lunch, reduced-price lunch, full-price lunch). Standard cut-offs for lunch status are determined by the U.S. Department of Agriculture's National School Lunch Program and correspond to 130% of the federal poverty level (FPL) for free lunch, 185% of the FPL for reduced-price lunch, and above 185% of the FPL for full-price lunch. DATA ANALYSIS Selection bias is the key difficulty in assessing the effects of any voluntary program, regardless of the poptilation it serves. Children whose parents enroll them in a voluntary pre-K program, for example, may differ in important ways from children whose parents do not enroll them. To the extent that these differences are measurable, their influence can be controlled. However, if some of these differences (e.g., children's intelligence or motivation and parental attitudes) are not measured, then any direct group comparison will be biased, as will the estimated effects of participation in the program. In effect, differences between enrolled and not-enrolled children that are ascribed to the pre-K program may actually be partially attributable to these types of preexisting sample differences. Regression-Discontinuity Design. To reduce selection bias, we utilized a regression discontinuity (RD) design to estimate the direct impact of pre- K participation on children with special needs and typically developing children, respectively, and to determine whether a differential effect of participation existed for the two subsamples. This approach builds on previous work with the TPS data (Gormley et al., 2008; Cormley & Cayer, 2005; Cormley, Cayer, Phillips, & Dawson 2005), others' work evaluating pre-K for typically 4 7 6 Summer 2012 F I G U R E 1 Hypothetical Illustration of Regression Discontinuity Design Test Score Cut-off Age Control (Entrants) Counterfactual (Entrants) Treatment (Alumni) developing childten (Barnett, Lamy, & Jung, 2005), and evaluations of language/reading interventions for developmentally at-risk children (Tuckwiller, Pullen, & Coyne, 2010; Vaughn, Wanzek, Murray, Scammacca, & Linan, 2009). RD substantially reduced selection bias by creating a treatment group that consisted of children who attended the pre-K program in the 2005—2006 school year and a comparison group that consisted of children who were about to begin the pre-K ptogtam at the beginning of the 2006-2007 school yeat. TPS enfotces a strict cutofïdate fot pre-K eligibility (Septembet 1) that is identical to the cut-off date fot kindergarten eligibility. This means that students who were born on or before September 1, 2001, wete eligible to participate in the pte-K ptogram for the 2005-2006 school yeat but students who wete born after that date wete not eligible to participate until the following yeat. This strict birthday qualification creates a situation where assignment into the pre-K treatment group is based solely on the cut-off vatiable, in this case age, which is untelated to the selection process. The associated inability of teseatchers or patents to manipulate a given child's assignment into the treatment ot compatison group increases confidence that treatment estimates are unbiased (Imbens & Lemieux, 2008; Lee ôiLemieux, 2010). Figure 1 provides a hypothetical illusttation of this design. The dotted line to the right of the cut-off date shows hypothetical test scores of the treatment group and the bold solid line to the left of the cut-off date shows hypothetical test scores of the comparison gtoup. The key challenge in estimating the effect of TPS pte-K is to estimate the countetfactual or test-scote outcomes fot treated children had they not been treated. The solid line to the right of the cut-off date depicts the countetfactual. The regression discontinuity design assumes that the countetfactual is continuous at the cut-off date, so any jump in estimated test scores for the treated children relative to the countetfactual can be Exceptional Children 477 TABLE 1 Demographic Characteristics of Pre-K Students With Special Needs: Alumni Versus Rntrants Variable Special needs ID before pre-K Special needs ID during pre-K Special needs ID during kindergarten Female Black White Hispanic Native American Lunch status Paid Reduced Free Maternal education No high school diploma High school diploma/CED Some college/associate's degree Bachelor's degree or higher Resident father status Internet access in the home M 0.290 0.516 0.194 0.339 0.395 0.355 0.097 0.145 0.008 0.298 0.081 0.621 0.137 0.274 0.508 0.500 0.476 Comparison SE 0.041 0.045 0.035 0.043 0.044 0.043 0.027 0.032 0.008 0.041 0.025 0.044 0.031 0.039 0.045 0.045 0.045 N 124 124 124 124 124 124 124 124 124 124 124 124 124 124 124 124 124 M 0.337 0.483 0.180 0.297 0.413 0.407 0.081 0.099 0.000 0.215 0.122 0.663 0.273 0.261 0.349 0.535 0.436 Treatment SE 0.036 0.038 0.029 0.035 0.038 0.038 0.021 0.023 0.000 0.031 0.025 0.036 0.044 0.031 0.035 0.038 0.038 N 172 172 172 172 172 172 172 172 172 172 172 172 172 172 172 172 172 Difference -.047 .033 .013 .042 -.018 -.052 .015 .046 .008 .083 -.041 -.042 -.168"* .013 .159*** -.035 .040 */> < 0.10. *> < 0.05. ***/> < 0.01. attributed to their participation in the pre-K program. The underlying assumption that supports using this analytic design to assess the impact of the pre-K program is that a child who just made the pre-K entrance cut-off date and a child who just missed the cut-off date should have similar characteristics (both measurable and immeasurable), except that one child has already received the treatment (Tulsa pre-K) and the other has not. We tested this assumption for measurable characteristics using the entire sample of children with special needs (reflected in a 12-month winclow in age around the cut-off date). As Table 1 indicates, there were no statistically significant differences between the children who had experienced pre-K and were entering kindergarten (the treatment group) and the children who were just entering pre-K (the comparison group) for gender, ethnicity, resident father status, Internet access, or free lunch status. There were, however, differences between the two groups for mother's education level. Specifically, the comparison group had more highly educated mothers. Inadequate sample size prevented tests to confirm whether imbalances diminish and the treatment effect remains for the children with special needs as the window (age in months) around the cut-off variable narrows. However, narrowing the window around the cut-ofFdate for the larger sample of typically developing children in this study and the entire sample of children (with and without special needs) who attended the pre-K program (Gormley & Gayer, 2005; Gormley et al., 2005; Gormley, 2011) confirmed that for these populations, imbalances diminished and test score differences remained. We are confident that the imbalances in the sample of children with special needs do not reflect a manipulation of children's assignment into the treatment or comparison group. Rather, they reflect a policy-induced loss of children with special needs whose mothers were poorly educated from the comparison group. The implications of 4 7 8 Summer 2012 these imbalances for the interpretation of the magnitude of pre-K program impacts for children with special needs are discussed later in this article. However, these imbalances do not affect the validity of utilizing RD in this study because they are not correlated with the cut-off date. Analytic Approach. The primary goal of this study was to estimate the effect of pre-K program participation for children with special needs. To do this, we ran a series of Ordinary Least Squares (OLS) regressions on the subsample of children with special needs, including both entrants to (A^ = 124) and alumni of {N = 172) the pre-K program. The initial model utilized a dichotomous variable (treatment) which captured whether the student was born on or before September 1, 2001, and represents treatment status equal to 1 for students who participated in the pre-K program in 2005-2006 and equal to zero for students who participated in the Tulsa pre-K program in 2006-2007, child's age (qualify) measured as the difference in days between the student's date of birth and September 1, 2001, and an interaction term for child's age by treatment. Covariates added to this model included race, gender, free lunch eligibility, maternal education, whether the child lived with his or her biological father, and whether the child had Internet access at home. Although, theoretically, the covariates should be uncorrelated with the treatment variable and are therefore unnecessary inclusions in the model if the underlying assumptions of the regression discontinuity design hold, we included them here to increase the precision of the estimated treatment effect. After estimating the effect of pre-K participation on the preacademic skills of children with special needs, we compared the impact of participation on children with special needs to that of their typically developing peers. To accomplish this, we ran a similar series of OLS regressions on the combined sample of special needs and typically developing children (A'^= 3,048). This series of regressions utilized a fully interacted model, which included students' special needs status; an interaction variable capturing the relative effect of treatment for students with special needs; and interaction terms for all of the covariates with special needs status. We implemented multiple imputation as a strategy to address the prevalence of missing data, particularly with regards to demographic variables that relied on parental survey report (Rubin, 1996), using the ice program in Stata (version 10.0) to create five imputed data sets that were then combined to produce estimates of the missing data values (Royston, 2004, 2005a, 2005b). In this study, missing values for gender, ethnicity, maternal education level (passively imputed), resident father status, and Internet access, were imputed. Missing values for WJ III test scores (outcome variables) and free lunch status were not imputed. Children with missing values fot either the outcome variable or for free lunch eligibility were dropped from the analyses. In each of the models, approximately half of the children who were dropped were missing WJ III test scores; the majority of these children failed to show up to school on the day they were being tested. Tests of the sensitivity of the estimates to different methods of dealing with missing data indicated that findings were robust. RESULTS IMPACT OF PRE-K FOR CHILDREN WITH SPECIAL NEEDS Table 2 presents tests of the hypothesis that children with identified special needs benefited from their enrollment in the pre-K program. Of the 296 entrants and alumni with special needs, 252 were included in the analysis for the WJ III Letter- Word Identification subtest, 243 were included in the analysis for the Spelling subtest, and 250 were included in the analysis for the Applied Problems subtest due to missing values (for test scores or free lunch eligibility). We calculated effect sizes by dividing the estimated treatment effect by the standard deviation of the comparison group. Children with special needs who had participated in the pre-K program had significantly higher Letter-Word Identification raw test scores {b = 3.433, SE = 1.109, p < 0.001) and Spelling raw test scores (b = 3.049, SE = 0.728, p < 0.001) than children who had selected into the program but had not yet experienced it, after controlling for age and other demographic variables. These Exceptional Children TABLE 2 Effect of Pre-K on Raw Test Scores: Children Witb Special Needs Variable Treatment Age in days Age in days X treatment Female Black Hispanic Native American Asian Reduced-price lunch Free lunch Maternal education No high school diploma Some college/ associate's degree Bachelor's degree or higher Lives with father Internet access in home Constant Number of observations Letter-Word Identification b 3.433*** 0.004 0.002 0.329 -1.468** 0.272 0.062 8.599*** -0.446 -0.505 -0.528 1.263* 3.338* 0.357 1.445** 3.333*** SE 1.109 0.003 0.005 0.632 0.607 1.112 0.765 0.999 1.033 0.844 0.916 0.737 1.859 0.709 0.667 1.233 252 Spelling b 3.049*** 0.004 -0.002 1.105*** -0.781 0.105 -0.390 1.160** -0.187*** -0.142 -0.446 0.734 1.016 0.728* 0.232 3.375*** 243 SE 0.728 0.002 0.004 0.413 0.505 0.678 0.602 0.586 0.781 0.577 0.580 0.857 0.857 0.412 0.514 0.839 Applied Problems b 1.295 0.006 0.009* 1.006 -2.451*** -2.373* -1.283 4.333*** -0.328 -0.704 1.594 2.067** 2.132 0.261 1.417* 6.428*** SE 1.347 0.004 0.006 0.739 0.828 1.366 1.106 1.223 1.195 0.967 1.118 0.828 1.508 0.861 0.837 1.725 250 */> < 0.10. **/> < 0.05. ***/> < 0.01. results indicate test score impacts of 3.433 points (a 1.093 effect size) for the Letter-Word scores and test score impacts of 3.049 points (a 1.155 effect size) for the Spelling scores. There were no differences between the two groups on the Applied Problems raw test scores (h = 1.295, SE = 1.547, p > 0.10). Figures 2 and 3 provide graphical depictions (scatterplots) of the RD findings. In both figures, the jump in test scores at the cutoff date represents the impact of the pre-K program on test scores, which we have estimated here. COMPARING PRE-K IMPACT EOR CHILDREN WITH AND WITHOUT SPECIAL NEEDS Table 3 presents a direct comparison of the effect of pre-K program participation on typically developing children and children with special needs for all three WJ III subtests (see Table 4 for fUll regression results). Due to missing values fot the outcome variables or free lunch eligibility, 2,746 of the complete sample of entrants and alumni (A'^ = 3,048) were included in the analysis for the Letter- Word Identification subtest, 2,641 were included in the analysis for the Spelling subtest, and 2,724 were included in the analysis for the Applied Problems subtest. There were no significant differences in the effect of program participation on the test scores of children with and without special needs for any of the subtests. Figure 4 illustrates the test-score gains associated with pre-K participation for children with special needs and typically developing children, converted into age equivalence scores in months. It is important to note that for WJ III the relationship of test-score points to months gained is non-linear. Children with special needs, who 48O Summer 2012 F I G U R E 2 Scatterplot of Regression Discontinuity for Letter-Word Identification Subtest t Score 15 tion Subtes 10 ttér-word Identifica 0 5 4L.-* • • • -• -365 f ir*—;"' -180 Age in • • 1 ( days from 05 • m 1 180 -06 pre-k cutoff date Test Score Fitted Line for Entrants Fitted Line for Alumni • • 365 Note. Each data point represents the average test score for groups of children born within a 10-day window. FIGU R E 3 Scatterplot of Regression Discontinuity for Spelling Subtest 8 CO o . Q) • S 3 CO O) .E ID - 'SQ . CO o - -365 • • • ( • A • * * • • 1 A A # • • • H• i • • •• • * -180 0 180 Age in days from 05-06 pre-k cutoff birthday • Test Score Fitted Line for Entrants Fitted Line for Alumni 365 Note. Each data point represents the average test score for groups of children horn within a 10-day window. Exceptional Children 4 8 1 TABLE 3 Comparison of the Effect of Pre-K on Raw Test Scores of Typically Developing Children and Children Witb Special Needs Variable Letter-Word Identification Spelling Applied Problems Typically Developing b 3.736*** 2.121*** 2.000*** SE 0.350 0.224 0.335 Special Needs b 3.433*** 3.049*** 1.295 SE 2.209 0.728 1.347 Difference (Treatment by Special Needs Status) b SE -0.303 1.137 0.928 0.758 -0.705 1.365 *p < 0.10. **/> < 0.05. ***p < 0.01. scored lower than childten without special needs on all three subtests ptior to pre-K participation, demonsttated latget age equivalence gains despite comparable gains in terms of raw test score points for the two gtoups. D I S C U S S I O N This study teptesents the fitst effort to look at the effects of school-based pte-K education on the school readiness of childten with special needs. It addressed the question of whether school-based pte-K offers young children with special needs, operationalized by IEPs, a head start towatds successful elementary school performance or whether it fails to advance the eatly leatning of these childten. The tesults are cause for optimism. Schoolbased pre-K education, as experienced in Tulsa, supported the eatly leatning of children with special needs, as hypothesized. Indeed, the impact of F I G U R E 4 Tulsa Pre-K Program Impacts in Monthly Equivalents for Typically Developing Children and Cbildren With Special Needs 70 • Equîva < il" 1 £ 30 • <co c o 20 • Je g 1 ^"- 0 • 8.27 } 1 55.19 7.48 f 51.14 1 Letter-word Identification 4.26 52.87 Spelling 9.25 51.84 Applied Problems Typically Developing Children 10.72 44.71 • Letter-word Identification 1 3.14 45.35 DGain a Baseline Spelling Applied 1 Problenns Special Needs Children 4 8 2 Summer 2012 TABLE 4 Regression Results Variable Treated Special education status Special education status X treated Qualify age in days Qualify X cut-off Special ed status X qualify age in days Special ed status X qualify X cut-off Reduced-price lunch Eree lunch Black Hispanic Native American Asian Eemale Lives with father Internet access in home Mother's education (ME) No high school Some college College grad and up Reduced price lunch X special ed status Eree lunch X special ed status Black X special ed status Hispanic X special ed status Native American X special ed status Asian X special ed status Female Lives with father X special ed status Internet access in home X special ed status ME: No high school X special ed status ME: Some college X special ed status ME: College grad and up X special ed status Constant Observations Letter-Word Identification b 3.735*** -2.490** -0.303 0.007*** 0.001*** -0.004 0.003 -1.224***' -1.164*** -0.043 -1.436*** -0.252 -0.134 0.838*** 0.157 1.064*** -0.326 0.664*** 1.804*** -0.778 -1.135*** -1.424** 1.708 0.315 8.734*** -0.510 0.200 0.381 0.203 0.599 1.534 5.823*** SE 0.351 1.254 1.137 0.001 0.002 0.003 0.005 0.337 0.251 0.218 0.243 0.305 0.585 0.164 0.184 0.207 0.252 0.208 0.354 1.065 0.888 0.632 1.110 0.807 1.147 0.636 0.733 0.653 0.953 0.744 1.840 0.380 2746 Spelling h 2.121*** -2.255*** 0.928 0.008*** -0.001 -0.005** -0.000 -0.376*** -0.562** -0.367** -0.178 0.090 0.552 1.108*** 0.227* 0.544*** -0.390* 0.297** 1.086*** 0.190 0.420 -0.415* 0.283 -0.479 0.608 -0.003 0.501 -0.312 -0.055 0.437 -0.070 5.630*** SE 0.224 0.864 0.758 0.001 0.001 . 0.002 0.004 0.336 0.163 0.147 0.159 0.187 0.467 0.106 0.133 0.149 0.200 0.146 0.213 0.786 0.163 0.512 0.681 0.617 0.738 0.416 0.413 0.556 0.625 0.513 0.861 0.276 2641 Applied Problems b 2.000*** -5.512*** -0.705 0.012*** 0.003* -0.007 0.014** -1.094*** -1.189*** -2.179*** -3.395*** -0.387 -2.996*** 0.536*** 0.049 0.938*** -0.801*** 0.667*** 2.104*** 0.767 0.485 0.272 1.022 -0.896 7.329*** 0.470 0.213 0.479 2.395** 1.400* 0.029 11.940*** : SE 0.335 1.718 1.365 0.001 0.002 0.004 0.006 0.295 0.239 0.222 0.263 0.288 0.637 0.166 0.222 0.222 0.266 0.230 0.317 1.201 0.970 0.222 1.356 1.113 1.357 0.738 0.897 0.814 0.266 0.841 1.498 0.418 Î724 *p < 0.10. **/> < 0.05. ***p < 0.01. Exceptional Children pte-K on the achievement test scores of the children with IEPs was not statistically different from the impact of pre-K on the scores of their typically developing classmates. The program appears to have its largest effects on the Letter-Word Identification subtest, which assesses prereading abilities, and the Spelling subtest, which assesses prewriting skills, for children with special needs. The Applied Problems subtest of early math reasoning and problem-solving abilities did not exhibit significant differences between the children with special needs who had and had not yet attended the pre- K program. This pattern of results also characterized the complete sample of children—those with and without special needs—who attended the Tulsa pre-K program during the 2005-2006 school year, although the Applied Problems subtest did produce significant group differences in this larger sample (Gormley et al., 2008). The relatively greater impact of the program on literacy than on numeracy skills may reflect the differential amount of classroom time that the Tulsa pre-K teachers devoted to these two domains of school readiness. To gain a better understanding of ptactices inside the school-based pre-K programs, we deployed highly trained observers to monitor the morning sessions of all classrooms (Phillips et al., 2009). On average, the pre-K teachers allocated 33% of their morning classroom time to prelitetacy and writing activities, whereas they allocated only 17% of their time to math activities. In addition, emerging evidence is revealing that early math skills, perhaps more so than early literacy skills, are affected by young children's executive function capacities, including their working memory and attention (Espy et al., 2004; Li-Grining, Raver, & Smith- Donald, 2010). These capacities pose special challenges to children with disabilities (Gathetcole, AUoway, Willis, & Adams, 2006; Liebman &C Goodman, 1995), which may tender typical premath instruction less effective than preliteracy instruction in fostering kindergarten readiness for children with special needs. It may also be the case that young children with special needs require more individualized math instruction than their peers without special needs, and than was provided in Tulsa pre-K classrooms. Not only did pre-K attendance bolster the school readiness of the children with special needs, but its impacts were comparable to those for the typically developing children. The effect sizes for the children with special needs were 1.093 and 1.155 for the Letter-Word Identification and Spelling tests, respectively. These effect sizes exceed those reported for other state-funded pre-K programs, which range from .17 to .68 (Wong, Cook, Barnett, & Jung, 2008), with data reported for complete samples that likely include some children with special needs but consist predominantly of typically developing children. Moreover, although the children with special needs scored lower on each of the WJ III subtests than the children without special needs both prior to and after attending pre-K, as expected, their test gains expressed in monthly equivalents were quite large for both the Lettet-Wotd Identification and Spelling subtests. Indeed, as compated to the children with special needs who had not yet experienced pre-K, those with pre-K experience gained over 9 months on the Letter-Word test and close to 11 months on the Spelling test. Although gains on the Applied Problems subtest were not significant for the children with special needs, it is notable that their scores on this test of pre-math skills placed them over 3 months ahead of children with special needs who had not yet attended the program. Nevertheless, in light of the importance of early mathematics concepts for children's continued academic success (Duncan et al., 2007), effotts to improve mathematics instruction within the context of inclusive programs that serve children with special needs are critical. There is reason to believe that these results may actually represent a conservative estimate of the impact of pre-K on the school readiness of children with special needs. The comparison group of children with special needs, who were just entering the pre-K program, as compared to the treatment children, had more highly educated mothers. These differences in maternal education appear to be the direct result of an administrative initiative within TPS to funnel all children with special needs, regardless of socioeconomic status, into the local Head Start program and away from the public pre-K program (A. McKenzie, personal communication, September 8, 2008). Highly educated mothers may have resisted the placement 4 8 4 Summer 2012 of their children with special needs into Head Start, thus creating the observed higher maternal education level, on average, in the TPS pre-K comparison group than the treatment group and resulting in a conservative estimate of the treatment efFect. Indeed, analyses designed to test this hypothesis (available from authors upon request) confirmed that impacts presented here are conservative. Although the current study did not examine the specific features of the Tulsa pre-K program that might account for its positive impacts on the early learning of children with special needs, it is noteworthy that, in prior analyses, the TPS pre-K classrooms scored higher on classroom quality on a range of indicators as compated to a multistate sample of school-based pre-K programs (Phillips et al., 2009). Quality of early educational environments clearly matters for children with special needs, as it does for all children (Conyers et al., 2003). More specifically, these indicators included observational assessments of time management, reliance on instructional techniques that maximize students' engagement and fostet higher order thinking skills, and providing feedback that expands understanding. These classroom features have been associated with children's early learning in previous studies (Howes et al., 2008; Mashburn et al., 2008). Tulsa pre-K teachers also devoted more classroom time, relative to pre-K teachers in other states, to literacy and math instruction. In light of evidence within the special education literature that children with special needs benefit from well-organized, structured, and sequenced instruction, as well as from intensive exposure to learning materials (Guralnick, 1998; Hill, Brooks-Gunn, & Waldfogel, 2003), these features of Tulsa pre-K are strong candidates for future efforts to identify the predictors of the gains demonstrated by the childten with special needs in these classrooms. Tulsa's almost exclusive reliance on inclusion, combined with the relatively high quality of instruction provided, may also contribute to the promising results (Bailey et al, 1998; Holahan & Costenbader, 2000; Odom et al., in press), although, we were not able to compare inclusive and noninclusive classrooms in this study. A relatively new literature examining peer effects in pre-K classrooms may also be pertinent in the context of Tulsa's reliance on inclusive classrooms. Henry and .Rickman (2007) found that the ability levels of children's peers within preschool classrooms were directly related to the variation in children's cognitive, prereading, and expressive language skills at kindergarten entry. Although peer effects have not been examined fot children with special needs, it is plausible that they may operate in a similar fashion within integrated classrooms. Full-day preschool attendance generates higher rates of developmental progress for children with special needs than does half-day attendance (Holahan & Costenbader, 2000); the fact that close to 90% of the children with special needs in the Tulsa pre-K program attended for a full day may also play a role in the gains made by these children. Finally, in addition to the evidence of high-quality instruction described above, it is notable that every pre-K teacher in the Tulsa pre- K program had a bachelor's degree, was earlychildhood certified (although not necessarily certified in special education), and was paid on the wage scale for public school teachers, suggesting a relatively well-educated early childhood teacher workforce that focused classroom time on instruction. Although these factors may account for the strong progress towards school readiness made by the children with special needs who attended the Tulsa pre-K program, they also limit the generalizability of the findings. The findings are also specific to pre-K that is both universal and inclusive. Other limitations concern an inability to examine variation in outcomes associated with specific subgroups within the population of children with special needs defined either by disability code or severity of disability, or by variation in family characteristics or environmental risk, due to both lack of information and the small sample size. Indeed, our findings should be viewed as generalizable only to young children with relatively mild to moderate disabilities. The enormous heterogeneity in both biological and environmental causes of disability, and its contribution to variability in response to early intervention, is an extremely pressing issue that this work cannot inform (Guralnick, 2005a). Unfortunately, we lacked information on other services received by the children in this Exceptional Children 4 8 S study, which may have contributed to their progress beyond the benefits of the pre-K program, as well as on the family circumstances of the children in our study. The developmental systems approach (Guralnick, 2005b, 2011) highlights the important contribution of the family in fostering the development of children with special needs. However, absent family-level data, we are unable to ascertain how family dynamics in either the comparison or treatment group tnay have contributed to our results. Yet the current results document that the TPS pre-K program is efficacious in preparing young children with a range of mild to moderate special needs for the demands of elementary school. In light of the rapid growth in special education enrollments within state pre-K programs, many of which rely at least in part on schoolbased classrooms, these are very promising results. Concerns that the needs of these children will be neglected in the context of pre-K programs that are not designed specifically with children with special needs in mind, or that state pre-K may even contribute to disparities in learning that emerge during the school years for these children, would appear to be largely unfounded when the quality of early education is high and when pre-K education is tightly integrated into the elementary education system. The challenge for the future is one of ensuring that this is not a unique pattern of results. This entails deciphering the "active ingredients" of the Tulsa program for children with special needs and taking the necessary steps to translate this information into programs across the nation that meet the needs of young children with early delays and disabilities. R E F E R E N C E S Administration for Children and Families. (1995). Passages to inclusion: Creating systems of care for all children. Monograph for state, territorial and tribal child care administrators. Washington, DC: Administration for Children, Youth, and Families. (ERIC Document Reproduction Service No. ED406819) Bailey, D. B., McWilliam, R. A., Buysse, V., &c Wesley, P. W. (1998). Inclusion in the context of competing values in early childhood education. Early Childhood Research Quarterly, 13, 27-47. http://dx.doi.org/10 .1016/S0885-2006(99)80024-6 Barnett, W. S., Epstein, D. J., Cardan, M. E., Fitzgerald, J., Ackerman, D. J., & Friedman, A. H. (2010). The state of preschool 2010. New Brunswick, NJ: Rutgers University, National Institute for Early Education Research. Barnett, W. S., Lamy, C, & Jung, K. (2005). The effects of state prekindergarten programs on young children's school readiness in five states. New Brunswick, NJ: Rutgers University, National Institute for Early Education Research. Blackorhy, J., Schiller, E., Mallik, S., Hebbeler, K., Huang, T., Javits, H., . . . Williamson, C. (2010). Patterns in the identification of and outcomes for children and youth with disabilities (NCEE 2010-4005). Washington, DC: U.S. Department of Education National Center for Education Evaluation and Regional Assistance. Bruder, M. B. (2005). Service coordination and integration in a developmental systems approach to early intervention. In G. J. Guralnick (Ed.), A developmental systems approach to early intervention: National and international perspectives (pp. 29-58). Baltimore, MD: Brookes. Buysse, V., Wesley, P W., Bryant, D., 6£ Gardner, D. (1999). Quality of early childhood programs in inclusive and noninclusive settings. Exceptional Children, 65, 301-314. Buysse, V., Wesley, P W, Keyes, L., & Bailey, D. (1996). Assessing the comfort zone of childcare teachers in serving young children with disabilities. Journal of Early Intervention, 20, 209-210. http://dx.doi.org /10.1177/105381519602000301 Campbell, FA., & Ramey, C. T. (1995). Cognitive and school outcomes for high-risk African-American students at midlife adolescence: Positive effects of early intervention. American Educational Research Journal, 32, 743-772. http://dx.doi.org/10.3102/000283120 32004743 Carlson, E., Daley, T., Bitterman, A., Heinzen, H., Keller, B., Markowitz, J., & Riley, J. (2009). Early school transitions and the social behavior of children with disabilities: Selected findings from the Pre-Elementary Education Longitudinal Study. Washington, DC: National Genter for Special Education Research. Ghase-Lansdale, L., MofFitt, R., Lohman, B., Cherlin, A., Goley, R., Pittman, L., . . . Vbtruba-Drzal (2003). Mothers' transitions from welfare to work and the well heing of preschoolers and adolescents. Science, 299, 1548-1552. http://dx.doi.org/10.1126/science .1076921 Gonyers, L. M., Reynolds, A. J., & Ou, S-R. (2003). The effect of early childhood intervention and suhse- 4 8 6 Summer 2012 quent special education services: Eindings from the Chicago Parent-Child Centers. Educational Evaluation and Policy Analysis, 25, 75-95. http://dx.doi.org /10.3102/01623737025001075 Division for Early Childhood, Council for Exceptional Children. (2007). Promoting positive outcomes for children with disabilities: Recommendations for curriculum, assessment, and program evaluation. Missoula, MT: Author Duncan, C. J. Dowsett, C. J. Claessens, A. Magnuson, K. Huston, A. C, Klebanov, P., . . . Brooks-Gunn, J. (2007). School teadiness and later achievement. Developmental Psychology, 43, 1428-1446. http://dx.doi .org/10.1037/0012-l649.43.6.l428 Dunst, C. J., Snyder, S. W, & Mankinen, M. (1989). Efficacy of early intervention. In M. C. Wang, M. C. Reynolds, ÔC G. H. Walbet (Eds.), Handbook of special education: Research and practice—Low incidence conditions (pp. 259-294). New York, NY: Pergamon. Dunst, C. J., &Trivette, C. M. (1997). Early intervention with young at-risk children and their families. In R. Ammerman & M. Hetsen (Eds.), Handbook of prevention and treatment with children and adolescents: Intervention in the real world (pp. 157—180). New York, NY: Wiley. Espy, K., McDiarmid, M. M., Cwik, M. E, Stalets, M., Hamby, A., & Stern, T. E. (2004). The conttibution of executive functions to emergent mathematic skills in preschool children. Developmental Neuropsychology, 26, 465-486. http://dx.doi.org/10.1207/sl5326942d n2601_6 Earran, D. (2000). Another decade of intervention for children who are low-income and disabled: What do we know now? In J. Shonkoff & S. Meisels (Eds.), Handbook of early intervention (2nd ed.). New York, NY: Cambridge University Press. Cathercole, S. E., Alloway, T. P, Willis, C, & Adams, A. (2006). Working memory in children with reading áisdihW'iúes. Journal of Experimental Child Psychology, 93, 265-281. http://dx.doi.Org/10.10l6/j.jecp.200 5.08.003 Gietzen, D., & Vetmeersch, J. A. (1980). Health status and school achievement of children from Head Start and free school lunch programs. Public Health Report, 95, 362-368. Gormley, W. (2011, January). Regression discontinuity: A view from Tulsa. Presentation at the Workshop on Regression Discontinuity Methods in Prekindergarten Evaluations, Cambridge, MA. Gormley, W, & Gayer, T. (2005). Promoting school readiness in Oklahoma: An Evaluation of Tulsa's pre-K . Journal of Human Resources, 40, 533—558. Gormley, W, Gayer, T., Phillips, D., & Dawson, B. (2005). The effects of universal pre-K on cognitive development. Developmental Psychology, 41, 872-884. http://dx.doi.Org/10.1037/0012-l649.4l.6.872 Gormley, W., Phillips, D., &c Gayer, T. (2008). Preschool programs can boost school teadiness. Science, 320, 1723-1724. http://dx.doi.org/10.1126/science .1156019 Gormley, W, Phillips, D., Newmatk, K., Welti, K., & Adelstein, S. (2011). Social-emotional effects of eatly childhood education programs in Tulsa. Child Development, 82, 2095-2109. http://dx.doi.0rg/lO.llll/j. 1467-8624.2011.01648.x Graham, M. A., &C Btyant, D. M. (1993). Characteristics of quality, effective service delivery systems for children with special needs. In D. M. Btyant & M. A. Graham (Eds.), Implementing early intervention: From research to effective practice (pp. 233-252). New York, NY: Guilfotd. Guarino, C, Buddin, R., Pham, C, & Cho, M. (2007). Working paper: Characteristics associated with the early identification of children with special needs in California (WR-439-EDU). Pittsburgh, PA: RAND Education. Guralnick, M. J. (1997). Second-generation research in the field of early intervention. In M. Guralnick (Ed.), The effectiveness of early intervention (pp. 3—23). Baltimore, MD: Brookes. Guralnick, M. J. (1998). Effectiveness of early intervention for vulnerable children: A developmental perspective. American Journal of Mental Retardation, 1.02, 319-345. http://dx.doi.org/10.1352/0895-8017 (1998)102<0319:EOEIEV>2.0.CO;2 Guralnick, M. J. (2001). Social competence with peers and early childhood inclusion. In M. J. Guralnick (Ed.), Early childhood inclusion: Eocus on change (pp. 3-35). Baltimore, MD: Brookes. Guralnick, M.J. (2005a). Early intervention for children with intellectual disabilities: Current knowledge and futute prospects. Journal of Applied Research in Intellectual Disabilities, 18, 313-324. http://dx.doi.org /10.1111/J.1468-3148.2005.00270.X Guralnick, M.J. (Ed.). (2005b). A developmental systems approach to early intervention: National and international perspectives. Baltimore, MD: Btookes. Guralnick, M.J. (2011). Why early intervention works: A systems perspective. Infants and Young Children, 24, 6-28. http://dx.doi.org/10.1097/IYC.0b013e31820 02cfe Hebbeler, K., Spiker, D., Bailey, D., Scarborough, A., Mallik, S., Simeonsson, . . . Nelson, L. (2007). Early Exceptional Children 4 8 7 intervention for infants and toddlers with disabilities and their families: Participants, services and outcomes, Einal report of the National Early Intervention Longitudinal Study. Menlo Park, CA: SRI International. Henry, C, Cordon, C. S., & Rickman, D. K. (2006). Early education policy alternatives: Comparing quality and outcomes of Head Start and state prekindergarten. Educational Evaluation and Policy Analysis, 28, 77-97. hrtp://dx.doi.org/10.1002/pam.20158 Henry, G. T., & Rickman, D. K. (2007). Do peers influence children's skills development in preschool? Economics of Education Review, 26, 100—112. http://dx.doi.org/10.1016/j.econedurev.2005.09.006 Hill, J. L., Brooks-Gunn, J., & Waldfogel, J. (2003). Sustained effects of high participation in an early intervention for low-birth-weight premature infants. Developmental Psychology, 39, 730-744. http://dx.doi.org /10.1037/0012-1649.39.4.730 Holahan, A., & Costenbader, V. (2000). A comparison of developmental gains for preschool children with disabilities in inclusive and self-contained classrooms. Topics in Early Childhood Special Education, 20, 224-236. http://dx.doi.org/ 10.1177/027112140002000403 Howes, C, Burchinal, M., Pianta, R., Bryant, D., Early, D., Clifford, R., & Barbarin, O. (2008). Ready to learn? Children's pre-academic achievement in pre- Kindergarten programs. Early Childhood Research Quarterly 23, 27-50. http://dx.doi.0rg/lO.lOl6/j.ecresq .2007.05.002 Imbens, G. W., & Lemieux, T. (2008). Regression discontinuity designs: A guide to practice, fournal of Econometrics, 142, 615-635. htrp://dx.doi.org/10.1016 /j.jeconom.2007.05.001 Kochanek, T. T, & Buka, S. L. (1999). Influential factors in inclusive versus non-inclusive placement for preschool children with disabilities. Early Education & Development, 10, 191-208. http://dx.doi.org/10.1207 /sl5566935eedlOO2_5 Krauss, M. W, & Hauser-Cram, P (1992). Two generations of family research in early intervention. In M. Guralnick (Ed.), The effectiveness of early intervention (pp. 611-624). Baltimore, MD: Brookes. Lee, D. S., & Lemieux, T. (2010). Regression discontinuity designs in economics, fournal of Econometric Literature, 48, 281-355. http://dx.doi.org/10.1257 /jel.48.2.281 Li-Grining, C. P., Raver, C. C, & Smith-Donald, R. (2010). School readiness and academic competence during preschool: Evidence from CSRP. Unpublished manuscript, Loyola University Chicago, Illinois. Liebman, J. L., & Goodman, J. F. (1995). Learning in early intervention programs: The generalization and maintenance of IEP objectives. Early Education and Development, 6, 127-143. http://dx.doi.org/10.1207 /sl5566935eed0602_2 Lordeman, A., & Jones, N. L. (2010). The Individuals with Disabilities Education Act (IDEA): Interactions with selected provisions of the No Child Left Behind Act (NCLB) (Congressional Research Service, No. 7-5700, RL32913). Washington, DC: U.S. Government Printing Office. Markowitz, J. B., Hehbeler, K., Larson, J. C, Cooper, J. A., & Edmisten, P. (1991). Using value-added analysis to examine short-term effects of early intervention. fournal of Early Intervention, 15, '377-389. hrtp://dx.doi.org/10.1177/105381519101500406 Mashburn, A. J., Pianta, R. C, Hamre, B. K., Downer, J. T, Barbarin, O., Bryant, D., . . . Howes, C. (2008). Measures of classroom quality in prekindergarten and children's development of academic, language, and social skills. Child Development, 79, 732-749. http://dx.doi.Org/10.llll/j.l467-8624.2008.01154.x Mather, N., & Woodcock, R. (2001). Examiner's manual, Woodcock-fohnson Achievement Test-Ill. Chicago, IL: Riverside. McDonnell, A. p., Brownell, K., & Wolery, M. (1997). Teaching experience and specialist support: A survey of preschool teachers employed in programs accredited by NAEYC. Topics in Early Childhood Special Education, 17, 263-285. http://dx.doi.org/10.1177/027112l4 9701700303 Newborg, J. (2004). Battelle Developmental Inventory, Second Edition. Rolling Meadows, IL: Riverside Publishing. Odom, S. L, & Diamond, K. E. (1998). Inclusion of young children with special needs in early childhood education: The research base. Early Childhood Research Quarterly, 13, 3-25. http://dx.doi.org/10.1016 /S0885-2006(99)80023-4 Odom, S. L., Pungello, E. P, & Gardner-Neblett, N. (Eds.). (In press). Implications of Developmental Health Science for Infant/Toddler Care and Poverty. The Guilford Press. Odom, S. L., Vitztum, J., Wolery, R., Lieber, J., Sandall, S., Hanson, M. J., . . . Horn, E. (2004). Preschool inclusion in the United States: A review of research from an ecological systems perspective, fournal of Research on Special Educational Needs, 4, 17-49. http://dx.doi.org/10.1111 /J. 1471 -3802.2004.00016.x Palfrey, J. S., Singer, J. D., Walker, D.K., & Buder, J. A. (1987) Early identification of children's special needs: A study in five metropolitan communiries./o«r- 4 8 8 Summer 2012 nal of Pediatrics, 111, 651-659. http://dx.doi .org/10.10l6/S0022-3476(87)80238-X Phillips, D. A., Gormley, W, & Lowenstein, A. (2009). Inside the pre-kindergarten door: Glassroom climate and instructional time allocation in Tulsa's pre-K programs. Early Childhood Research Quarterly, 24, 213-228. http://dx.doi.Org/10.1016/j.ecresq.2009 .05.002 Puma, M., Bell, S., Cook, R., Heid, G., Lopez, M., Zill, N., . . . Bernstein, H. (2005, May). Head Start impact study: First year findings. Washington, DC: U.S. Department of Health and Human Services, Administration for Children and Families. Redden, S. G., Ramey, S. L., Ramey, G. T., Forness, S. R., & Brezausek, G. M. (2003). Special education placements among former Head Start children: A descriptive multi-site study. Education & Treatment of Children, 26, 1-21. Royston, P. (2004). Multiple imputation of missing values. The Stata Journal, 4, 227-241. Royston, P. (2005a). Multiple imputation of missing values: Update. The Stata Journal, 5, 1-14. Royston, P. (2005b). Multiple imputation of missing values: Update of ice. The Stata Journal, 5, 527-536. Rubin, D. B. (1996) Multiple imputation after 18+ years (With discussion). Journal of the American Statistical Association, 91, 473-489. http://dx.doi.org /10.2307/2291635 Schulting, A. B., Malone, P S., & Dodge, K. A. (2005). The effect of school-based kindergarten transition policies and practices on child academic outcomes. Developmental Psychology, 41, 860-871. http://dx. doi.org/10.1037/0012-l649.4l.6.860 Schwartz, I. S., Garra, J. J., & Grant, S. (1996). Examining the use of recommended language intervention practices in early childhood special education classrooms. Topics in Early Childhood Special Education, 16, 251-272. http://dx.doi.org/10.1177/027112l4 9601600208 Schweinhart, L. J., Barnes, H. V, & Weikart, D. P (1993). Significant benefits: The High/Scope Perry Preschool study through age 15. Ypsilanti, MI: High/Scope. ShonkofF, J. P, & Hauser-Gram, P (1987). Early intervention for disabled infants and their families: A quantitative analysis. Pediatrics, 80, 650-658. Shonkoff, J. P, & Meisels, S. J. (2000). Handbook of early childhood intervention. New York, NY: Gambridge University Press. Smith, T., Kleiner, A., Parsad, B., & Farris, E. (2003) Prekindergarten in U.S. public schools: 2000-2001. Washington, DG: Department of Education, National Center for Education Statistics. Tuckwiller, E. D., Pullen, P G., & Goyne, M. D. (2010). The use of the regression discontinuity design in tiered intervention research: A pilot study exploring vocabulary instruction for at-risk kindergartners. Learning Disabilities Research & Practice, 25, 137-150. http://dx.doi.Org/10.llll/j.1540-5826.2010.00311.x U.S. Department of Education. (2006). Twenty-eighth annual report to Congress on the implementation of the Individuals with Disabilities Education Act. Washington, DG: Author. Retrieved from http://www2.ed.gov /about/reports/annual/osep/2006/parts-b-c/index.html U.S. Department of Education, National Genter for Education Statistics. (2007). The nation's report card. Retrieved from http://nces.ed.gov/nationsreportcard/ U.S. Departnient of Education, Office of Special Education and Rehabilitative Services, Office of Special Education Programs. (2005, April). Twenty-fifih annual report to Congress on the irnplementation of the Individuab with Disabilities Education Act, Vol. 1. Washington, DG: Author. U.S. Department of Health and Human Services, Administration for Ghildren and Families (2010, January). Head Start impact study. Einal report. Washington, DG: Author. Vaughn, S., Wanzek, J., Murray, G. S., Scammacca, N.; & Linan, S. (2009). Response to early reading intervention: Examining higher and lower responders. Exceptional Children, 75, 163-183. Wakeman, S. Y, Browder, D. M., Meier, I., & McGoll, A. (2007). The implications of No Ghild Left Behind for students with developmental disabilities. Mental Retardation and Developmental Disabilities Research Reviews, 13, 143-150. http://dx.doi.org/10.1002 /mrdd.20147 Whitebook, M., Phillips, D., Bellm, D., Growell, N., Alamzaraz, M., & Jo, J. Y. (2004). Two years in early care and education: A community portrait of quality and workforce stability. Berkeley, GA: Genter for the Study of Ghild Gare Employment. Wolery, M., & BaUey, D. B. (2002). Early childhood special education research. Journal of Early Intervention, 25, 88-99. http://dx.doi.org/10.1177/10538151020 2500204 Wolery, M., Holcombe, A., Brookfield, J., Huffman, K, Schroeder, G., Martin, G. G., . . . Fleming, L. A. (1993). The extent and nature of preschool mainstreaming: A survey of general early educators. The Journal of Special Education, 27, 222-234. http://dx.doi .org/10.1177/002246699302700205 Exceptional Children 4 8 9 Wolery, M., Holcombe, A., Venn, M. L., Brookfield, J., Huffman, K., Schroeder, G., . . . Fleming, L. A. (1993). Preschool mainstreaming: Current status and relevant issues. Young Children, 49(1), 78-84. Wolery, M., &c Wilbers, J. S. (1994). Including children with special needs in early childhood programs. Washington, DG: National Association for the Education of Young Ghildren. Wong, V. G., Gook, T. D., Barnett, W S., & Jung, K. (2008). An effectiveness-based evaluation of five state pre-kindergarten programs. Journal of Policy Analysis and Management, 27, 122—154. http://dx.doi.org /10.1002/pam.20310 Woodcock, R. W, McGrew, K. S., & Mather, N. (2001). Woodcock-Johnson III NU tests of achievement. Rolling Meadows, IL: Riverside. Woodcock, R. W, Muñoz-Sandoval, A. F., McGrew, K. S.; Ô£ Mather, N. (2005). Batería IIIWoodcock-Muñoz. Rolling Meadows, IL: Riverside. ABOUT THE AUTHORS DEBORAH A. PHILLIPS, Professor; and MARY E. MELOY (DC CEC), Doctoral Student, Department of Psychology, Georgetown University, Washington, DC. Address correspondence concerning this article to Deborah A. Phillips, Psychology Department, Georgetown University, 302F White Gravenor Hall, 3700 O St. NW, Washington, DC 20007 (e-mail: Deborah.dap4@gmail.com). This research was conducted by scholars affiliated with the Center for Research on Children in the United States at Georgetown University. We received helpful comments from Garolyn Hill, William Gormley, Michael Guralnick, and W. Steven Barnett. We greatly appreciate the cooperation we have received from the Tulsa Public Schools and its teachers and principals. Finally, we thank the Foundation for Child Development, the David and Lucile Packard Foundation, the Spencer Foundation, the A. L. Mailman Family Foundation, the National Institute for Early Education Research, and the Pew Charitable Trusts for their generous financial support. We alone are responsible for the contents of this report. • Manuscript received February 2011; accepted June 2011. ADVERTISING INFORMATION EXCEPTIONAL. CHILDREN Exceptional Cbildren (EC), published four times a year, is one of the official publications of CEC. EC is a peer-reviewed journal that publishes original research on the education and development of infants, toddlers, children, and youth with exceptionalities and articles on professional issues of concern to special educators. Insightful pioneering research, topical issues, and broad perspectives by leaders in the field forimore than 75 years have made EC one of the most respected scholarly journals in special education. Classifieds are $24 per line with a 12-line minimum. Black and white ads start at $773. Other advertising opportunities TEACHING Exceptional Children New! Online Advertising CEC SmartBrief specialedcareers. org To download our marketing tool kit, go to www.cec.sped.org/advertising or contact Victor Erickson at 703-264-9454 or via e-mail at victore@cec.sped.org INDEX OF ADVERTISERS Council for Exceptional Children, cover 2, 389, 448, 470, cover 3 4 9O Summer 2012 Copyright of Exceptional Children is the property of Council for Exceptional Children and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. PLACE THIS ORDER OR A SIMILAR ORDER WITH US TODAY AND GET AN AMAZING DISCOUNT :)