Abstract
The purpose of this study is to examine the effects of Head Start on early literacy skills relevant to school readiness of English language learners compared to their peers. The comparisons of literacy outcomes were conducted between English language learners and non-English language learners when both groups participated and were not in Head Start. A total of 47 covariates were involved in propensity score analysis, and average treatment effects for the treated individuals were used to estimate the literacy outcome differences from the comparisons. The results indicated that early literacy outcomes of English language learners and non-English language learners were significantly different in recognizing alphabetic letters and rhyming words regardless of whether or not both English language learners and non-English language learners attend in Head Start. Being in Head Start did not contribute to reducing the gap between English language learners and non-English language learners. Finally, the limitations of this study and future directions for research and practice are discussed.
As the number of children from culturally and linguistically diverse (CLD) backgrounds in US schools increases (August and Shanahan, 2006), educators and policymakers have become increasingly concerned with developing educational programs that advance equitable outcomes for CLD students. The accountability movement, especially No Child Left Behind (NCLB), has focused schools on reducing achievement gaps, such that all students are prepared for academic success (Barnett et al., 2004). A large proportion of CLD students who are at risk of academic failure are English language learners (ELLs; Klingner and Artiles, 2006; Lesaux and Siegel, 2003; Zill, 1995).
ELLs have an especially high risk for poor literacy skills because the languages they use at home are different from those taught in school and because of limited exposure to English (Hammer and Miccio, 2006; National Center for Education Statistics, 2011). The impact of limited English proficiency on the achievement gap increases with age, as students are increasingly expected to use English to learn content in other subject areas (Coyne et al., 2001). Moreover, the challenges experienced by ELLs are not only linguistic; they are also intricately linked with socioeconomic status (SES; Hammer and Miccio, 2006; Roberts et al., 2010). Children from low-SES homes and neighborhoods have limited opportunities to develop essential school readiness skills compared to children from higher SES communities. Given that many ELLs are from disadvantaged backgrounds (Hammer and Miccio, 2006; Garcia, 2000), these students are at especially high risk of academic failure (Klingner and Artiles, 2006; Lesaux and Siegel, 2003).
Because the achievement gap increases with age, early intervention is essential for reducing the persistent underachievement of ELL students in the area of literacy (Klingner and Artiles, 2006; Lesaux and Siegel, 2003; Zill, 1995). Scholars have consistently demonstrated that early interventions have a more significant and durable impact on students’ learning than interventions in later years (Coyne et al., 2001). Head Start (HS) is intended to fulfill this role, by providing comprehensive early child development services to low-income children and families, thus promoting school readiness for students from low-SES backgrounds (Currie and Thomas, 1995). Scholars are interested in the effectiveness of HS, but most of the research on HS programs has focused on native speakers, despite the fact that many preschool ELL students attend HS classrooms (Hammer et al., 2003).
Unfortunately, there is little evidence about the school readiness skills of ELLs before entering kindergarten; more specifically, prior research on the effectiveness of HS has failed to consider the possibility that HS may have differential outcomes on school readiness skills for ELL students. Given the public investment in HS programs, it is essential to determine whether HS is effective at increasing the school readiness of the many ELL students who are served in HS. Therefore, the primary purpose of this study is to investigate the effects of HS on the literacy-related school readiness skills of ELL students. Specifically, we examine the following research questions:
Among children who participated in HS, do literacy-related school readiness skills differ between ELLs and non-ELLs?
Among children who did not participate in HS, do literacy skills differ between ELLs and non-ELLs?
Among ELLs, do early literacy skills differ between those who did and did not attend HS?
This study used propensity score (PS) methods to deal with the fact that systematic differences between children who are ELL and who are not ELL are likely to impact school readiness skills, as are systematic differences between children who do and do not attend HS. Rosenbaum and Rubin (1983) define PSs as the predicted probability of assignment to a condition given observed covariates. When comparing two groups who systematically differ in ways that are related to the outcome of interest, PSs can be calculated based on observed covariates. The observed difference between treatment and control conditions at a given PS level is an unbiased estimate of the treatment effect (Rosenbaum and Rubin, 1983). We used two PS methods, optimal propensity score matching (OPSM) and propensity score weighting (PSW), to deal with systematic differences between students who are and are not ELLs and between students who did and did not attend HS.
Early intervention services and the effectiveness of HS
Substantial evidence indicates that early literacy skills are crucial for lifelong learning outcomes (Duncan et al., 2007; Gauntlett et al., 2001; Scarborough, 2001). Early experiences affect academic and social development throughout school years; as such, providing effective literacy interventions in the early stages of a child’s life increases the developmental and educational gains for the child far more than the provision of such services later in life (Gauntlett et al., 2001). Moreover, early intervention services are more effective than those provided later (Coyne et al., 2001). For instance, Wanzek and Vaughn (2007) conducted a meta-analysis to compare the effects of intensive, evidence-based literacy interventions for struggling readers in kindergarten through third grade to equally intensive, evidence-based interventions for struggling readers in 4th through 12th grades. They found that the effect sizes in the early grades were significantly larger than effect sizes in later grades. Effect sizes for comprehension, for example, were .46 in early grades, and .14 in later grades. Such findings have revealed that early intervention services are essential for enabling children who are at risk of learning challenges to achieve many academic goals.
The purpose of HS is to enhance the lifelong learning outcomes of students who are at risk of school failure, by providing early intervention services on a national scale. HS provides a variety of services, including health, nutritional, educational, and social services, for young children and their families, in order to support children’s social and cognitive development. HS specifically targets their services to promote reading and math-related school readiness skills for preschool students from low-SES backgrounds (US Department of Health and Human Services, Administration for Children and Families, 2010).
Much of the research on HS indicates that it is successful in advancing literacy-related school readiness skills. According to the report A Year in HS, children who attended HS made significant progress in developing letter–word awareness compared to same-age peers (Aikens et al., 2010). When compared to children who were not enrolled in HS, 3-year-olds in HS demonstrated more improvement in letter–word knowledge and 4-year-olds showed more development in comprehensive vocabulary, early writing, and mathematics abilities while attending HS for their first year (West et al., 2010). West et al. (2010) found that children who attended HS showed improvement in a number of school readiness skills, including literacy; the children’s literacy skills during their final period of HS attendance were significantly correlated with the children’s literacy development at the end of their kindergarten year. Similarly, Lee (2011) studied the relationship between children’s outcomes and durations of enrollment in HS. He found that children who attended HS for 2 years (e.g. from ages 3 to 5) had better outcomes in a variety of subjects, including literacy, compared to children who were enrolled in HS for only 1 year. This result suggests that early exposure to HS, which leads children to attend HS for longer, has positive effects on children’s school readiness skills. As a result of these findings, Lee (2011) argued that children are likely to receive more advantages when they begin attending HS at an earlier age and stay longer. Although the long-term benefits of HS are still a topic of debate (e.g. Currie and Thomas, 1995; Ludwig and Phillips, 2008), the collective evidence on the early benefits of HS is quite strong (Deming, 2009).
Preschoolers’ school readiness
The meaning of school readiness has been a controversial issue in the field of early childhood education, particularly among early childhood teachers, policymakers, and researchers. Some scholars agree that school readiness indicates the fundamental abilities and skills, in both academic and social domains, that enable children to smoothly transition to school settings (Hair et al., 2006). Other scholars disagree, contending that school readiness implies both academic and social aspects of development (Edwards, 1999, Lewit and Baker, 1995). However, the debate surrounding the term focuses on the more “traditional” definition of cognitive readiness versus a broader definition of school readiness that considers regulatory readiness more important for academic success than cognitive readiness (Blair, 2002; Blair and Diamond, 2008). Despite these debates, most researchers agree that school readiness supports long-term academic success. However, the curricula and strategies to achieve academic success are different depending on the definition used. For example, the Tools of the Mind program (Bodrova and Leong, 2007) is a curriculum based on promoting regulatory readiness to achieve academic success. In this article, the more traditional definition of school readiness (cognitive skills) is emphasized, focusing more on academic readiness skills, especially literacy-related skills (e.g. reading, writing) that are closely connected to academic achievement and educational outcomes.
Recent studies have shown that young children in kindergarten classrooms who attended preschool had much higher academic achievement than those who did not attend preschool (Perry, 1999; Warden, 1998), leading scholars to conclude that providing a developmentally appropriate and well-designed preschool program is vital for helping preschoolers transition successfully to kindergarten. A large volume of studies have shown that a well-designed preschool program also plays a crucial role in facilitating young children’s cognitive development (Garces et al., 2002; Mushel, 2001). As a result, educational scholars and policymakers have become increasingly focused on developing children’s school readiness skills, and states have increased their investment in preschool programs (Schumacher et al., 2005).
Despite ample evidence of the general importance of preschool, research about ELL students’ participation in preschool is very limited. Hammer et al. (2007) investigated the relationship between the receptive language development of bilingual children attending HS and their later reading outcomes in kindergarten. Results showed that children’s early reading abilities in Spanish were nearly one standard deviation below the test mean or lower. Their findings imply that more intensive and comprehensive preschool programs are needed to target ELL children’s improvement in English. Therefore, this study specifically purports to extend existing knowledge regarding the literacy-related school readiness skills of ELL students who participated in HS program.
Methods
Data sources and data collection instruments
In an effort to quantify the effects of HS on ELLs’ early literacy skills, data from the School Readiness Survey of the 2007 National Household Education Surveys Program (SR-NHES) were used. The SR-NHES is a large-scale, nationally representative sample of US preschool children. This survey focuses on the experiences and developmental accomplishments of children who were 3–6 years old as of 31 December 2006 and who had not started kindergarten yet. The survey included information about (a) the participation of young children in preschool or other types of center-based care or educational programs; (b) parental plans for kindergarten enrollment and their assessment of what they believe they should do to prepare their children for kindergarten; (c) children’s developmental accomplishments and difficulties, including emerging literacy and numeracy skills; (d) family activities with children in and out of the home; and (e) children’s television-viewing habits (National Center for Education Statistics (NCES), 2007).
The SR was conducted by trained interviewers, over the phone. First, a screener survey was used to determine eligibility for the extended interviews, which collected data on the areas of interest. The respondent for the SR interview was the parent or guardian in the household who was the most knowledgeable about the child’s care and education (NCES, 2007).
Participants
The term ELL indicates a person who is in the process of acquiring English, has a first language other than English, and has had limited exposure to English (Klingner et al., 2006). This group of children is also referred to as “limited English proficient” in the NCLB Act of 2001, indicating individuals who are still developing proficiency in English. This term, therefore, may include both children who are beginning to learn English, and those who have already learned some English, but still are limited in English proficiency (LEP). The single most common reason why students are considered LEP is limited exposure to an English-speaking environment (Hammer and Miccio, 2001). While native-speaking students benefit from frequent exposure to English language and literacy both at home and at school, LEP students mainly speak a language other than English at home because their parent(s) or caregivers do not speak English when communicating with them. Hence, children’s use of their home language when communicating with their parents or caregivers can be considered one of the important factors that impact their English language development and academic achievement using English literacy. These criteria provide guidance for defining the ELL group.
The variable on languages children speak most at home (CSPEAK) was used to define a group of ELLs. If a respondent answered that their child uses English or English and another language equally at home, these children were not included in the group of ELLs. When a child used other languages more often than English at home, she or he was considered an ELL. The variable indicating whether or not a child ever attended HS (e.g. HS, early HS, or home HS) was used to specify the type of program the child attends.
Study design and data analysis
An experimental design with random assignment is the optimal way to remove bias from unobserved variables when estimating the differences in outcomes between two groups. However, when experimental designs are not feasible, as is often the case in educational research, quasi-experimental designs are appropriate. PS methods can be the most effective way to deal with selection bias in quasi-experimental studies (Caliendo and Kopeinig, 2008; Heckman et al., 1998). The following sections explain the concept and implementation of PSs and specify the criteria used to conduct PS methods for this investigation.
OPSM and PSW
Rubin’s (1974) potential outcome framework is commonly used to understand causal estimation in observational studies. The basic principle of this framework is that treated and control individuals have potential outcomes in both the presence and the absence of treatment. To illustrate, let
The estimate of the average treatment effects for the treated (ATT) individuals will only be unbiased when two assumptions are met. First, the assignment to the treatment should be independent of the potential outcomes. This condition is known as strong ignorability of treatment assignment (Rubin, 1974). Random assignment meets this assumption, but in social research, the randomized experiment is frequently either unfeasible or unethical. Second, it is also necessary that the stable unit treatment value assumption (SUTVA) should be also met. This assumption requires that the potential outcome of one unit should be unaffected by the particular treatment assignment of other units (Guo and Fraser, 2010). In observational studies, both strong ignorability of treatment assignment and SUTVA may be violated, which leads to biased estimates of the ATT and poor internal validity of the study (Shadish et al., 2002). However, if the selection mechanism or outcome is modeled completely, the estimate of the ATT can be unbiased (Guo and Fraser, 2010). PS methods attempt to model the selection mechanism completely, thereby enabling unbiased estimates of the ATT. Regarding the second assumption, we assume that data do not have a nested nature, and we have also employed the sampling weights in the analysis to assure the generalizability of the results.
PSs are typically estimated using logistic regression (Guo and Fraser, 2010). For this study, we have estimated PSs using Bayesian logistic regression because the logistic regression algorithm did not converge for any comparison. After PSs are calculated, the treated and control group participants are matched on the PSs using one or more of a variety of matching methods. OPSM was developed by Rosenbaum (2002) and Hansen (2004, 2007). This method is superior to other PS methods, such as greedy matching, because it uses a network flow approach, developing “matched sets in such a way that the matching optimizes or minimizes the total distance for a given dataset and prespecified structure” (Guo and Fraser, 2010: 151). Therefore, we first conducted the analysis using OPSM.
In addition, we also conducted the analysis a second time with a different PS method because if we are able to draw the same conclusions using both methods, the results will be cross-validated and we will have greater confidence in the findings of this study. Furthermore, we will see the results of two methods, one of which uses part of the sample (OPSM) and the other of which uses the whole sample (PSW).
The second method selected was PSW. Hernán et al. (2000) extended the concept of inverse probability weighting, creating PSW as a method to control for selection bias in observational studies. To illustrate, in order to estimate ATT, let Ti be the treatment indicator with Ti = 1 as a treatment group participant and Ti = 0 as a control group participant,
Rationale for covariates selected in the model
PS methods, as described above, attempt to reduce the confounding effects of measured covariates. PS values are dependent on a vector of observed covariates that are associated with treatment assignment. Therefore, selecting appropriate covariates to be included in the PS model is critical for reducing potential bias. In order to reduce bias, the covariates used in the PS model should, ideally, be true confounders that are related to both the selection mechanism and the outcomes. However, covariates that are related to the outcome alone will also reduce bias (Brookhart et al., 2006). The following section explains the covariates used in the PS models for this investigation.
There have been numerous attempts to determine factors that influence school readiness and early academic achievement. Different researchers may conceive of significant factors for early literacy development slightly differently. Despite differences in scholars’ understanding of the factors that influence early literacy development, several common variables are associated with potential academic outcomes in school-aged children’s reading skills. For this analysis, eight categories of covariates were selected, based on research demonstrating a relationship between these covariates and students’ reading skills: (a) demographic characteristics (Ramani and Siegler, 2011; Scarborough and Dobrich, 1994; Scarborough et al., 1991; Zill, 1995); (b) developmental or emotional characteristics (Coplan et al., 1999); (c) impacts of native language characteristics (Aunio et al., 2008, 2009); (d) frequency of home literacy activities and the time spent on reading with family (Bus et al., 1995; Evans et al., 2000; Taylor, 1983; Whitehurst et al., 1994); (e) reading-related materials and home literacy environment (Burgess et al., 2002; Dickinson and McCabe, 2001; Mason, 1980; Snow et al., 1991; Weinberger, 1996; Williams and Rask, 2003); (f) preschooler’s play and experience outside of school and home (Kim, 1999; Pellegrini, 1984; Roskos, 1990; Roskos and Neuman, 1998); (g) birth-related and parental health–related issues (Malacova et al., 2008); and (h) learning experience in preschools or other programs (Aunio et al., 2008; Hojnoski et al., 2009).
Predictors of literacy skills in ELLs and non-ELLs
To maximally reduce the potential for selection bias, 47 covariates drawn from both theory and prior empirical research were included in the model as follows: (a) demographic characteristics related to the child (e.g. sex, age, location of child’s household); (b) developmental characteristics related to the child (e.g. frequency of being active, or unable to sit, a position of holding a pencil); (c) frequency of home literacy activities and the time spent on reading with family (e.g. frequency or length of reading with family members); (d) reading-related materials and home literacy environment (e.g. parents’ beliefs about the importance of reading, providing feedback when child makes errors); (e) preschooler’s play and experience outside of school and home (e.g. visiting zoo or library, playing sports or games, arts and crafts); (f) birth- and parental-related issues (e.g. parent’s education, parent’s marital status, birth weight).
Outcome measures and ATT
Five early literacy skills used in the SR were selected as outcome measures: recognizing alphabetic letters (RAL), writing first name (WFN), rhyming words (RW), recognizing beginning sounds of words (RBSW), and reading story books (RSB) on their own. OPSM and PSW were used in order to allow for contrasts of early literacy outcomes of children who were ELLs and non-ELLs in HS, as well as of the outcomes of ELLs who did and did not receive HS to estimate the effects of HS for ELL students.
Data analysis procedures
The analysis was undertaken in three steps: (a) estimating the PS using Bayesian logistic regression, (b) implementing OPSM and PSW methods, and (3) computing ATT for each dependent variable. We used optmatch library to perform OPSM (Hansen, 2007) and design library to perform PSW in the R 2.14.0 environment. In order to answer each research question, we selected proper treatment and control groups and ran the whole analysis.
Results
Study’s analytical sample
The final sample of students contained 1900 native English speakers and 280 ELLs. Of the 1900 non-ELLs participants used for the analysis, 1021 were female and 879 were male. Of the 280 students who were identified as ELLs used for the analysis, 166 were female and 114 were male. A large majority of ELLs in this study are Hispanic, followed by Asian or Pacific Islander. The children’s ages ranged from 3 to 6 years, but most children were aged 3–5 years. Table 1 shows essential demographic information of the sample of preschool children selected for this study.
Demographic information for ELLs and non-ELLs.
ELL: English language learners; HS: Head Start.
The numeric values mean the number of children included in each group.
Do early literacy skills differ for ELL students and native speakers who attend HS?
For the first research question, we compared early literacy outcomes of ELL students to the outcomes of native speakers, when both groups participated in HS. The group of ELL students in HS is considered as the treatment group and the group of native speakers in HS as the control group. Using OPSM, the mean probability of being successful in RW (i.e. a child can rhyme words) for children in the ELL group was .355 and for those in the native-speaking group was .786 (p < .001). The mean probability of being successful in RAL (i.e. a child can read all or most of the letters of the alphabet) for children in the ELL group was .313 and for those in the native-speaking group was .669 (p =.005). These findings are also supported by the PSW analysis. When using PSW, the mean probability of being successful in RW for children in the ELL group were .355 and for those in the native-speaking group were about .794, indicating that there is a significant difference between the two groups (p < .001). Similarly, the mean probability of being successful in RAL for children in the ELL group was about .355 and for those in the native-speaking group were .604 (p = .014). Therefore, it was revealed that there is a significant difference between ELL students and native speakers on two literacy skills (specifically RW and RAL), with native speakers showing better performance than ELLs in these two skills; however, no significant difference among other literacy skills (i.e. RBSW, WFN, and RSB) was found between these two groups.
In order to examine differences in early literacy skills between ELL students and native speakers when both groups were not exposed to HS, both OPSM and PSW were used. When estimating the probabilities of being successful in the five literacy skills using OPSM, the mean probability of being correct in RW was .388 for the ELL group and .592 (p < .001) for the native-speaking group. The mean probability of being successful in RBSW was .551 for the ELL group and .648 for native-speaking group (p = .021). The mean probability of being correct in RAL was .287 for the ELL group and .451 for native-speaking group (p = .003). These results were not substantiated by the PSW analysis, which found no significant differences between groups.
The third research question was about the difference, if any, among early literacy skill outcomes between ELL students who were in HS and ELL students who were not exposed HS. Neither OPSM nor PSW captured any significant difference between groups for any dependent variables.
Conclusion and discussion
The primary purpose of this study was to examine the effects of HS on the literacy-related school readiness skills of ELL students. The results demonstrated that for students who attended HS, native English speakers performed significantly better on RAL and RW than ELL students. It is interesting that other comparisons between these two groups did not show any significant results; in other words, native English speakers who attended HS did not perform any better than ELL students who attended HS on the other three literacy skills, RBSW, WFN, and RSB on their own. These findings might be a result of effective HS programs, or they might result from relationships between children’s native language and their second language (e.g. English; Geva et al., 2000; Klingner et al., 2006); early literacy skills such as phonological awareness may be easily transferable between two different languages (Akamatsu, 2003; Lindsey et al., 2003). In other words, at the beginning stage of reading, children’s literacy skills tend to rely on phonological awareness, and both ELL students and native speakers likely have similar level of phonological awareness skills, regardless of their native languages or involvement in preschool programs. This possibility is supported by the findings of Geva et al.’s (2000) study. Their examination of the reading performance of a large sample of native English speakers and ELL students found that a child’s level of success on phonological awareness, rapid naming, and word recognition was similar across both groups. The comparison of ELL students and native speakers who did not attend HS indicates that the gap in early literacy skills exists regardless of HS attendance. Significant differences between ELL students and native speakers who did not attend HS were always in favor of native speakers.
Comparisons of ELL students who did and do not attend HS yielded mostly nonsignificant results. The only significant difference was for the WFN skill; on this skill, ELL students in HS performed better than ELL students who were not exposed to HS before entering kindergarten. These results indicate that HS is effective at teaching ELL students to write their first name, but participating in HS does not make a difference in ELL students’ other early literacy outcomes, including RAL, RW, RBSW, and RSB.
These findings have implications with respect to the effectiveness of HS at reducing the achievement gap between ELL students and native speakers. The mixture of significant and nonsignificant differences between ELL students and native speakers who attend HS indicates that the gap between these two groups remains for some essential early literacy skills; for those skills in which there is no gap, it is unclear whether HS is responsible for equivalent performance or whether these skills are less impacted by language differences. Furthermore, the nonsignificant results of the comparison between ELL students who do and do not attend HS indicate that HS does not seem to impact the early literacy skills of these students. Collectively, these findings imply that HS might not have reading/literacy programs that are appropriate for ELL students and their diverse language backgrounds. These findings align with the results discussed earlier, of Hammer et al.’s (2007) investigation of bilingual students’ early literacy development in HS programs, which revealed that early reading skills in students’ native language was also well below the mean.
One of the limitations of our findings is that NHES data are collected through a telephone survey of children’s parents or guardians, who may not consistently provide accurate and honest answers to the questions. This form of measurement can result in biased estimates of effects. We suggest that future research on early childhood intervention consider the complexities involved in becoming literate in English and focus more on cultural and contextual factors that affect children’s school readiness before entering schools (Klingner et al., 2006). Also, to better understand the effects and implications of HS for ELL students, future studies should address the nature of the curriculum or specific program used in HS and the associations between these specific programs and student achievement. Such research should, ideally, be conducted using methods that account for cultural and linguistic factors. Estimating causal effects from large datasets using PSs is a way to reduce bias in using nonexperimental designs (Guo and Fraser, 2010), but, in spite of its advantages, PS estimates have limitations. PS methods can only reduce bias associated with observed covariates included in the dataset, and unobserved covariates can increase bias. It is possible that other unknown covariates may affect the validity of our model. The PS methods also require adequate overlap between groups in order to be generalizable to the population. For the first comparison, comparing ELL students to native speakers, all of whom attended HS, 25 of the 45 treatment group participants did not have overlapping PSs. For the second comparison of ELL students who did and did not attend HS, 3 of the 45 treatment group participants did not have overlapping PSs. For the last comparison of ELL students and native speakers, all of whom did not attend HS, 57 of the 235 treatment group participants did not have overlapping PSs. This limits the generalizability of our findings.
Suggestions for improving HS services for school readiness of ELLs
The HS approach to school readiness aims to ensure that children are ready for school, families are ready to support their children’s learning, and schools are ready for children. HS views school readiness as children possessing the skills, knowledge, and attitudes necessary for success in school and for later learning and life (US Department of Health and Human Services, Office of Head Start, 2011b). Historically, HS been a model early childhood program, with a strong, clear, and comprehensive focus on all aspects of healthy development, including physical, cognitive, social, and emotional development, all of which are essential to children’s school readiness. All HS agencies are required to establish school readiness goals, or “the expectations of children’s status and progress across domains of language and literacy development, cognition and general knowledge, approaches to learning, physical health and well-being and motor development, and social and emotional development that will improve readiness for kindergarten goals” and that “appropriately reflect the ages of children, birth to five, participating in the program” (Head Start Act, 2007).
With respect to families, school readiness means families are engaged in the long-term, lifelong success of their child. HS recognizes that parents are their children’s primary teachers and advocates. Programs are required to consult with parents in establishing school readiness goals (Head Start Act, 2007, 45 CFR 1307.3 (b) (1) (iii), as amended). When HS programs and schools work together to promote school readiness and to engage families in the transition to kindergarten, schools are more likely to be ready for children (US Department of Health and Human Services, Office of Head Start, 2011b). In the following sections, we provide suggestions for how HS can better meet these goals for young children and families who are ELLs.
Assessment
HS leaders and researchers interested in the efficacy of HS programs would do well to collect data from practitioners about how they are addressing the needs of young ELL students and their families. How have practitioners in HS programs responded to large influxes of language minority children? Anecdotal reports and informal observations suggest that practitioners may be engaging in a range of practices to address the needs of ELL students, with little guidance regarding the efficacy of their practices (e.g. consistently using young bilingual children as language brokers between the teachers and other ELL students). In addition, more data regarding the needs of practitioners in this area, from their perspective, are required to inform future research efforts (e.g. US Department of Health and Human Services, Office of Head Start, 2010).
Culturally and linguistically responsive curriculum
In general, research on effective programs for ELL students suggests that the most effective programs and curricula are responsive to the needs of specific children, families, and communities (Galinsky, 2006). Programs that integrate children’s home language and home culture result in better overall outcomes for children. Children in preschool programs in which staff spoke children’s home language showed better development of social skills (Chang et al., 2007). Programs that support the development of English and of children’s home language help low-income children who are dual language learners increase proficiency in both languages (Winsler et al., 1999).
HS programs should provide a framework to guide curriculum, assessment, and other programming decisions, keeping in mind that they are serving children who need to continue to develop their first language while they acquire English. Ongoing assessment of children’s progress on each of the domain elements in the HS program is essential. When assessing children who are ELLs, staff should be aware that the purpose of assessment is to learn what a child knows and is able to do. Unless they are specifically assessing a child’s English language development, assessment does not depend on a child’s understanding or speaking abilities in English, but on the specific knowledge, skills, or abilities that the assessment measures. For example, a child can demonstrate an understanding of book knowledge or science concepts in his or her home language. Assessing a child who is ELL only in English will rarely give an accurate or complete picture of what the child knows or can do. HS staff should choose assessment instruments, methods, and procedures that use the language or languages that most accurately reveal each child’s knowledge, skills, and abilities. The assessment data gathered in the home language can be used to inform instructional practices and curriculum decisions in order to maximize the child’s learning. In addition to selecting an appropriate language for assessment, HS programs should use culturally appropriate assessments to capture what children who are ELLs know and can do (US Department of Health and Human Services, Office of Head Start, 2011a).
At the same time, HS programs need to promote the acquisition of English for children who are learning English. The program of English Language Development applies only to these children and contains domain elements that focus on their receptive and expressive language skills and their participation in literacy activities. Children’s progress in learning English will vary depending upon their past and current exposure to English, their temperament, their age, and other factors.
There are still many unresolved issues when determining the best education for ELL students. Considering HS’s important preventive role in supporting students who are at risk, HS should provide educational services that take ELL students’ linguistic and cultural backgrounds into account. In addition, educational policy should rethink not only how to involve ELLs in educational systems but also the best practices for serving ELLs.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.