g model article in press - george mason universitywinslerlab.gmu.edu/pubs/winslereclsb2014.pdf ·...

E

Ela

ALa

b

c

d

e

f

a

ARRA

KEDILE

1

p[aiBb

Cft(

e

h0

ARTICLE IN PRESSG ModelARCHI-697; No. of Pages 15

Early Childhood Research Quarterly xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Early Childhood Research Quarterly

arly development among dual language learners: The roles ofanguage use at home, maternal immigration, country of origin,nd socio-demographic variables�

dam Winslera,∗,1, Margaret R. Burchinalb, Hsiao-Chuan Tienb, Ellen Peisner-Feinbergb,inda Espinosac, Dina C. Castrod, Doré R. LaForettb, Yoon Kyong Kime, Jessica De Feyter f

George Mason University, United StatesUniversity of North Carolina Chapel Hill, United StatesUniversity of Missouri, United StatesArizona State University, United StatesBSC Best Solution for Competency Issues, South KoreaICF International, United States

r t i c l e i n f o

rticle history:eceived 10 January 2012eceived in revised form 15 February 2014ccepted 22 February 2014

eywords:CLS-B

a b s t r a c t

Using nationally representative data from the ECLS-B, we examined children’s outcomes and growth from9 to 65 months as a function of language used in the home at 24 months (English only n = 7300; Englishand another language n = 1500; other language only n = 400). We also examined whether demographicvariables moderated the effects of DLL status in predicting child outcomes. Results revealed substantialvariation within the DLL population within and across language groups in immigration status, heritagecountry, child outcomes, and family socioeconomic risk. DLL status was associated with differential out-

ual language learners (DLLs)mmigrantanguagearly childhood

comes, gains over time, and processes in complex ways. Maternal birth outside of the U.S., child gender,and parental education moderated relations between home language and child outcomes. Use of theheritage language at home served as a protective factor for children of immigrant families for a fewoutcomes. Gender and parental education were more strongly associated with child outcomes amongEnglish-speaking households than among DLLs.

© 2014 Elsevier Inc. All rights reserved.

. Introduction

Children who are dual language learners (DLLs) are in therocess of learning a majority language (i.e., English in the U.S.L2]) in addition to a (minority) language (L1) that is spokent home. About 25% of children in the U.S. grow up in a home

Please cite this article in press as: Winsler, A., et al. Early develouse at home, maternal immigration, country of origin, and socio-dehttp://dx.doi.org/10.1016/j.ecresq.2014.02.008

n which a language other than English is used (U.S. Censusureau, 2010). The number of DLL children in schools doubledetween 1997 and 2008, and continues to grow faster than the

� This paper was completed as part of the secondary data analysis activities of theenter for Early Care and Education Research: Dual Language Learners (CECER-DLL)

unded by the Office of Planning, Research, & Evaluation (OPRE) in the Administra-ion for Children & Families (ACF), U.S. Department of Health and Human ServicesDHHS) Cooperative Agreement 90YR0041/01.∗ Corresponding author.

E-mail address: [email protected] (A. Winsler).1 This paper went through a standard blind review process with an associate

ditor (independent of the author/editor) with 5 external reviewers and 4 revisions.

ttp://dx.doi.org/10.1016/j.ecresq.2014.02.008885-2006/© 2014 Elsevier Inc. All rights reserved.

general population (National Clearinghouse for English LanguageAcquisition, 2010). Understanding and promoting positive earlydevelopment among the growing population of DLLs is critical forour nation’s future. Indeed, many scholars have urgently calledfor research and policy attention on dual language learners andtheir early educational needs (August & Shanahan, 2006; Garcia &Jensen, 2009; Hernandez, Denton, & McCartney, 2007; Takanishi,2010).

Research on DLLs is in its infancy for a number of reasons.First, there are fundamental definitional and assessment challengespresent in research with this heterogeneous population. Muchexisting research does not provide enough information about thechild participants or the language environment of the home toproperly determine whether children are being raised in the con-text of two or more languages. Also, DLL children are frequently

pment among dual language learners: The roles of languagemographic variables. Early Childhood Research Quarterly (2014),

excluded from studies because protocols often require childrento have sufficient proficiency in English (i.e., NICHD ECCRN stud-ies). Further, when DLL children are included in studies, they areoften not included in the analyses because assessments conducted

dx.doi.org/10.1016/j.ecresq.2014.02.008


http://www.sciencedirect.com/science/journal/08852006

mailto:[email protected]


ING ModelE

2 Resea

iteWeatdE

fifpclohEDM2rct2

sfadiiioeficmccmiab

owcctSmoeo

vcptnocdn

ARTICLEARCHI-697; No. of Pages 15

A. Winsler et al. / Early Childhood

n another language are not considered equivalent/comparable tohose in English (Chien et al., 2010; Espinosa & López, 2007; Howest al., 2008), or they are analyzed separately (Chang et al., 2007;insler et al., 2008). Additionally, lack of validated assessments

xisting in L1, and the lack of bilingual norms for interpretingssessment results have severely limited our understanding of theypical developmental trajectories of DLL children across multipleomains of development (Barrueco, López, Ong, & Lozano, 2011;spinosa & López, 2007; Francis & Rivera, 2007; Solórzano, 2008).

Also challenging are differences between DLL families and otheramilies on other factors known to predict early development. It ismportant to tease apart differences due to DLL status from otheractors such as ethnicity, country of origin, immigration status, andoverty status. Socio-economic differences are perhaps the largestonfound that need to be addressed. As a group, DLLs are moreikely to live in poverty and have less-educated parents: almost 70%f DLL students are considered low-income, and more than a thirdave parents with less than a high-school education (Matthews &wen, 2006). These sociocultural factors may, in part, explain whyLLs are less likely to perform well in school (Gandara, Rumberger,axwell-Jolly, & Callahan, 2003; U.S. Department of Education,

008). Similarly, DLL families are more likely to have immigratedecently and to retain beliefs and practices from their heritageountry, and thus it is important to distinguish DLL effects from cul-ural heritage effects (August & Shanahan, 2006; Garcia & Jensen,009; Hernandez et al., 2007).

The present study investigated variability in developmentalkills and gains over time, and potential moderators of outcomes,or DLLs from diverse backgrounds by conducting secondary datanalysis of the Early Childhood Longitudinal Survey-Birth Cohortata (ECLS-B; Najarian, Snow, Lennon, & Kinsey, 2010). The ECLS-B

s a nationally representative dataset of children born in the U.S.n 2001, and one of the few data sets that, although still limited,ncludes sufficient questions about home language use, countryf origin, immigration status, and other demographic variables toxamine specifically the development of DLL children in their firstve years of life. In this paper, we describe the extent to whichognitive, academic, and social development in early childhoodight vary as a function of home language use, examine heritage

ountry and maternal birth in the U.S. as moderator variables, andontrol for a variety of other demographic factors (child gender,arital status, parental education, family income). We also exam-

ne the extent to which relations between family demographicsnd child outcomes may differ for children with different languageackgrounds.

Recent research and theoretical models of minority child devel-pment emphasize the need to explore developmental processesithin distinct cultural groups, in order to understand which pro-

esses and outcomes are universal across different language andultural groups and which vary according to particular cultural con-ext (Garcia Coll et al., 1996; Quintana et al., 2006; Rogoff, 2003;hweder et al., 2006). Culture and language influence develop-ent, cognition, and behavior through socially constructed ways

f knowing and understanding the world (Rogoff, 2003; Shwedert al., 2006). There are a number of reasons to suspect that aspectsf child development might be different for DLL children.

The contexts of DLL children’s development expose them to aariety of vulnerability and protective factors. For instance, DLLhildren in the U.S. are more likely than other children to experienceoverty and to have immigrated, or have parents who immigratedo the U.S. Poverty, the immigration experience, and one’s eth-ic/cultural heritage are each associated with variance in family


pportunities, parenting values and practices, and, thus, child out-omes. Poverty is well known to negatively influence multipleomains of child development, both directly through biological,utritional, and health mechanisms, but also indirectly through

PRESSrch Quarterly xxx (2014) xxx–xxx

the effects that the stress of poverty has on undermining opti-mal parenting practices (Aber, Morris, & Raver, 2012; Duncan &Brooks-Gunn, 1997; McLoyd, 1990). On the other hand, immigra-tion status and fewer years of residency in the United States havealso been associated with some positive child outcomes (Bender &Castro, 2000; Garcia Coll & Marks, 2012). Below, we briefly discussunique features of immigrant families that may influence the earlydevelopment of DLLs as well as differences due to cultural heritage.

1.1. Immigrant families

There are well-documented cultural and linguistic differences inthe skills and abilities emphasized and reinforced by foreign-bornparents (Crosnoe, 2007; Fuligni, 1997; Perreira, Chapman, & Stein,2006; Portes, 1999; Suárez-Orozco & Suárez-Orozco, 2001). Immi-grant advantage refers to children of immigrant parents surpassingchildren of native U.S.-born parents in terms of developmental out-comes (Garcia Coll & Marks, 2012). Such differences are often seenas ‘paradoxical’ because strong outcomes for children of immi-grants are still observed in certain domains and certain age groupsdespite the personal, economic, and psychological hardships typi-cally involved with the experience of immigrating to a new country,and because with increased duration of time in the U.S. (eitherontogenetically or across generations), outcomes often decline forchildren of immigrant families (Garcia Coll & Marks, 2012). Evenas early as infancy, immigrant advantage is seen in health benefitsamong children of immigrant parents when compared to infantsof U.S.-born parents (Landale, Oropesa, Llanes, & Gorman, 1999).Personal and psychological strengths present among immigrantfamilies, such as increased motivation, optimism, and hope areoften considered potential mechanisms (Crosnoe, 2013).

However, immigrant “disadvantage” is also found especiallyamong those with less economic means, particularly with regardto cognitive skills. Glick, Bates, and Yabiku (2009), using ECLS-Bdata with U.S.-born children, show that the early cognitive out-comes of children of immigrant parents are poorer than those fromUS-born mothers and that this nativity effect is mediated by theparenting practices and cognitive stimulation at home and the ageat which the mother arrived to the U.S. Parenting, the home envi-ronment, and child cognitive outcomes are more similar to thoseof U.S.-born mothers when immigrant mothers arrive in the U.S.before the age of eight. By the entrance to kindergarten, childrenof immigrants show stronger social and emotional skills and fewerbehavior problems than U.S.-born children but still show poorercognitive and language skills (Crosnoe, 2007: De Feyter & Winsler,2009). Although main effects for immigrant advantage or disad-vantage are often found in child outcomes, some find such effectsto be moderated by parental education or country/region of ori-gin. Jackson, Kiernan, and McLanahan (2012) show that immigrantadvantage for health and behavioral problems is intensified amonglow-educated parents in the U.S. and that immigrant disadvantagefor internalizing behavior problems in the U.K. is stronger amongchildren of mothers with less education. Finally, and as discussedin the next section, outcomes for immigrant and DLL children canalso vary by heritage country of origin.

1.2. Heritage country of origin

The experiences and outcomes of DLL children also vary con-siderably as a function of ethnic heritage and/or country of origin.Children from immigrant families from Asian countries, for exam-ple, often fare better than those from Mexico during early childhood


in the U.S. partly due to advantages in family educational and eco-nomic resources, but also perhaps due to differences in teachingpractices in the home (Fortuny, Hernandez, & Chaudry, 2010; Han,Lee, & Waldfogel, 2012; Jung, Fuller, & Galindo, 2012). There are


ING ModelE

Resea

uwsmvietftwcOtn(e1

tbAaaECKcK&sipw(eftetoCm

ttUicwECi‘cet&pftttg1g



nique sociocultural, economic, and historical factors associatedith country of origin. Children’s educational outcomes, such as

chool readiness (De Feyter & Winsler, 2009) and verbal develop-ent (Leventhal, Xue, & Brooks-Gunn, 2006) have been found to

ary by heritage country of origin within Spanish-speaking Latinommigrants from various parts of Central and South America. Forxample, Dominican, Mexican, and Central American immigrantsypically have different immigration experiences, different reasonsor leaving their home country, and different patterns of transna-ional ties to their home country which can affect the extent tohich family involvement and support can serve as a buffer for

hildren (Araújo Dawson, Perez, & Suárez-Orozco, 2012; Suárez-rozco & Paez, 2002). Once immigrant families have settled in

he receiving country, the density of compatriot families of similarational heritage in the neighborhood in which the family settlesresiding in an “ethnic niche”) can influence the social, cultural, andconomic supports available to the child (Hao & Bonstead-Bruns,998; Portes, 1999).

Others have shown, both with ECLS-B and with other data sets,hat country of origin/heritage differences in parenting behavior,eliefs, and values lead to differential outcomes for young children.lthough country of origin differences for some child outcomesre typically reduced when socioeconomic factors and/or parentalcculturation (i.e., length of time in the US or degree of parentalnglish proficiency) is controlled (Barrueco, López, & Miles, 2007;abrera, Shannon, West, & Brooks-Gunn, 2006; Glick et al., 2009;eels, 2009), ethnic heritage differences in parenting and child out-omes often remain even with these controls (Barrueco et al., 2007;eels, 2009; Tamis-LeMonda, Song, Smith Leavell, Kahana-Kalman,

Yoshikawa, 2012). For example, Puerto Rican parents have beenhown to have affectively more positive interactions with theirnfants than Cuban parents (Field & Widmayer, 1981). Dominicanarents have been described as stricter than Puerto-Rican parents,ith Puerto Rican parents reporting receiving more social support

Wasserman, Brunelli, Rauh, & Alvarado, 1994). Tamis-LeMondat al. (2012) showed that African-American and Latina mothersrom Mexico and from the Dominican Republic each differed inheir use of early language and gesture with their infants, differ-nces that were associated with child language outcomes at agewo. Also, Mexican-Americans have poorer access to and qualityf health care in the U.S. compared to those with Puerto-Rican oruban roots, which can impact multiple domains of child develop-ent (Durden & Hummer, 2006).Diversity is found within Asian countries of origin as well, in

erms of parenting values and practices, child outcomes, and rela-ions between family capital and child outcomes (Han et al., 2012).sing ECLS-B data, Han et al. (2012) found that Chinese Americans

n particular showed strong academic performance in kindergartenompared to other groups (including other Asian groups), whichas related to greater parental education, income, health, parental

nglish proficiency and English use in the home found amonghinese immigrant families. Also, among specifically Chinese fam-

lies (and not other groups), the culturally specific constructs ofchiao shun’ (training, including both education and the adoption ofulturally appropriate behavior) and ‘guan’ (caring, loving, and gov-rning) relate stronger to children’s psychosocial outcomes thanhe construct of authoritative parenting (Chao, 1994, 2001; Lim

Lim, 2003). For Vietnamese immigrant children, early academicerformance is related to parental educational expectations, butor Chinese children, the same outcomes are more strongly relatedo SES (Shah, 2011). Korean American families show stronger cul-ural homogeneity and adherence to the norms and practices of


heir indigenous culture, and strong values for continued reli-ious education than other Asian American groups (Hurh & Kim,990; Min, 1992) which could potentially lead to differential lan-uage practices in the home and child outcomes. Further, Korean

PRESSrch Quarterly xxx (2014) xxx–xxx 3

cultural values that deemphasize individuality and self-expressionmay be related to the observation of less frequent use of pretendplay among children (Farver, Kim, & Lee, 1995).

As is clear from the above, it is important not to simply look atpan-ethnic categories such as Latino and Asian, but also to exam-ine different cultural heritages or countries of origin to understandthe experiences of DLL children. Toward that end, in the currentstudy, we examine families from Puerto Rico, Mexico, Cuba, Viet-nam, China, Japan, Korea, and India. Although the primary goal forthis study was to examine specifically whether DLL status, thatis, language use in the home, is uniquely associated with childoutcomes and gains (once family socioeconomic status, culturalheritage, parental education, and immigrant status are controlled),we also were interested in whether there were differences in theseassociations for different heritage groups. Finally, it is importantto acknowledge that, depending on the language use practices inthe home, not all children of immigrants are dual language learn-ers. The experience of learning two languages during the first fewyears of life may differentially influence DLLs in some ways relativeto monolingual children – the topic to which we now turn.

1.3. Dual language learners

Studies on bilingual language and literacy development amongyoung DLLs reveal that DLLs develop a distinct language systemfor each language, that the two language systems interact, and thatdevelopment and proficiency in the first language (L1) affect secondlanguage (L2) development and vice versa (Coppola, 2005; Fabiano-Smith & Barlow, 2010; Kuhl, 2001; Kupisch, 2007; Mishina-Mori,2005; Paradis, 2001). Also noted is that early language develop-ment for DLLs for both languages depends much on the quantityand quality of the language input they receive from parents andothers at home (Hammer, Lawrence, Rodriguez, Davison, & Miccio,2011; Scheele, Leseman, & Mayo, 2010). The age at which young DLLchildren reach certain language milestones in terms of vocabulary,syntax, and phonology in both of their languages is different thanthose found for monolingual speakers of each of their languages(Bland-Stewart & Fitzgerald, 2001; Fabiano-Smith & Goldstein,2010; Genesee, Paradis, & Crago, 2010; Paez, Tabors, & Lopez, 2007;Tabors, Paez, & Lopez, 2003). That is, children who are learning twolanguages during the preschool years, compared to monolingualchildren, seem to follow different developmental trajectories. Forexample, when receptive vocabulary in DLL children is assessed, thenumber of English words they ‘know’ is often considerably fewerthan those of monolingual English-speaking children of the samechronological age; but, if the number of Spanish words that thebilingual child knows is included along with the number of Englishwords, the total conceptual vocabulary score can be just as highor higher than that of monolingual children (Genesee & Nicoladis,1995; Oller & Jarmulowicz, 2007; Oller, Pearson, & Cobo-Lewis,2007). Thus, depending on the age and amount of early exposure,the expected developmental milestones for many language devel-opment skills for DLL children can be quite different from those ofmonolingual (English) children (Hoff et al., 2012). Further addingto the complexity is that language development can also be differ-ent for DLLs who learn their two languages simultaneously frombirth or early infancy as opposed to the sequential or successivebilingual who learns the second language after the first languagehas been established, usually after two-three years of age (Geneseeet al., 2010).

Given that language has far reaching transformative effects onother domains of development, such as cognition, behavior, and


social skills (Nelson, 2005; Vygotsky, 1986; Winsler, Fernyhough,& Montero, 2009) and in light of cultural/linguistic differencesin socialization (Shweder et al., 2006), it is likely that DLL chil-dren are different than monolingual English-speaking children in


ING ModelE

4 Resea

oebbc22aamsicerd

afodttDM&ca

cyhgdtdyIwsstddertica

2

2

iCCwgAtty



ther (non-language) domains as well. A small body of literature ismerging showing that young DLL children in immigrant families,efore school entry, may show stronger socio-emotional skills andetter behavior control compared to monolingual English-speakinghildren (De Feyter & Winsler, 2009; Galindo & Fuller, 2010; Han,010; Han & Huang, 2010; Luchtel, Hughes, Luze, Bruna, & Peterson,010). Fully bilingual children and adults, who have consider-ble experience inhibiting one language while speaking the othernd appropriately switching between languages, are known to beore advanced in inhibitory control and executive functioning

kills compared to monolingual children (Bialystok, 2009). Thesenhibitory control skills likely manifest themselves both in theognitive and social/behavioral domains. Thus, it is important toxamine potentially different developmental profiles and trajecto-ies for DLL children compared to monolingual children in otheromains of development as well.

Finally, DLL language status itself can be a source of consider-ble acculturative stress, which can affect developmental outcomesor children. Children in the early stages of English acquisitionften experience much pressure to speak English, stigma andiscrimination from their teachers and peers, and a variety ofensions related to language and cultural identity that cause poten-ially damaging stress (Araújo Dawson & Williams, 2008; Araújoawson, Williams, & Lopez-Humphreys, 2007; Padilla, Cervantes,aldonado, & Garcia, 1988; Smart & Smart, 1995; Suárez-Orozco

Suárez-Orozco, 2001). Such stress, experienced uniquely by DLLhildren could potentially lead to delays or differences in variousspects of child development.

The ECLS-B dataset is a large, nationally representative birth-ohort sample of US-born children who were followed for fiveears. The dataset includes information collected on children’some language use, ethnic heritage/country of origin, and immi-ration history of the family, as well as measures of children’sevelopment over time. Thus, it provides us with an opportunityo examine patterns of development and factors influencing thatevelopment across a wide variety of domains over the first fewears of life for a large sample of both DLL and monolingual children.n this paper, we examine a native-English-speaking group (those

ho speak only English at home) and two DLL groups (those whopeak only a non-English language at home, and those who speakome combination of English and a heritage language at home)o address the following questions: (1) Are skill levels and earlyevelopmental gains over time across multiple domains of childevelopment (cognitive, motor, language, social, behavior, math,arly literacy) similar for DLL and non-DLL children, controlling forelevant demographic variables? (2) Do child and family charac-eristics (gender, mother born in the US, heritage country of origin,ncome, marital status, parental education, maternal English profi-iency) show the same pattern of association with child outcomesmong young DLLs and non-DLLs?

. Method

.1. Participants

The ECLS-B is a nationally representative sample of approx-mately 14,000 births in the United States in the year 2001.hildren were sampled via registered births from the Nationalenter for Health Statistics vital statistics system. The childrenere recruited from diverse socioeconomic and racial/ethnic back-

rounds with oversamples of Asian and Pacific-Islander children,


merican Indian and Alaska Native children, Chinese children,wins, and low- and very-low-birth-weight children. The ECLS-Barget population was all children born in the United States in theear 2001 except children born to mothers less than 15 years of age,


and children who either died or were adopted before the 9-monthassessment. Baseline parent interviews and child assessments wereconducted when the child was approximately nine months old with10,700 children having complete 9-month data, and this group wasthen followed longitudinally with additional data collection tak-ing place when the child was approximately 24 months, four years(preschool wave), and 5/6 (kindergarten wave) of age. Additionaldetails on the full ECLS-B sample are available (National Centerfor Education Statistics, n.d.). As is required by ECLS-B data usageagreements, sample sizes are rounded to the nearest 50.

The sample included children from the following her-itages/countries: Mexico, Puerto Rico, Cuba (each with a reportablesample size 50 < n < 1050, resulting in a Hispanic/Latino pan-ethnicgroup of about n = 1200); India, China, Japan, Korea, and Vietnam(each 100 < n < 450 for a Pan-Asian group of about n = 1000). Theanalyses provided here included a total of 9250 children with dataon at least one child outcome, home language, heritage country,and selected demographic variables. The home language questionswere administered at 24 months, and attrition between 9 and 24months accounted for most of the missing data. The analysis sam-ple represents 86% of the original ECLS-B sample with 9-monthdata. Comparison between the subjects included and excluded inthe analysis sample indicated similarity among gender, socioeco-nomics, immigrant status, and languages spoken at home. Parentsfrom the analysis sample have slightly higher education, and areless fluent in English than parents not in the analysis sample.

Wave-specific sample weights are used in analyses that accountfor attrition and adjust back to the nationally representative sampleat each wave. The weights were selected according to the com-bination of components used in the models. We used child (i.e.,W1CO, W2CO, . . .) strata (i.e., W1RSTR, W2CSTR, . . .), and clus-ter (i.e., W1RRPSU, W2CPSU, . . .) weights supplied for each wavefor all analysis models. The weights were selected to account forthe stratified sampling design that involved oversampling in somestrata and to account for families that were recruited in the baselinewave, but did not participate in later waves of data collection. Thisaccounted for the original sampling strata by upweighting under-represented strata and for the loss to follow-up by upweightingdemographic groups that were disproportionately less likely to bein the follow-up samples. The weights do not, however, correctfor item-level missing data. This would occur when the ECLS-Bincludes some, but not all, data for a given child in a given wave.Overall, such missing data occurred in only 9.8% of the 9-monthdata, 2.3% of the 24-month data, 5% of the preschool data, and 4%of the kindergarten data. Descriptive tables are weighted.

2.2. Measures

Data were collected using direct assessments of the children,observations of parent-child interactions, and extensive interviewswith the primary caregiver, typically the mother, when the childrenwere 9 months old, 24 months old, a year prior to kindergarten, andin kindergarten. The parent interviews included questions aboutfamily demographic characteristics, home language use, and thechildren’s social skills. Parents/guardians were interviewed andasked to provide key information about their children and them-selves. The parent interview included two instruments: the parentinterview instrument and the parent self-administered question-naire (PSAQ). The first was conducted in person in by trained fieldinterviewers using computer-assisted personal interviewing (CAPI)as part of the home visit, and these interviews were conducted


in the parent’s home language when possible (especially true forSpanish or Chinese). The PSAQ was a paper and pencil instrument,presented during the parent CAPI instrument for the respondentto complete and return in a provided envelope, and contained 23


ING ModelE

Resea

q(

2

ddtvohblogHuwSaparw

2

oEFrwmwrtcorftgbgVawf

isbsptisiwi

2

tmE



uestions on topics some people might prefer to answer privatelyNord, Edwards, Andreassen, Green, & Wallner-Allen, 2006).

.2.1. Language group/DLL statusAccurately characterizing the home language environment of

iverse DLL families is extremely difficult since two, three, or moreifferent languages can be used in the home by different speakerso different family members. Ideally, in the absence of direct obser-ation, aggregates of multiple self-reported proportional measuresf language use (% of time each language is used) by everyone in theome when speaking with each other person in the home woulde available. The 24-month self-reported interview items about the

anguages spoken at home available in the ECLS-B, however, werenly sensitive enough to categorize families into three home lan-uage groups: (English only – EO n = 7750; heritage language only –O n = 450; or some combination of English and a heritage languagesed regularly at home – EH n = 1600). Both groups of children whoere not in the English-only language group are considered DLLs.

pecifically, in the 24-month parent interview, parents were firstsked, “What is the primary language spoken in your home?” Ifarents answered any language other than English, they were thensked, “Is English also spoken in your home?” Overall weightedesponse rates for the 24 month interview for the entire sampleas 93.1%

.2.2. Demographic covariatesIn addition, the parental interviews provided information about

ther relevant factors for DLL children - how well DLL parents spokenglish, the mother’s immigration status, and country of origin.or parents who spoke a language other than English, their self-ating of how well they spoke English (1 = very well, 4 = poorly)as used as a measure of parental English proficiency. In the 24-onth parent interview, mothers were asked, “In what countryere you born?” and could select “United States,” “U.S. territo-

ies,” or “some other country.” Mothers were coded as immigrants ifhey said no to the U.S. (parents who moved from Puerto Rico wereounted as immigrants for the purposes of this study). Regardlessf mothers’ birth place, all were also asked to identify their family’sace/country of origin by asking respondents to select one or morerom a long list of categories/countries on cards that best describedheir racial group heritage. The following country/territory of ori-in groups were included in this study from those items on theasis that there were more than 50 participants represented in eachroup: Mexico, Puerto Rico, Cuba, China, Japan, Korea, India, andiet Nam. For the purposes of analysis, dummy variables were cre-ted that compared each of these to the reference group of thoseho did not report a heritage country/territory of origin different

rom the U.S. [n = 7050].The interview also asked about demographic variables. We

ncluded information about family income, child gender, maritaltatus, and parental education. To determine level of education,oth parents were asked, “What is the highest grade or year ofchool that you have completed?” ranging from “no schooling com-leted = 0” to “5 or more years of college” = 17. We averaged acrosshe original ECLSB maternal and paternal education variables yield-ng a continuous variable representing average years of parentalchooling. This variable was represented by one parent’s educationf both parents did not report an educational level. Family income

as measured by the question, “What was your total householdncome last year, to the nearest thousand?” (Nord et al., 2006).

.2.3. Direct child assessments


The ECLS-B direct assessments included shortened versions ofhe Bayley Mental and Motor Development scales (cognitive and

otor T scores) at 9 months (weighted response rate for fullCLS-B sample = 95.6%) and 24 months (weighted response rate for


full ECLS-B sample = 94.2%) (West & Andreassen, 2002) developedthrough item response theory modeling. The shortened versionincludes 29 cognitive items at 9 months ( ̨ = .81) and 33 cogni-tive items at 24 months ( ̨ = .88), and 35 motor items at 9 months( ̨ = .94) and 32 motor items at 24 months ( ̨ = .73) (Andreassen &Fletcher, 2007; West & Andreassen, 2002). For the 9- and 24-monthassessments, Spanish, Chinese, or other languages assessmentswere conducted for those children who were not proficient enoughin English, according to a series of language-routing questions(Andreassen & Fletcher, 2007). A dummy variable indicated thelanguage of administration, with English as the reference cell.

Children’s academic achievement at preschool (weightedresponse rate for full ECLS-B sample = 98.3%) and kindergarten(weighted response rate for full ECLS-B sample = 98.6%) was mea-sured using direct assessments developed specifically for theECLS-B, comprised of items drawn from well-validated standard-ized instruments. The literacy measure included items used inthe ECLS-Kindergarten battery from the Woodcock Johnson Scalesof Academic Achievement-III and Peabody Individual AchievementTest, and added items from the Peabody Picture Vocabulary TestThird Edition, the PreLAS-2000, and the Preschool ComprehensiveTest of Phonological & Print Processing. The math scale includeditems from ECLS-K math assessments from the WJ and PIAT, withitems added from the Test of Early Mathematics Ability – III. Theearly literacy assessment ( ̨ = .92) consisted of 74 items that mea-sured early literacy and language skills, including letter knowledge,word recognition, print conventions, and phonological awareness.The math assessment ( ̨ = 92) consisted of 58 items focused onnumber sense, properties, operations, and probability. The presentsecondary data analyses utilized the IRT scores calculated by theECLS-B for these assessments. Literacy and math assessments wereconducted in either English or Spanish, but given that the numberof children who failed the language routing questions (and thusreceived assessments in Spanish) was too small in the last twowaves for IRT analyses, only data from the English-assessed chil-dren for the math and literacy measures are included in the two lastwaves of the ECLS-B (Najarian, Snow, Lennon, Kinsey, & Mulligan,2010).

2.2.4. Observations and parent ratingsPositive engagement with a parent was observed and rated dur-

ing parent-child interactions at the 2-year and the prekindergartenwaves of data collection (weighted response rate for full ECLS-Bsample = 93.1%). A parent, typically the mother, was asked to inter-act with her child during a semi-structured play situation based onthe two-bags procedure (National Center for Education Statistics,n.d.). Engagement with parent reflected the extent to which thechild shows, initiates, and maintains interactions with the parentand to which the child shows positive regard and affect towardthe parent. To be certified, coders had to demonstrate 85 percentagreement within 1 point with the master coders on all scales.Parents and children spoke in whatever languages they wanted toduring these observations and transcribers and coders who spokethe relevant language were used (Najarian, Snow, Lennon, Kinsey,& Mulligan, 2010).

Child social skills and behavior during the preschool years werealso measured using parental ratings (weighted response rate forfull ECLS-B sample = 91.3%). The parent rated the child using itemsfrom widely used rating scales such as the Social Skills Rating Scale(SSRS) and the Preschool and Kindergarten Behavior Scales (PKBS)at prekindergarten and kindergarten. Because the 26 items werenot scored in the ECLS-B dataset, we conducted a factor analysis.


The plot of the eigenvalues indicated an elbow at two factors, sotwo factors were retained. After dropping one item (worries) thatdid not load on either factor, the factor solution was the same atboth ages. The first factor ( ̨ = 80 at PK and .83 at K) consisted of


IN PRESSG ModelE

6 Research Quarterly xxx (2014) xxx–xxx

soatss

3

3

tpgjtfahgaiwf

aFseiacoR

TF

TF



ocial skill items such as “invites other to play,” “tries to understandthers,” “comforts other children,” and task orientation items suchs “pays attention” and “works independently.” The second fac-or ( ̨ = .76 at PK and .80 at K) consists of problem behavior itemsuch as “physically aggressive,” “annoys other children,” and “childeems unhappy.”

. Results

.1. Descriptive analyses

Before answering our research questions, we felt it importanto descriptively examine the diversity in family characteristicsresent in our sample by heritage country of origin and by lan-uage group. Given that the goal of these preliminary analyses wasust description and we wanted to minimize interpretation, andhe fact that, with such a large sample size, even very small dif-erences would be statistically significant, significance test resultsre not reported. As is shown in Table 1, there was considerableeterogeneity among families within the three home languageroups, and in Fig. 1 and Table 2, there was considerable variabilitymong heritage country-of-origin groups in family characteristics,ncluding languages spoken at home, maternal English proficiency,

hether the mother was born in the U.S., parental education, andamily income.

In terms of language use in the home, there was marked vari-bility in home language use across the different heritage groups.amilies with a heritage from Puerto Rico, Japan, India, and Koreahowed the lowest rates (2–11%) of speaking other languagesxclusively at home (‘other language only’), and most of these her-tage groups also showed the largest proportions using only English


t home (25–74%). A large proportion of families from all heritageountries (21–71%) reported that they speak some combinationf English and a heritage language at home. Mothers from Puertoican, Indian, Japanese and Korean heritages reported the strongest

able 1amily characteristics by home language groups (weighted).

Heritage language only (HO) (N = 400)

Parental education M (SE) 10.2 (0.3)Income (in 1000s) M (SE) 31.1 (1.9)

Maternal English proficiencya M (SE) 3.1 (0.1)

Mother U.S. born % 5

Country of originMexico % 64

Puerto Rico % <5

Cuba % <5

India % <5

China % <5

Japan % <5

Korea % <5

Vietnam % <5

Other % 23

a Lower numbers indicate greater English proficiency.

able 2amily characteristics by heritage country of origin (weighted).

Mexico Puerto Rico Cuba IndN 1050 100 50 25

Home languageHeritage only % 19 3 29 7English + heritage % 49 31 40 67English only % 32 66 30 26

Maternal English proficiencya M (SE) 2.2 (0.1) 1.2 (0.1) 1.6 (0.2) 1Mother U.S. born % 39 73 34 7Parental education M (SE) 10.5 (0.2) 12.7 (0.5) 14.0 (0.6) 16Income (in 1000s) M (SE) 31.0 (1.0) 47.0 (5.6) 73.3 (16.2) 77

a Lower numbers indicate greater English proficiency.

Fig. 1. Percentage of families who speak only English, only another language, orboth English and another language at home, by heritage country.

English proficiency while those from Vietnamese and Mexicanheritages were least proficient in English (note – lower numbersindicate greater English proficiency). Families with heritage tiesto India, China, and Vietnam rarely had mothers born in the U.S.(5–10%), whereas the majority of mothers with Japanese and PuertoRican heritage were born in the continental U.S. Parental educationwas lowest for those of Mexican origin and highest for those fromChina, India, Japan, and Korea. Family income followed a similarpattern and was greatest for those of Chinese, Indian, Japanese, andKorean origin, and lowest for families of Mexican and Puerto Ricandescent. In general, families from the Pan-Latino groups (the firstthree columns of the table) had lower levels of parental educationand income than those from the Pan-Asian groups.

Table 1 shows the family background variables as a function oflanguage used in the home, which is the focus of this investigation:those who speak only English at home (EO; n = 7300), those who


speak a combination of English and another language at home (EH;n = 1500), and those who speak exclusively a language other thanEnglish at home (HO; n = 400). Each variable varied by language

English + heritage language (EH) (N = 1500) English only (EO) (N = 7300)

11.6 (0.2) 14.1 (0.1)40.4 (1.8) 58.3 (1.4)

2.2 (0.0) –17 95

51 <10<5 <5<5 0<5 0<5 0<5 0<5 0<5 034 91

ia China Japan Korea Vietnam Other0 450 50 100 100 7050

19 8 13 17 1 50 12 30 66 6 31 80 58 17 93.5 (0.1) 1.8 (0.1) 1.4 (0.1) 1.4 (0.1) 2.1 (0.1) 1.1 (0.0)

14 69 32 4 90.5 (0.2) 16.4 (0.2) 15.3 (0.4) 16.0 (0.3) 12.7 (0.4) 14.1 (0.1).1 (3.8) 102.3 (6.8) 80.3 (9.2) 74.6 (6.5) 48.8 (5.1) 58.0 (1.5)


ARTICLE IN PRESSG ModelEARCHI-697; No. of Pages 15

A. Winsler et al. / Early Childhood Research Quarterly xxx (2014) xxx–xxx 7

Table 3Child outcomes by home language group (weighted).

Heritage languageonly (HO)

English + heritagelanguage (EH)

English only(EO)

Time 1 Time 2 Time 1 Time 2 Time 1 Time 2

Bayley mental (9 and 24 m) NM (SE)

40049.5 (0.5)

35044.1 (0.6)

150049.6 (0.4)

140045.6 (0.4)

730050.3 (0.3)

695051.2 (0.2)

Bayley motor (9 and 24 m) NM (SE)

40048.7 (0.7)

35048.0 (0.7)

150049.1 (0.4)

140048.9 (0.4)

725050.3 (0.3)

690050.3 (0.2)

Engagement (24 m and PK) NM (SE)

3003.9 (0.1)

2004.3 (0.1)

11004.1 (0.1)

9004.3 (0.0)

59004.7 (0.0)

57504.5 (0.0)

Literacy (PK and K) NM (SE)

25022.1 (1.0)

25032.8 (1.5)

110022.6 (0.5)

100037.2 (0.7)

620025.7 (0.3)

485039.3 (0.5)

Math (PK and K) NM (SE)

25027.8 (0.9)

25036.5 (0.8)

110028.0 (0.5)

100038.9 (0.5)

615029.6 (0.3)

485041.0 (0.3)

Social skills (PK and K) N 300 2500.0)

1200 1000 6300 4850

0.1)

ga

3

btmmfkaoturgsp

nrwutoTttsspsA

TC

M (SE) 3.7 (0.1) 3.7 (Problem behavior (PK and K) N

M (SE)3002.4 (0.0)

2502.4 (

roup with a consistent pattern of higher socio-demographic statusmong families who spoke more English at home.

.2. Children’s skills over time by home language group

Our first research question involved examining differencesetween DLL and non-DLL children in various skills. Table 3 showshe means (and SDs) at time 1 and time 2 for each of the child assess-

ents as a function of home language group. Note that for someeasures, T1 refers to 9 months with T2 being 24 months, whereas

or other assessments, TI refers to 24 months and T2 means pre-indergarten; still for others, T1 refers to the age-four pre-k periodnd T2 refers to the age-five kindergarten year. For cognitive devel-pment (Bayley scores), the English-only group made gains overime, whereas for the other groups (where other languages weresed in the home), Bayley scores went down slightly over time oremained the same. These descriptive results show that all groupsenerally made small gains on most measures, except for socialkills and problem behaviors from the pre-k to kindergarten timeeriods, where there was little change over time.

Given the longitudinal nature of the study, the last compo-ent of our initial descriptive analysis included here reports on theelative stability of the measurements across time points (T1–T2eighted correlations), and the inter-correlations between meas-res at the same time point. Table 4 shows these correlations, withhe diagonal having the stability coefficients and the upper rightf the table displaying the T2 correlations across the measures.he repeated assessments showed modest to high correlations overime. The literacy (r = .69) and math (r = .74) assessments from pre-ko kindergarten were the most stable. Parent ratings of child socialkills (r = .66) and problem behavior (r = 66) also were relatively


table. Infant cognitive (r = .26), toddler positive engagement witharent (r = .20), and infant motor (r = 17) measures were the leasttable. Correlations were generally very modest across domains.s expected, the two direct assessments were the most strongly

able 4orrelations across child outcomes at the second assessment, and T1-T2 stabilities (weigh

Bayley mental Bayley motor Engagement

Bayley mental .25 .43 .21

Bayley motor .14 .05

Engagement .20

Social skills

Problem behavior

Literacy

Math

a Values on the diagonal are T1–T2 correlations over time.

3.8 (0.0) 3.8 (0.0) 3.9 (0.0) 4.0 (0.0)12002.3 (0.0)

10002.3 (0.0)

63002.4 (0.0)

48502.3 (0.0)

related to each other, literacy and math (r = .82) and Bayley mentaland motor (r = .43). The parent ratings of social skills and problembehavior were modestly negatively correlated (r = −.34).

3.3. Hierarchical linear model analyses

Hierarchical linear model analyses were conducted to addressquestion 2; namely, to examine levels of, and change in, each ofthe developmental outcomes based on two repeated measures foreach assessment by language group. We accounted for correla-tions over time by estimating a random intercept for each child.The model included gender (female), marital status, parental edu-cation, and family income as covariates, and tested the extent towhich dummy-coded heritage country of origin (eight countrieswith no non-U.S. country given as the reference group), home lan-guage group (English only as the reference group), birthplace ofmother (born in any of the 50 states or DC), and maternal reportof English proficiency predicted both the intercept (adjusted finalstatus at T2) and slope (developmental gains) from T1 to T2. The ref-erence cell for home language was a family in which only Englishwas spoken at home (EO), so the two home language dummy vari-ables described differences between either children in which thefamily only spoke the heritage language (HO), or those in whichthe family spoke both English and the heritage language (EH), com-pared to the EO group. The dummy coding of the heritage countryof origin described differences between children of that heritagewith children whose parents did not indicate any of the eightnon-US heritage countries described above. We were able to esti-mate the association between maternal proficiency in English andoutcomes only for children whose families spoke the heritage lan-guage at home either solely or in combination with English by


taking advantage of the coding of the home language dummy vari-ables. Using the same approach economists often use to deal withmissing data, we assigned the mean score of English proficiencyto English-only families who were not asked this question and the

ted).a

Social skills Problem behavior Literacy Math

.28 −.16 .27 .34

.18 −.07 .06 .08

.14 −.08 .15 .20

.67 −.35 .21 .26.65 −.16 −.17

.67 .80.72


ING ModelE

8 Resea

mdadiwvtaEDt0cDtdatc

dgaeicgiga

asimnt(rnvbpBefsscctbb

3

osrowa(hl



odel computed the coefficient for English proficiency from theata for the DLL families only (NICHD ECCRN & Duncan, 2003). Allnalyses adjusted for sampling weights that accounted for missingata and attrition across time. Finally, analyses of the Bayley scores

ncluded a dummy variable that indicated whether the assessmentas conducted in English. Preliminary analyses involved dummy

ariables for Spanish, Chinese, and other languages, but estima-ion problems resulted due to the small cell sizes for the Chinesend other language groups. To remedy these problems, all non-nglish language groups were combined into a single group for theLL × heritage country interactions. All child outcomes and con-

inuous predictor variables were standardized to have a mean of and standard deviation of 1 for these analyses so the regressionoefficients could be interpreted as effect sizes (NICHD ECCRN &uncan, 2003). Thus, for categorical variables, coefficients describe

he mean difference between the groups in terms of standardeviation units in the dependent variable, and for continuous vari-bles, they describe the difference in standard deviation units inhe dependent variable associated with a one standard deviationhange in the predictor.

The goal of the analyses was to examine the extent to whichevelopmental outcomes were different as a function of languageroup, controlling for all of the other relevant demographic vari-bles. Also of importance was to determine whether DLL groupffects interacted with key background variables. Thus, the modelncluded language group (EO, HO, EH), a host of covariates that wereontrolled for, time, and interaction terms involving both languageroup and time and language group and covariates. We only exam-ned interactions between heritage country of origin and languageroup where there were a sufficient number of families who spoke

language other than English (Mexico, India, China, and Korea).The regression coefficients and standard errors from the models

re reported in Table 5. Consistent with our focus, we empha-ize below results for DLL-related predictors for each outcome andnteractions involving DLL status. The slope (‘time’) refers to how

uch children increased (positive numbers) or decreased (negativeumbers) from T1 to T2. Each of the other rows or estimates refero how much one has to increase or decrease either the interceptthe first two pages of the table underneath the shaded ‘intercept’ow), or the slope/time effect (the latter pages of the table under-eath the shaded ‘time/slope’ row) to represent the effect of thatariable on either final status or change over time. The interactionetween time and another variable refers to whether children’srogress over time was different across levels of that other variable.olded entries highlight statistically significant coefficients. Mainffects of each of the covariates on outcomes are listed in Table 5or completeness, but they are not discussed below since they areerving as control variables and because of the way the model waset up, they only refer to the English-only group. As seen from theoefficients in the table for the main effects of covariates on out-omes for English-only children, demographic factors were relatedo outcomes in largely expected directions (i.e., girls performingetter than boys on most outcomes, parental education and incomeeing positively associated with many outcomes).

.3.1. Home language effectsThe average skill levels of DLLs (those with either all or some

f another language occurring at home) compared to English-peaking children are shown in the 6th and 7th lightly shadedows of Table 5. These indicate how much children’s final statusn each outcome increased or decreased if their home languageas either exclusively a different language (other vs. English) or


combination of languages including English was used at homemultiple vs. English). Overall, only a few main effects appeared forome language. Whether the family spoke English and/or another

anguage at home was related to the direct assessment of mental


development at 24 months, observed engagement in interac-tions with mother at the preschool visit, and parent report ofproblem social skills and behavior at kindergarten. English-onlychildren scored higher on the developmental assessment than bothgroups of DLL children (−.49 < B < −.41). EO only children werealso observed to show more engagement with mothers (B = 20) atthe preschool visit and were reported to have better social skills(B = .19) at entry to kindergarten than children who spoke bothEnglish and a home language at home, and had more behavior prob-lems (B = .63) than children who only spoke the heritage languageat home.

3.3.2. Developmental gains for DLL and non-DLL childrenThe interactions between time and home language status (the

two lightly shaded rows below ‘time’ in Table 5) compare the devel-opmental gains for DLL to non-DLL children. DLL groups showeddifferent patterns of change over time on mental development andearly literacy. The children who spoke both English and a her-itage language in the home showed fewer gains between 9 and24 months on the infant cognitive measure (B = −.45) comparedto English-only children (B = .03). The children who only spoke theheritage language in the home showed larger gains in early literacy(B = 1.66) over time than did English-only children (B = .77). Thus,there is some evidence that developmental trajectories for DLLs inthe areas of cognition and early literacy are somewhat differentthan those for monolingual English children, after controlling forall other demographic variables in the model.

3.3.3. Moderators involving DLL statusOur second research question asked about whether demo-

graphic variables were related to child outcomes in the sameway for the different language groups. Interactions between childand family characteristics (i.e., child gender, mother born in U.S.,heritage country of origin, income, parental education, maternalEnglish proficiency, marital status) and DLL status were investi-gated to examine moderators of children’s outcomes and gains overtime. There were many of these, suggesting a complicated picturewith demographic variables appearing to operate differently fordifferent home language and immigrant groups, and some effectsof home language on child outcomes differing as a function of othervariables.

Parental education appeared to operate differently by languagegroup, and the direction of effects varied across type of outcome.Compared to children whose families spoke only English at home(EO), parental education was a weaker predictor of cognitive skillsat 24 months for children whose families spoke both English andHeritage languages (EH), and children whose families spoke theHeritage language only (HO), (EO B = .22; HO B = .04; EH B = −.01).The same pattern was found for engagement with mother at 52months (EO B = .14; HO B = −.02) and literacy at kindergarten entry(EO B = .28; HO B = .10; HE B = .17). In contrast, parental educationwas a stronger positive predictor of problem behaviors amongchildren in the Heritage-language-only group compared to theEnglish-only group (EO B = −.06; HO B = .14). Parental educationpredicted rate of change over time differently for early cogni-tive skills depending on language group. The association betweenparental education and change over time in cognitive skills wasmore positive for the English-only infants (B = .17) than for theEnglish and Heritage language infants (B = −.08).

There were complex interactions for child outcomes betweenimmigrant status (mother being born in the U.S. or not) andlanguage use in the home. All findings are reported below, but


it should be noted that because there were very few mothers inthe Heritage-language-only group who were U.S. born (n < 50),comparisons of native vs. immigrant mothers for this group shouldbe interpreted cautiously. Fig. 2 shows the unadjusted least-square


Please cite this article in press as: Winsler, A., et al. Early development among dual language learners: The roles of languageuse at home, maternal immigration, country of origin, and socio-demographic variables. Early Childhood Research Quarterly (2014),http://dx.doi.org/10.1016/j.ecresq.2014.02.008


A. Winsler et al. / Early Childhood Research Quarterly xxx (2014) xxx–xxx 9

Table 5Regression coefficients: HLM analyses of child outcomes.

Bayley mental(9 m, 24 m)

Bayley motor(9 m, 24 m)

Engagementw/parent(24 m, 52 m)

Social skills(52 m, 65 m)

Problembehavior(52 m, 65 m)

Early literacy(52 m, 65 m)

Math(52 m, 65 m)

B (SE) B (SE) B (SE) B (SE) B (SE) B (SE) B (SE)

Intercept (F) 0.15 (0.02)*** 0.10 (0.02)*** 0.02 (0.02) 0.17 (0.02)*** −0.13 (0.02)*** 0.47 (0.02)*** 0.51 (0.02)***

Gender (Girl) 0.36 (0.02)*** 0.18 (0.02)*** 0.13 (0.02)*** 0.36 (0.02)*** −0.42 (0.02)*** 0.19 (0.02)*** 0.10 (0.02)***

Parental education 0.22 (0.02)*** 0.06 (0.02)** 0.14 (0.02)*** 0.11 (0.02)*** −0.06 (0.02)** 0.28 (0.02)*** 0.27 (0.02)***

Income/1000 0.10 (0.01)*** 0.03 (0.01)* 0.06 (0.02)*** 0.06 (0.02)** −0.06 (0.02)*** 0.15 (0.01)*** 0.15 (0.01)***

Home language (F) 21.08*** 1.05 3.70* 4.07* 3.97* 1.72 1.88Home language: other vs.

English−0.49 (0.21)* 0.09 (0.22) −0.16 (0.20) 0.17 (0.22) −0.63 (0.22)** 0.35 (0.19) 0.31 (0.18)

Home language: multiple vs.English

−0.41 (0.07)*** −0.09 (0.07) −0.20 (0.07)** −0.19 (0.07)** −0.02 (0.07) 0.01 (0.06) −0.05 (0.06)

Mom English proficiency −0.01 (0.01) −0.00 (0.01) −0.02 (0.01) −0.02 (0.01) 0.02 (0.01) 0.01 (0.01) 0.02 (0.01)+

Mother born US 0.23 (0.05)*** 0.04 (0.05) 0.08 (0.06) 0.02 (0.06) 0.15 (0.06)* 0.10 (0.05)+ 0.18 (0.05)***

Marital status (married) 0.11 (0.03)*** 0.01 (0.03) 0.01 (0.03) −0.03 (0.03) −0.17 (0.03)** 0.04 (0.02) 0.02 (0.02)Assessment language not

English−0.07 (0.11) 0.28 (0.12)*

Heritage (F) 2.91** 1.88+ 1.79+ 2.87** 3.61*** 6.42*** 4.02***

Mexico −0.09 (0.04)* −0.05 (0.04) 0.11 (0.04) 0.04 (0.04) 0.06 (0.04) 0.02 (0.04) −0.04 (0.04)Puerto Rico −0.11 (0.08) −0.29 (0.09) 0.03 (0.10) 0.22 (0.09)* 0.31 (0.10)** 0.03 (0.08) −0.04 (0.08)Cuba 0.19 (0.14) −0.06 (0.14) 0.02 (0.16) −0.12 (0.09) −0.33 (0.18)+ 0.03 (0.15) 0.14 (0.15)India 0.09 (0.12) −0.18 (0.12) −0.08 (0.15) 0.21 (0.13) −0.47 (0.14)*** 0.56 (0.11)*** 0.21 (0.11)+

China 0.41 (0.12)*** −0.07 (0.12) −0.42 (0.16) −0.45 (0.14)** −0.06 (0.14) 0.60 (0.12)*** 0.50 (0.11)***

Japan −0.13 (0.16) −0.04 (0.17) 0.04 (0.20) −0.14 (0.19) −0.03 (0.19) −0.15 (0.16) −0.19 (0.15)Korea 0.03 (0.17) −0.19 (0.18) −0.23 (0.27) 0.03 (0.22) −0.31 (0.22) 0.16 (0.19) 0.25 (0.18)Vietnam 0.01 (0.15) −0.11 (0.16) 0.04 (0.25) −0.23 (0.18) 0.01 (0.19) 0.30 (0.15)+ 0.31 (0.15)*

Girl × home language (F) 7.42** 10.09*** 2.96+ 3.44* 11.15*** 6.06** 5.70**

Girl × other vs. English −0.28 (0.09)** −0.43 (0.10)*** 0.09 (0.12) 0.03 (0.12) −0.15 (0.12) 0.04 (0.10) −0.07 (0.10)Girl × multiple vs. English −0.15 (0.05)** −0.10 (0.06)+ −0.15 (0.07) −0.16 (0.06)** 0.28 (0.06)*** −0.18 (0.05)*** −0.17 (0.05)***

Parental education × homelanguage (F)

28.98*** 0.09 4.55* 0.03 7.20*** 11.33*** 14.17***

Parental education × other vs.English

−0.16 (0.04)** −0.02 (0.05) −0.16 (0.06)** 0.01 (0.05) 0.20 (0.05)*** −0.18 (0.04)*** −0.21 (0.04)***

Parental education × multiplevs. English

−0.23 (0.03)*** −0.01 (0.03) −0.06 (0.04) −0.01 (0.04) 0.06 (0.04) −0.11 (0.03)*** −0.11 (0.03)***

Income × home language (F) 0.72 1.62 1.71 2.59+ 2.47+ 7.52*** 1.09Income × other vs. English −0.11 (0.11) 0.21 (0.11) 0.21 (0.12) 0.04 (0.12) −0.22 (0.12) + 0.38 (0.10)*** 0.14 (0.10)Income × multiple vs. English −0.03 (0.04) 0.01 (0.04) 0.04 (0.05) −0.10 (0.05) 0.06 (0.05) −0.03 (0.04) −0.02 (0.04)

Mom born US × home language(F)

4.88** 0.88 1.21 3.15* 5.60** 2.24 8.34***

Mom born US × other vs.English

−0.52 (0.21)* −0.30 (0.23) −0.09 (0.27) 0.59 (0.30)+ −1.03 (0.31)*** 0.45 (0.25) 0.51 (0.25)*

Mom born US × multiple vs.English

−0.21 (0.09)* −0.04 (0.09) −0.16 (0.10) −0.13 (0.10) −0.05 (0.11) −0.08 (0.09) −0.28 (0.09)**

Marital status × home language(F)

1.61 2.27 0.23 1.16 2.28 1.58 1.68

Marital status × other vs.English

0.18 (0.10) 0.22 (0.11) −0.00 (0.13) −0.01 (0.12) 0.26 (0.12) 0.17 (0.11) 0.17 (0.10)

Marital status × multiple vs.English

0.03 (0.06) 0.07 (0.07) −0.05 (0.07) 0.10 (0.07) 0.05 (0.07) 0.05 (0.06) 0.05 (0.06)

Assessment language – nonEnglish × home language (F)

1.39 4.05*

Assessment language – nonEnglish × other vs. English

−0.07 (0.19) −0.57 (0.20)**

Assessment language – nonEnglish × multiple vs. English

0.15 (0.13) −0.18 (0.14)

Time 0.03 (0.02) 0.00 (0.02) −0.14 (0.02)*** 0.09 (0.02)*** −0.16 (0.03)*** 0.88 (0.02)*** 0.91 (0.02)***

Time × girl 0.23 (0.03)*** 0.14 (0.03)*** −0.13 (0.04)*** 0.02 (0.03) −0.05 (0.03) 0.07 (0.03)* −0.01 (0.03)Time × parental education 0.17 (0.02)*** 0.06 (0.02)* −0.02 (0.03) −0.01 (0.03) 0.08 (0.03)* 0.03 (0.02) −0.03 (0.02)Time × income 0.06 (0.02)*** 0.07 (0.02)*** −0.02 (0.02) −0.01 (0.02) −0.00 (0.02) 0.07 (0.02)*** 0.04 (0.02)*

Time × home language (F) 13.33*** 1.62 0.15 1.52 2.92+ 5.17*** 1.02Time × other vs. English −0.52 (0.31)+ 0.15 (0.33) 0.14 (0.32) 0.14 (0.28) −0.69 (0.29) 0.78 (0.25)** 0.34 (0.24)Time × multiple vs. English −0.48 (0.10)*** −0.17 (0.10) 0.04 (0.11) −0.16 (0.10) 0.02 (0.10) 0.10 (0.09) −0.00 (0.08)Time × maternal English

proficiency0.00 (0.02) −0.00 (0.02) 0.04 (0.02)+ −0.04 (0.02)* 0.02 (0.02) 0.00 (0.02) 0.00 (0.02)

Time × mother born in US 0.14 (0.07)* −0.08 (0.08) −0.07 (0.08) 0.02 (0.08) 0.15 (0.08)+ 0.11 (0.07) 0.18 (0.07)**

Time × marital status 0.08 (0.04)* 0.18 (0.04)*** −0.15 (0.04)*** −0.01 (0.04) −0.13 (0.04)** 0.01 (0.03) 0.01 (0.03)Time × Assessment language,

non-English−0.06 (0.17) 0.08 (0.18)



10 A. Winsler et al. / Early Childhood Research Quarterly xxx (2014) xxx–xxx

Table 5 (Continued)

Bayley mental(9 m, 24 m)

Bayley motor(9 m, 24 m)

Engagementw/parent(24 m, 52 m)

Social skills(52 m, 65 m)

Problembehavior(52 m, 65 m)

Early literacy(52 m, 65 m)

Math(52 m, 65 m)

B (SE) B (SE) B (SE) B (SE) B (SE) B (SE) B (SE)

Time × heritage (F) 3.24*** 2.86** 1.09 0.38 0.73 1.30 0.64Time × Mexico −0.13 (0.05)** 0.12 (0.05)* 0.08 (0.06) 0.01 (0.06) −0.01 (0.06) 0.10 (0.05) 0.07 (0.05)Time × Puerto Rico −0.21 (0.12)+ −0.36 (0.12)** −0.02 (0.14) 0.10 (0.13) 0.14 (0.14) 0.02 (0.11) −0.01 (0.11)Time × Cuba 0.33 (0.20)+ 0.43 (0.21)* 0.15 (0.22) −0.27 (0258) −0.35 (0.24) 0.20 (0.20) 0.31 (0.20)Time × India −0.01 (0.17) −0.24 (0.18) 0.21 (0.21) 0.10 (0.19) −0.19 (0.19) 0.36 (0.16) 0.17 (0.16)Time × China 0.61 (0.17)*** −0.05 (0.18) −0.30 (0.21) 0.01 (0.19) 0.00 (0.20) 0.20 (0.17) 0.08 (0.16)Time × Japan 0.13 (0.23) −0.06 (0.25) 0.45 (0.28) 0.22 (0.26) −0.07 (0.27) −0.05 (0.22) 0.04 (0.22)Time × Korea 0.07 (0.04) −0.33 (0.25) −0.30 (0.34) 0.12 (0.30) −0.29 (0.31) 0.13 (0.26) 0.03 (0.25)Time × Vietnam 0.12 (0.23) 0.02 (0.24) 0.13 (0.31) −0.05 (0.25) 0.23 (0.26) 0.13 (0.22) 0.03 (0.21)

Time × girl × home language (F) 0.17 2.38+ 4.59* 0.61 0.15 1.32 0.63Time × other vs. English × Girl −0.07 (0.14) −0.29 (0.15)+ 0.51 (0.17)*** 0.05 (0.16) −0.03 (0.17) 0.18 (0.15) 0.00 (0.15)Time × multiple vs.

English × Girl−0.03 (0.08) 0.06 (0.08) −0.03 (0.10) 0.09 (0.09) 0.05 (0.09) −0.08 (0.08) −0.08 (0.07)

Time × parentaleducation × home language(F)

15.23*** 1.98 0.36 0.47 1.16 0.25 1.69

Time × other vs. English × P. Ed −0.11 (0.06)+ −0.13 (0.07) −0.06 (0.08) −0.07 (0.07) 0.01 (0.07) 0.01 (0.07) 0.03 (0.06)Time × multiple vs.

English × parental education−0.25 (0.05)*** −0.03 (0.05) −0.03 (0.05) −0.00 (0.05) −0.07 (0.05) 0.03 (0.04) 0.08 (0.04)

Time × income × home language(F)

1.63 0.88 0.14 2.25 1.11 2.11 1.25

Time × other vs.English × income

−0.17 (0.15) 0.03 (0.16) 0.10 (0.19) 0.14 (0.16) 0.01 (0.17) 0.13 (0.14) 0.01 (0.14)

Time × multiple vs.English × income

0.08 (0.06) 0.09 (0.06) 0.00 (0.08) −0.13 (0.07) 0.10 (0.07) −0.11 (0.06) −0.09 (0.06)

Time × mom born US × homelanguage (F)

1.66 0.47 0.53 0.98 4.82** 4.57* 2.76+

Time × other vs.English × mother born in US

−0.40 (0.31) −0.24 (0.33) 0.04 (0.42) 0.24 (0.39) −1.23 (0.40)** 0.98 (0.35)** 0.52 (0.33)

Time × multiple vs.English × mother born in US

−0.19 (0.13) 0.07 (0.14) −0.15 (0.15) −0.16 (0.14) −0.14 (0.15)* −0.07 (0.12) −0.17 (0.12)

Time × marital status × homelanguage (F)

0.02 1.16 1.96 0.42 0.53 1.43 0.62

Time × marital status × othervs. English

0.03 (0.15) −0.06 (0.16) −0.35 (0.18) 0.14 (0.17) 0.17 (0.17) 0.28 (0.17) 0.11 (0.16)

Time × maritalstatus × multiple vs. English

0.01 (0.09) −0.14 (0.10) −0.06 (0.11) 0.05 (0.10) 0.04 (0.10) −0.02 (0.09) −0.07 (0.08)

Time × assessment language,non-English × home language(F)

0.67 5.98**

Time × assessment language,non-English × other vs.English

−0.16 (0.28) −0.77 (0.29)**

Time × assessment language,non-English × multiple vs.English

0.10 (0.19) 0.08 (0.20)

Additional interactionsMexico × English vs. other 0.12 (0.07)+ 0.04 (0.08) −0.06 (0.09) 0.10 (0.08) −0.20 (0.08)* 0.04 (0.07) 0.05 (0.07)Time × Mexico × English vs.

other−0.02 (0.10) −0.20 (0.11)+ −0.26 (0.12)** 0.08 (0.11) −0.21 (0.12)+ −0.01 (0.10) 0.01 (0.10)

India × English vs. other −0.09 (0.23) −0.16 (0.25) 0.23 (0.30) 0.41 (0.27) −0.28 (0.27) −0.09 (0.23) 0.01 (0.22)Time × India × English vs. other −0.31 (0.34) −0.05 (0.36) 0.22 (0.41) −0.11 (0.38) 0.14 (0.39) −0.07 (0.32) 0.16 (0.31)China × English vs. other −0.37 (0.23) 0.10 (0.25) 0.04 (0.32) 0.14 (0.28) 0.29 (0.28) −0.26 (0.23) −0.48 (0.23)*

Time × China × English vs.other

−0.30 (0.34) −0.52 (0.36) −0.21 (0.43) 0.06 (0.39) −0.01 (0.40) −0.00 (0.33) −0.12 (0.32)

Korea × English vs. other −0.26 (0.33) −0.17 (0.35) 0.32 (0.55) 0.27 (0.43) 0.38 (0.45) 0.12 (0.37) −0.19 (0.36)Time × Korea × English vs.

other−0.25 (0.48) 0.02 (0.51) 0.43 (0.68) 0.04 (0.60) 0.04 (0.62) −0.08 (0.51) −0.12 (0.50)

+ p < .10.

miHpotbc

* p < .05.** p < .01.

*** p < .001.

eans on the Bayley cognitive measure for immigrant and non-mmigrant families as a function of language used in the home.aving a mother born outside of the U.S. was associated withoorer cognitive skills for infants and toddlers (and lower rates


f growth), but this immigrant disadvantage effect was especiallyrue when only English was spoken in the home. This differenceetween children in immigrant and non-immigrant homes onognitive skills was smaller when a combination of English and the

heritage language was spoken at home. When only the heritagelanguage was spoken at home, children in immigrant familiesoutperformed those from native US-born families.

There were somewhat similar interactions among immigration


and DLL group on preschool academic outcomes. Children scoredhigher on math among the English-only (B = .18) and the heritage-only (B = .69) groups if their mother was native born, but the Englishand heritage language group scored higher if their mother was an


ARTICLE ING ModelEARCHI-697; No. of Pages 15

A. Winsler et al. / Early Childhood Resea

Fo

iebreol

iliiflepEfcoUtdiat

ghmalmghslf

olpggEfge(

ig. 2. Bayley cognitive scores for immigrant and native-born families as a functionf language used at home.

mmigrant (B = −.10). Children showed larger gains over time in lit-racy among the Heritage-only group if their mothers were nativeorn, gains over time in the other two language groups were notelated to whether the mother was born in the U.S. This later differ-nce was difficult to interpret because it was driven by comparisonsf a very small group - the group who only spoke the heritageanguage at home when the mother was U.S. born.

For children’s behavior problems, significant interactionsnvolving DLL status and immigration status emerged for bothevels of child behavior problem at kindergarten entry and formprovement over time in children’s behavior. For overall behav-or problems at kindergarten entry, the levels of problem behavioror children of immigrant families were the same across the threeanguage groups. However, for children of native US-born moth-rs, fewer behavior problems were reported by the small group ofarents who only used the heritage language at home. Also, for thenglish-only and English and heritage language groups, childrenrom immigrant families showed fewer behavior problems thanhildren of U.S.-born mothers. The interaction involving changever time in behavior problems was also driven by the small cell of.S.-born mothers speaking only the heritage language at home –

hey reported the highest level of problem behavior in their chil-ren at 52 months and this group showed the greatest reduction

n behavior problems over time by kindergarten entry, althoughll groups reported some improvement in child behavior overime.

In summary (and downplaying the findings involving the smallroup of U.S.-born heritage-language only group), compared toaving a mother born in the U.S., having an immigrant mother wasore negatively related to early cognitive and preschool math skills

mong the English-only group than among the English and heritageanguage group. Similarly, for math outcomes, having an immigrant

other was related to lower math skill among the English-onlyroup but not among the group that used both languages in theome. Thus, for early cognitive development and math skills uponchool entry, there is some evidence here that use of the heritageanguage at home was a protective factor for children of immigrantamilies.

Gender differences varied depending on what language wasccurring in the home. Gender differences favoring girls on the Bay-ey mental and motor scores, ratings of social skills and behaviorroblems, and assessments of literacy and math at entry to kinder-arten were larger among the English-only group than amongroups with other languages spoken in the home (‘Girl × Other vs.nglish,’ and ‘Girl × Multiple vs. English’ rows in Table 5). The dif-


erence between girls and boys was significantly different by DLLroup for early cognitive skills (EO B = .36; HO B = .08; EH B = .21),arly motor skills (EH B = .18; HO B = −.25; EH B = .08), social skillsEO B = .36; EH B = .20), and problem behaviors (EO B = −.42; EH


B = −.14), literacy (EO B = .19; EH B = .01), and math (EO B = .10; EHB = −.07).

Also visible in Table 5 is that language of assessment (Heritagelanguage vs. English) for the 9- and 24-month Bayley motor per-formance assessments interacted with DLL status. Children testedin their heritage language scored higher overall in motor skills,especially if the heritage language was spoken exclusively at home.When the heritage language was spoken exclusively at home andchildren were tested in the heritage language, they showed higherlevels of motor skills at 9 and 24 months compared to other chil-dren, and those differences were larger at 9 months than at 24months.

Finally, we were interested in asking whether home languagepredicted outcomes differently depending on heritage country. Thetwo DLL groups were combined in these interaction terms due tosmall cell sizes in some groups. Nevertheless, not all interactionscould be computed because there were still too few families in somecells (e.g., very few families who reported Japan as their countryof origin also reported speaking Japanese at home, and very fewfamilies who reported Puerto Rico as their place of origin reportedspeaking English only at home). There was sufficient variation inhome language among individuals from Mexico, India, China, andKorea to examine this issue. Results are shown on the final rowsof Table 5. Few significant differences emerged. For families witha Mexican heritage, the finding that DLL children were more posi-tively engaged in interactions with a parent, and had fewer problembehaviors compared to children who spoke English at home, wasmore true for families that came from Mexico than for familiesfrom other heritage countries. Also, the difference in math scoresfavoring DLL families compared to English-only families was muchstronger among children of Chinese heritage than children fromother countries.

4. Discussion

These secondary analyses of ECLS-B data were designed to addto the limited information available about the early developmentof children who are dual language learners (DLLs). Our findingsillustrate large differences among DLL children in terms of demo-graphic characteristics and the extent to which the home languageis spoken depending on country of heritage. For example, familiesof Puerto Rican, Japanese, Indian, and Korean descent used the mostEnglish at home, on average, but there was clearly variance withineach heritage country group in terms of family characteristics aswell. Mothers with Puerto Rican, Cuban, and Indian heritages hadthe strongest English proficiency while those of Vietnamese andMexican descent were the least proficient in English. The majorityof mothers from Japanese and Puerto Rican heritages were born inthe continental U.S. but families originally from India, China, andVietnam rarely had mothers who were born in the U.S. Parentaleducation and family income were relatively low for those of Mex-ican origin and highest for those with Chinese, Indian, Japanese,and Korean heritage. These results replicate prior work of othersshowing that immigrant and language-minority families are a het-erogeneous group and further examination of subgroups is needed(De Feyter & Winsler, 2009; Hernandez, Denton, & Blanchard,2011). These findings suggest that simple global categories suchas ‘immigrant’ or ‘DLL’ families may obscure important variabilitywithin these groups in children’s outcomes as well as in under-standing processes and predictors of outcomes.

Our descriptive analysis of variations in family demographics


as a function of language use in the home replicated prior demo-graphic work showing broad family social and economic capitaladvantages as English language use and parental English languageproficiency in the home increase (Hernandez et al., 2011). Children


ING ModelE

1 Resea

wglatansmwddpa

dhgamgEaanhamiwfimdmns

iofiaatmeamo

iig22efidcwhswlt


2 A. Winsler et al. / Early Childhood

ho used only English at home had higher parental education,reater family income, greater English proficiency, and were moreikely to have the mother born in the U.S., whereas those speaking

combination of English and another language at home were inhe middle range, and those who spoke only a heritage languaget home were at greater socioeconomic risk. It was interesting toote, however, that while parental education and/or income wereignificant factors for every child outcome measured, significantain effects for DLL group on child outcomes and change over timeere only present for fewer than half of the child outcomes. Familyemographic factors, thus, appear to play a larger role in early childevelopment than does home language use, and that DLL statuser se, once other demographic variables are controlled, does notppear to have large and broad effects on child development.

That being said, it is still notable that after controlling for manyemographic factors, as well as immigrant status and country oferitage, there were a few significant effects attributable to lan-uage use in the home on both cognitive and socio-emotionalspects of early child development. For early cognitive develop-ent at 9 months, DLL children scored lower (and showed slower

rains between 9 and 24 months) than children speaking onlynglish at home, regardless of the language of assessment, but over-ll main effects for language group were not observed on the mathnd literacy assessments later, at the end of preschool. Also, it wasoted that children exposed exclusively to the heritage language atome made particularly strong gains in preschool on (English) liter-cy relative to the other groups. DLLs performed better on the earlyotor assessment at 24 months if they were tested in their her-

tage language as opposed to in English, and this was especially truehen the heritage language was spoken exclusively at home. Thisnding points to the importance of matching language of assess-ent with that used at home with the child. There were no reliable

ifferences, however, among the DLL groups on motor skills at 24onths or early literacy and math skills at kindergarten entry, and

o differences in change over time from 9 to 24 months in motorkills or from preschool to kindergarten entry in math.

There were several complex patterns of interaction, indicat-ng that gender, parental education, and immigration status mayperate differently across different language groups for at least aew outcomes. This suggests that attention to home language ismportant for understanding early development, both in terms ofpparent differences related to learning more than one language,nd in terms of potential differences in how various family fac-ors predict early development. Future exploration of the context of

ultiple language use in the home is needed, with studies that canxplore in more detail the relevant processes and mediators thatre at play. The goal of the present study was simply first to deter-ine the extent to which there might be broad brush differences

n a variety of child outcomes as a function of DLL status.Home language use and immigrant status interacted in interest-

ng ways to predict some child outcomes. Other work has shown anmmigrant disadvantage in early cognitive outcomes but an immi-rant advantage for some socio-emotional outcomes (Crosnoe,007; De Feyter & Winsler, 2009; Magnuson, Lahaie, & Waldfogel,006). In the present study, a pattern emerged showing that differ-nces between children in immigrant and non-immigrant familiesor a few outcomes may depend somewhat on the language(s) usedn the home. Controlling for heritage country of origin and familyemographics such as income and education, more positive earlyognitive outcomes for children in immigrant families appearedhen some amount of the heritage language was occurring in theome compared to only English. As displayed in Fig. 2, if English was


poken exclusively in the home, being from an immigrant familyas linked with poorer early cognitive outcomes, but when other

anguages were spoken in the home, children with U.S.-born andhose with foreign-born mothers were performing more similarly,


even when controlling for other variables in the models. Anotherway of thinking about this pattern is that for children with U.S.-born parents, the more English used in the home, the better itwas for early cognitive outcomes, however, for immigrant fami-lies, language use in the home was not as critical for performance.A similar pattern was found for preschool math outcomes. Thus,for early cognitive development and math skills upon school entry,this study provides some preliminary evidence that use of the her-itage language at home may be a protective factor for children ofimmigrant families. It is important to qualify, however, that theseresults are quite preliminary and in need of replication given thatthis pattern was not found for other outcomes and when the patternwas present, it involved some groups with relatively small samplesizes.

Potential explanations for these findings are that they reflectrecency of migration, degree of acculturation, or both. A motherwho is foreign born but who migrated to the U.S. as a young child(also known as the 1.5 generation because characteristics and out-comes are similar to the 2nd generation [Rumbaut, 2004]) is likelyto be more acculturated and speak more English when her childis born than a mother who migrated at an older age. The chil-dren of both of these mothers are considered members of the2nd-generation, however each may experience a very differenthome environment depending on their mother’s age at migration,including the amount of the heritage language used in the homeand the quality of the English input provided to the child. Forimmigrant families and children, especially those who are simi-lar to disadvantaged U.S. minority groups, acculturation may notalways be a positive experience. Declining outcomes with increas-ing acculturation and time in the country, sometimes called theimmigrant paradox (Berry, Phinney, Sam, & Vedder, 2006), havebeen found in a number of areas, including maternal health andstress (Landale et al., 1999), parental living arrangements (Brandon,2002), infant mortality and birth weight (Osypuk, Bates, & Acevedo-Garcia, 2010), infant cognitive development (Fuller et al., 2009), andyouth academic motivation and performance (Suárez-Orozco andSuárez-Orozco, 1995). Although we do not know from the ECLS-B whether families who spoke only English had been in the U.S.longer than families who spoke only their native language, it is apossible explanation for the interactions between immigrant statusand home language found here.

Parental education was more strongly associated with childcognitive, literacy, and engagement with mother outcomes forEnglish-speaking families than for children with other languagesspoken at home. However, the pattern was different for behav-ior problems with parental education being more strongly relatedto improvements in behavior for children whose families spokeanother language at home. The finding of differences for DLLs inthe impact of parental education on different aspects of develop-ment (positive association with socio-emotional outcomes but notwith cognitive/language outcomes) is new to the literature. Simi-larly, the finding that gender differences varied depending on homelanguage group is also novel. When English was used at home, gen-der differences (favoring girls) were observed in most cognitive,educational, social and behavioral outcomes, whereas in homeswhere other languages were spoken, no gender differences wereobserved. Since heritage country of origin was controlled for in theanalyses, known cultural/heritage differences in gender socializa-tion cannot be the explanation for these findings. This suggests thatthere is something specific about heritage language use in the homethat relates to child outcomes differentially for boys and girls thatperhaps should be a topic for future research.


The findings from this investigation have potentially impor-tant implications for parental practices at home for DLL familiesand can inform debates about language-minority parental lan-guage choices at home (Kim & Starks, 2010; Luo & Wiseman,


ING ModelE

Resea

2eEifigalrotpitfl2

tsuPiceosai

dgt2rmbsd(tahiista2

ciybit2LyeadpGabD



000; Mancilla-Martinez & Lesaux, 2011). Many immigrant par-nts are concerned with the speed with which their child will learnnglish and prefer to use English at home rather than their her-tage language (Kim & Starks, 2010; Van Broekhuizen, n.d.). Thendings of this study suggest that if one is a more recent immi-rant (mother was born outside the U.S.), positive child outcomescross several domains are associated with using the heritageanguage at home, either alone or in conjunction with English,elative to the relatively rare group of recent immigrants whonly speak English at home. Furthermore, it is well-documentedhat proficiency in more than one language, or bilingualism,romotes cognitive development in several domains such as

nformation processing, meta-linguistic awareness, perspectiveaking, inhibitory control, executive functioning, and cognitiveexibility (Bialystok, 2009, 2010; Goetz, 2003; Hamers & Blanc,000).

Though there is certainly evidence that early English acquisi-ion is necessary for academic success, research also indicates that atrong foundation in one’s native language is an important contrib-tor to successful English acquisition (August & Shanahan, 2006).erhaps what is most important to general language developments not a particular language that is spoken in the home, but thathildren receive rich exposure to at least one language in the home,nabling them to build a strong linguistic foundation upon whichther languages can be learned (Scheele et al., 2010). This rich expo-ure may be more difficult for immigrant parents to provide if theyre speaking only English to the child and they are not fully fluentn English themselves.

Studies of heritage language maintenance and loss and theevelopment of English proficiency show that use of heritage lan-uage at home is the strongest predictor of children maintainingheir home language (Hakuta & D’Andrea, 1992; Luo & Wiseman,000; Pease-Alvarez, 2002). Children from immigrant families whoeceive considerable exposure to high-quality English in the com-unity early on, such as at preschool or during early schooling also

ecome proficient in English (Tabors & Snow, 2001). English inputpecifically from native speakers of English is also important for L2evelopment, even controlling for total amount of English exposurePlace & Hoff, 2011). Bilingual education provides a promising solu-ion for further supporting children’s language development andchievement (Lindholm-Leary, 2001), but perhaps a more compre-ensive approach is needed to include outreach to parents on the

mportance of heritage language use in the home, as well as increas-ng efforts to bridge culture and language gaps between home andchool (August & Shanahan, 2006). This approach would be consis-ent with research showing that retention of ethnic identity, values,nd community ties can be beneficial for immigrant youth (Akiba,007; Bankston & Zhou, 1997; Rumbaut, 1997).

Coming to a better understanding of the development of DLLhildren is not only scientifically and theoretically important, butt is also critical for practice. With the increasing diversity ofoung children attending pre-k programs, preschools, and home-ased childcare settings, the field of early childhood education

s working hard to establish what constitutes high-quality prac-ices for the education of DLL children (Castro, Espinosa, & Paez,011; Chen, Kyle, & McIntyre, 2007; Espinosa, 2007; Espinosa &ópez, 2007). Although studies on evidence-based practices foroung DLLs (i.e., children ages five and younger) are only nowmerging, current research on evidence-based practices for DLLscross a wide range of ages finds that what benefits non-DLL chil-ren is also beneficial for DLLs, however, DLL students may needarticular supports (August & Shanahan, 2006; Espinosa, 2010;


enesee, Lindholm-Leary, Saunders, & Christian, 2006). There arelso specific suggestions for strategies and practices that mightetter meet the particular needs of DLL children (Castro, Páez,ickinson, & Frede, 2011; Garcia & Jensen, 2009; Garcia Coll et al.,


1996; Goldenberg, 2008; Shivers, Sanders, & Westbrook, 2011;Tang, Dearing, & Weiss, 2012), such as having teachers who canspeak the heritage language (Chang et al., 2007), the presenceof home-language materials in the classroom (Ginsberg, Honda,& O’Neil, 2011), frequent bilingual language interactions (Yates,2011), explicit support to transfer learning from one language tothe other by using L1 strategically in instruction (August et al.,2006; Gillanders & Castro, 2011), and exposure to text with familiarthemes and contexts (Floyd & Carrell, 1987).

The findings of the present study are notable, given that thisstudy used a nationally representative sample with sufficient num-bers of families to examine a variety of heritage country groupsand diversity in language use at home, and utilized complex, mul-tivariate models that allowed us to tease apart unique contributionsand interactions among relevant variables while controlling forthe influence of confounding variables. However, there are also, ofcourse, several limitations. First, as discussed in the introduction,we are limited by the assessment measures that were available atthe time and used in the evaluation of the DLL children in the ECLS-B. Although the measures used in the ECLS-B were commonly usedwithin the DLL literature, at least at that time, they are limited.Most direct assessments were administered in only one language(English or Spanish), so development in both languages for DLLchildren could not be examined. Also, for some cells, there wererelatively small sample sizes (n = 50) based on a particular her-itage country of origin and language group. Further, there wererelatively few children in most heritage groups where no Englishwas occurring at home, so these results have limited generalizabil-ity to children from what is sometimes referred to as ‘linguisticallyisolated households’ (Capps, Fix, Ost, Reardon-Anderson, & Passel,2004). Finally, the ECLS-B surveys included a range of questionsabout language use in the home, but no data were collected on howmuch exposure children received to any particular language, andtherefore, we were unable to examine the effects of this potentiallyimportant factor. More information about the quantity and qualityof language exposure in the home is needed if we are to understandfully early development among dual language learners.

References

Aber, L., Morris, P., & Raver, C. (2012). Children, families and poverty: Definitions,trends, emerging science and implications for poverty. Social Policy Report, 26(3),1–29.

Akiba, D. (2007). Ethnic retention as a predictor of academic success. Lessons fromthe children of immigrant families and Black children. The Clearing House, 80(5),223–226.

Andreassen, C., & Fletcher, P. (2007). Early Childhood Longitudinal Study, Birth Cohort(ECLS – B) psychometric report for the 2-year data collection (NCES 2007-084).Washington, DC: National Center for Education Statistics, Institute of EducationSciences, U.S. Department of Education.

Araújo Dawson, B., Perez, R. M., & Suárez-Orozco, C. (2012). Exploring differences infamily involvement and depressive symptoms across Latino adolescent groups.Journal of Human Behavior in the Social Environment, 22(2), 153–171.

Araújo Dawson, B., & Williams, S. (2008). The impact of language status as an accul-turative stressor on internalizing and externalizing symptoms among Latino/achildren: A longitudinal analysis from school entry through third grade. Journalof Youth and Adolescence, 37(4), 399–411.

Araújo Dawson, B., Williams, S., & Lopez-Humphreys, M. (2007). Examining therelationship between language barriers and life satisfaction among immigrantchildren: An evaluation of a pilot intervention. Journal of Immigrant and RefugeeServices, 5(3), 61–76.

August, D., & Shanahan, T. (2006). Developing literacy in second-language learners:Report of the National Literacy Panel on Language-Minority Children and Youth.Mahwah, NJ: Erlbaum.

August, D., Snow, C., Carlo, M., Proctor, P., de San Francisco, A., Duursma, E., et al.(2006). Literacy development in elementary school second-language learners.Topics in Language Disorders, 26, 351–364.

Bankston, C. L., & Zhou, M. (1997). The social adjustment of Vietnamese Ameri-


can adolescents: Evidence for a segmented assimilation approach. Social ScienceQuarterly, 78(2), 508–523.

Barrueco, S., López, M. L., & Miles, J. C. (2007). Parenting behaviors in the first yearof life: A national examination of Latinos and other cultural groups. Journal ofLatinos and Education, 6, 253–265.


http://refhub.elsevier.com/S0885-2006(14)00013-1/sbref0005








































































































































































































































































ING ModelE

1 Resea

B

B

B

B

B

B

B

C

C

C

C

C

C

C

C

C

C

C

C

D

D

D

E

E

E

F

F

F

F

F

F


4 A. Winsler et al. / Early Childhood

arrueco, S., López, M. L., Ong, C. A., & Lozano, P. (2011). Assessing young childrenwithin and across languages: Approaches and measures. Baltimore, MD: Paul H.Brookes.

ender, D., & Castro, D. (2000). Explaining the birth weight paradox: Latina immi-grants’ perceptions of resilience and risk. Journal of Immigrant Health, 2(3),155–173.

erry, J. W., Phinney, J. S., Sam, D. L., & Vedder, P. (2006). Immigrant youth in cul-tural transition: Acculturation, identity and adaptation across national contexts.Mahwah, NJ: Erlbaum.

ialystok, E. (2009). Bilingualism: The good, the bad, and the indifferent. Bilingual-ism: Language and Cognition, 12, 3–11.

ialystok, E. (2010). Global-local and trail-making tasks by monolingual and bilin-gual children: Beyond inhibition. Developmental Psychology, 46, 93–105.

land-Stewart, L., & Fitzgerald, S. (2001). Use of Brown’s 14 grammatical morphemesby bilingual Hispanic preschoolers: A pilot study. Communication Disorders Quar-terly, 22, 171–186.

randon, P. D. (2002). The child care arrangements of preschool children in immigrantfamilies in the United States. Working paper series. New York, NY: Foundation forChild Development.

abrera, N., Shannon, J., West, J., & Brooks-Gunn, J. (2006). Parental interactionswith Latino infants: Variation by country of origin and English proficiency. ChildDevelopment, 77, 1190–1207.

apps, R., Fix, M., Ost, J., Reardon-Anderson, J., & Passel, J. S. (2004). The health andwell-being of young children of immigrants. Washington, DC: Urban Institute.

astro, D., Espinosa, L. M., & Paez, M. (2011). Defining and measuring quality earlychildhood practices that promote dual language learners’ development andlearning. In M. Zaslow, I. Martinez-Beck, K. Tout, & T. Halle (Eds.), Next steps inthe measurement of quality in early childhood settings (pp. 257–280). Baltimore,MD: Paul H. Brookes.

astro, D. C., Páez, M., Dickinson, D., & Frede, E. (2011). Promoting language and liter-acy in dual language learners: Research, practice and policy. Child DevelopmentPerspectives, 5(1), 15–21.

hao, R. K. (1994). Beyond parental control and authoritarian parenting style:Understanding Chinese parenting through the cultural notion of training. ChildDevelopment, 65, 1111–1120.

hao, R. K. (2001). Extending research on the consequences of parenting style for Chi-nese Americans and European Americans. Child Development, 72(6), 1832–1843.

hang, F., Crawford, G., Early, D., Bryant, D., Howes, C., Burchinal, M., et al.(2007). Spanish-speaking children’s social and language development in pre-kindergarten classrooms. Early Education and Development, 18(2), 243–269.

hen, C. T., Kyle, D. W., & McIntyre, E. (2007). Helping teachers work effectively withEnglish language learners and their families. School Community Journal, 18(1),7–19.

hien, N., Howes, C., Burchinal, M., Pianta, R., Ritchie, D., Bryant, D. M., et al. (2010).Children’s classroom engagement and school readiness gains in prekinder-garten. Child Development, 81, 1534–1549.

oppola, J. (2005). English language learners: Language and literacy developmentduring the preschool years. Journal of the New England Reading Association, 41(2),18–23.

rosnoe, R. (2007). Early child care and the school readiness of children from Mexi-can immigrant families. International Migration Review, 41(1), 152–181.

rosnoe, R. (2013). Preparing the children of immigrants for early academicsuccess. Washington, DC: Migration Policy Institute. Retrieved from:http://www.migrationpolicy.org/pubs/COI-EarlyAcademicSuccess.pdf

e Feyter, J., & Winsler, A. (2009). The early developmental competencies andschool readiness of low-income, immigrant children: Influences of genera-tion, race/ethnicity, and national origins. Early Childhood Research Quarterly, 24,411–431.

uncan, G. J., & Brooks-Gunn, J. (1997). The consequences of growing up poor. NewYork, NY: Russell Sage Foundation.

urden, T. E., & Hummer, R. A. (2006). Access to healthcare among working-agedHispanic adults in the United States. Social Science Quarterly, 87(5), 1319–1343.

spinosa, L. M. (2007). English-language learners as they enter school. In R. Pianta,M. Cox, & K. Snow (Eds.), School readiness and the transition to kindergarten in theera of accountability (pp. 175–196). Baltimore, MD: Paul H. Brookes.

spinosa, L. M. (2010). Classroom teaching and instruction: What are best practices foryoung English language learners? New York, NY: Teachers College Press.

spinosa, L., & López, M. (2007). Assessment considerations for young Englishlanguage learners across different levels of accountability. Retrieved fromhttp://www.pewtrusts.org/our work report detail.aspx?id=31164

abiano-Smith, L., & Barlow, J. A. (2010). Interaction in bilingual phonological acqui-sition: Evidence from phonetic inventories. International Journal of BilingualEducation and Bilingualism, 13, 81–97.

abiano-Smith, L., & Goldstein, B. (2010). Early-, middle-, and late-developing soundsin monolingual and bilingual children: An exploratory investigation. AmericanJournal of Speech-Language Pathology, 19, 66–77.

arver, J. A. M., Kim, Y. K., & Lee, Y. (1995). Cultural differences in Korean- and Anglo-American preschoolers’ social interaction and play behaviors. Child Development,66, 1088–1099.

ield, T. M., & Widmayer, S. M. (1981). Mother-infant interactions among lower SESBlack, Cuban, Puerto Rican and South American immigrants. Culture and Early


Interactions, 41–62.loyd, P., & Carrell, P. (1987). Effects on ESL reading of teaching cultural content

schemata. Language Learning, 37, 89–108.ortuny, K., Hernandez, D. J., & Chaudry, A. (2010). Young children of immigrants: The

leading edge of America’s future. Washington, DC: Urban Institute.


Francis, D. J., & Rivera, M. O. (2007). Principles underlying English language pro-ficiency tests and academic accountability for ELLs. In J. Abedi (Ed.), Englishlanguage proficiency assessment in the nation: Current status and future practice(pp. 13–31). Davis, CA: University of California, Davis.

Fuligni, A. (1997). The academic achievement of adolescents from immigrant fami-lies: The roles of family background, attitudes, and behavior. Child Development,68, 351–363.

Fuller, B., Bridges, M., Bein, E., Jang, H., Rabe-Hesketh, S., Halfon, N., et al. (2009). Thehealth and cognitive growth of Latino toddlers: At risk or immigrant paradox?Maternal and Child Health Journal, 13, 755–768.

Galindo, C., & Fuller, B. (2010). The social competence of Latino kindergarteners andgrowth in mathematical understanding. Developmental Psychology, 46, 579–592.

Gandara, P., Rumberger, J., Maxwell-Jolly, R., & Callahan, R. (2003). English learnersin California schools: Unequal resources, unequal outcomes. Educational PolicyAnalysis Archives, 11, 1–54.

Garcia, E., & Jensen, B. (2009). Early educational opportunities for children of His-panic origins. SRCD Social Policy Report, 23(2), 1–20.

Garcia Coll, C., Lamberty, G., Jenkins, R., McAdoo, H. P., Crnic, K., Wasik, B. H., et al.(1996). An integrative model for the study of developmental competencies inminority children. Child Development, 67, 1891–1914.

Garcia Coll, C., & Marks, A. K. (Eds.). (2012). The immigrant paradox in children and ado-lescents: Is becoming American a developmental risk?. Washington, DC: AmericanPsychological Association.

Genesee, F., Lindholm-Leary, K., Saunders, B., & Christian, D. (2006). Educating Englishlanguage learners: A synthesis of research evidence. New York, NY: CambridgeUniversity Press.

Genesee, F., & Nicoladis, E. (1995). Language development in bilingual preschoolchildren. In E. Garcia, & B. McLaughlin (Eds.), Meeting the challenge of linguisticand cultural diversity in early childhood education (pp. 18–33). New York, NY:Teachers College Press.

Genesee, F., Paradis, J., & Crago, M. B. (2010). Dual language development and disorders:A handbook on bilingualism and second language learning (2nd ed.). Baltimore,MD: Paul H. Brookes.

Gillanders, C., & Castro, D. C. (2011). Storybook reading for young English languagelearners. Young Children, (January), 91–95.

Ginsberg, D., Honda, M., & O’Neil, W. (2011). Looking beyond English: Linguis-tic inquiry for English language learners. Language and Linguistics Compass, 5,249–264.

Glick, J. E., Bates, L., & Yabiku, S. T. (2009). Mother’s age at arrival in the UnitedStates and early cognitive development. Early Childhood Research Quarterly, 24,367–380.

Goetz, P. (2003). The effects of bilingualism on theory of mind development. Bilin-gualism: Language and Cognition, 6(1), 1–15.

Goldenberg, C. (2008). Improving achievement for English language learners. In S.Neuman (Ed.), Educating the other America: Top experts tackle poverty, literacy,and achievement in our schools (pp. 139–162). Baltimore, MD: Paul H. Brookes.

Hakuta, K., & D’Andrea, D. (1992). Some properties of bilingual maintenance and lossin Mexican background high-school students. Applied Linguistics, 13, 72–99.

Hamers, J. F., & Blanc, M. H. A. (2000). Bilinguality and bilingualism (2nd ed.). NewYork, NY: Cambridge University Press.

Hammer, C., Lawrence, F., Rodriguez, B., Davison, M., & Miccio, A. (2011). Changein language usage of Puerto Rican mothers and their children: Do gender andtiming of exposure to English matter? Applied Psycholinguistics, 32, 275–297.

Han, W. J. (2010). Bilingualism and socioemotinoal well-being. Children and YouthServices Review, 32, 720–731.

Han, W. J., & Huang, C. C. (2010). The forgotten treasure: Bilingualism and Asianchildren’s emotional and behavioral health. American Journal of Public Health,100, 831–838.

Han, W. J., Lee, R., & Waldfogel, J. (2012). School readiness among children of immi-grants in the US: Evidence from a large national birth cohort study. Children andYouth Services Review, 34, 771–782.

Hao, L., & Bonstead-Bruns, M. (1998). Parent-child differences in educational expec-tations and the academic achievement of immigrant and native students.Sociology of Education, 71, 175–198.

Hernandez, D. J., Denton, N. A., & Blanchard, V. L. (2011). Children in the UnitedStates of America: A statistical portrait by race-ethnicity, immigrant origins,and language. Annals of the American Academy of Political and Social Science, 633,102–127.

Hernandez, D. J., Denton, N. A., & McCartney, S. E. (2007). Children in immigrantfamilies – The U.S. and 50 states: National origins, language, and early education.Washington, DC: Child Trends.

Hoff, E., Core, C., Place, S., Rumiche, R., Senor, M., & Parra, M. (2012). Dual languageexposure and early bilingual development. Journal of Child Language, 39, 1–27.

Howes, C., Burchinal, M., Pianta, R., Bryant, D., Early, D., Clifford, R. M., et al. (2008).Ready to learn? Children’s pre-academic achievement in pre-Kindergarten pro-grams. Early Childhood Research Quarterly, 23, 27–50.

Hurh, W. M., & Kim, K. C. (1990). Religious participation of Korean immigrants in theUnited States. Journal for the Scientific Study of Religion, 29, 19–34.

Jackson, M. I., Kiernan, K., & McLanahan, S. (2012). Immigrant-native differences inchild health: Does maternal education narrow or widen the gap? Child Develop-ment, 83, 1501–1509.


Jung, S., Fuller, B., & Galindo, C. (2012). Family functioning and early learning prac-tices in immigrant homes. Child Development, 83, 1510–1526.

Keels, M. (2009). Ethnic group differences in Early Head Start parents’ parentingbeliefs and practices and links to children’s early cognitive development. EarlyChildhood Research Quarterly, 24, 381–397.
















































































































































































































































































































































































































http://www.migrationpolicy.org/pubs/COI-EarlyAcademicSuccess.pdf






















































































































http://www.pewtrusts.org/our work report detail.aspx?id=31164
















































































































































































































































































































































































































































































































































































































































































































































































































































ING ModelE

Resea

K

K

K

L

L

L

L

L

L

M

M

M

M

M

M

N

N

N

N

N

N

N

O

O

O

P

P

P

P

based childcare and public school pre-kindergarten programs. Early ChildhoodResearch Quarterly, 23, 314–329.



im, S. H., & Starks, D. (2010). The role of fathers in language maintenance andlanguage attrition: The case of Korean-English late bilinguals in New Zealand.International Journal of Bilingual Education and Bilingualism, 13, 285–301.

uhl, P. K. (2001). Speech, language, and developmental change. In F. Lacerda, C. vonHofsten, & M. Heimann (Eds.), Emerging cognitive abilities in early infancy (pp.111–134). Mahwah, NJ: Erlbaum.

upisch, T. (2007). Determiners in bilingual German-Italian children: What theytell us about the relation between language influence and language dominance.Bilingualism: Language and Cognition, 10, 57–78.

andale, N. S., Oropesa, R. S., Llanes, D., & Gorman, B. K. (1999). Does Americanizationhave adverse effects on health? Stress, health habits, and infant health outcomesamong Puerto Ricans. Social Forces, 78(2), 613–641.

eventhal, T., Xue, Y., & Brooks-Gunn, J. (2006). Immigrant differences in school-agechildren’s verbal trajectories: A look at four racial/ethnic groups. Child Develop-ment, 77, 1359–1374.

im, S., & Lim, B. K. (2003). Parenting style and child outcomes in Chinese and immi-grant Chinese families: Current findings and cross-cultural considerations inconceptualization and research. Marriage & Family Review, 35(3–4), 21–43.

indholm-Leary, K. J. (2001). Dual language education. Clevedon, UK: MultilingualMatters.

uchtel, M., Hughes, K., Luze, G., Bruna, K., & Peterson, C. (2010). A comparison ofteacher-rated classroom conduct, social skills, and teacher-child relationshipquality between preschool English learners and preschool English speakers.NHSA Dialog, 13, 92–111.

uo, S., & Wiseman, R. (2000). Ethnic language maintenance among Chinese immi-grant children in the United States. International Journal of Intercultural Relations,24, 307–324.

agnuson, K. A., Lahaie, C., & Waldfogel, J. (2006). Preschool and school readinessof children of immigrants. Social Science Quarterly, 87(5), 1241–1262.

ancilla-Martinez, J., & Lesaux, N. (2011). Early home language use and later vocab-ulary development. Journal of Educational Psychology, 103, 535–546.

atthews, H., & Ewen, D. (2006). Early education programs and children of immigrants:Learning each other’s language. Washington, DC: The Urban Institute. Retrievedfrom http://www.urban.org/UploadedPDF/412205-early-education.pdf

cLoyd, V. C. (1990). The impact of economic hardship on black families and chil-dren: Psychological distress, parenting, and socioemotional development. ChildDevelopment, 61, 331–346.

in, P. G. (1992). The structure and social functions of Korean immigrant churchesin the United States. International Migration Review, 26(4), 1370–1394.

ishina-Mori, S. (2005). Autonomous and interdependent development of twolanguage systems in Japanese/English simultaneous bilinguals: Evidence fromquestion formation. First Language, 25, 291–315.

ajarian, M., Snow, K., Lennon, J., & Kinsey, S. (2010). Early Childhood Longitudi-nal Study, Birth Cohort (ECLS-B): Preschool-kindergarten. Washington, DC: U.S.Department of Education.

ajarian, M., Snow, K., Lennon, J., Kinsey, S., & Mulligan, G. (2010). Early ChildhoodLongitudinal Study, Birth Cohort (ECLS-B): Preschool-kindergarten psychometricreport. Washington, DC: U.S. Department of Education.

ational Center for Education Statistics. (n.d.). Two bags task in the ECLS-B. Retrievedfrom http://nces.ed.gov/pubs2007/2007084 C6.pdf

ational Clearinghouse for English Language Acquisition (NCELA). (2010). The grow-ing numbers of English learner students, 1997/98–2007/08. Washington, DC: U.S.Department of Education.

ational Institute for Child Health and Human Development Early Child CareResearch Network, & Duncan, G. J. (2003). Modeling the impacts of child carequality on children’s preschool cognitive development. Child Development, 74,1454–1475.

elson, K. (2005). Cognitive functions of language in early childhood. The develop-ment of social cognition and communication. In B. Homer, & C. Tamis-LeMonda(Eds.), The development of social cognition and communication (pp. 7–28). Mah-wah, NJ: Erlbaum.

ord, C., Edwards, B., Andreassen, C., Green, J. L., & Wallner-Allen, K. (2006). EarlyChildhood Longitudinal Study, Birth Cohort (ECLS-B): User’s manual for the ECLS-Blongitudinal 9-month–2-year data file and electronic codebook (NCES 2006-046).Washington, DC: National Center for Education Statistics.

ller, D. K., & Jarmulowicz, L. (2007). Language and literacy in bilingual children inthe early school years. In E. Hoff, & M. Shatz (Eds.), Blackwell handbook of languagedevelopment (pp. 368–386). Malden, MA: Blackwell.

ller, D. K., Pearson, B. Z., & Cobo-Lewis, A. B. (2007). Profile effects in early bilinguallanguage and literacy. Applied Psycholinguistics, 28(2), 191–230.

sypuk, T. L., Bates, L. M., & Acevedo-Garcia, D. (2010). Another Mexican birth-weight paradox? The role of residential enclaves and neighborhood povertyin the birthweight of Mexican-origin infants. Social Science & Medicine, 70(4),550–560.

adilla, A. M., Cervantes, R. C., Maldonado, M., & Garcia, R. E. (1988). Coping responsesto psychosocial stress among Mexican and Central American immigrants. Journalof Community Psychology, 16(4), 418–427.

aez, M., Tabors, P., & Lopez, L. (2007). Dual language and literacy development ofSpanish-speaking preschool children. Journal of Applied Developmental Psychol-ogy, 28, 85–102.


aradis, J. (2001). Are object omissions in Romance object clitic omissions? Bilin-gualism: Language and Cognition, 4, 36–37.

ease-Alvarez, L. (2002). Moving beyond linear trajectories of language shift andbilingual language socialization. Hispanic Journal of Behavioral Sciences, 24,114–137.


Perreira, K. M., Chapman, M. V., & Stein, G. L. (2006). Becoming an American par-ent: Overcoming challenges and finding strength in a new immigrant Latinocommunity. Journal of Family Issues, 27, 1383–1414.

Place, S., & Hoff, E. (2011). Properties of dual language exposure that influence 2-year-olds’ bilingual proficiency. Child Development, 82, 1834–1849.

Portes, P. (1999). Social and psychological factors in the academic achievement ofchildren of immigrants: A cultural history puzzle. American Educational ResearchJournal, 36(3), 489–507.

Quintana, S. M., Aboud, F. E., Chao, R. K., Contreras-Grau, J., Cross, W. E., Hudley, C.,et al. (2006). Race, ethnicity, and culture in child development: Contemporaryresearch and future directions. Child Development, 77, 1129–1141.

Rogoff, B. (2003). The cultural nature of human development. Oxford, UK: OxfordUniversity Press.

Rumbaut, R. G. (1997). Paradoxes (and orthodoxies) of assimilation. SociologicalPerspectives, 40(3), 483–511.

Rumbaut, R. G. (2004). Ages, life stages, and generational cohorts: Decomposingthe immigrant first and second generations in the United States. InternationalMigration Review, 38(3), 1160–1205.

Scheele, A., Leseman, P., & Mayo, A. (2010). The home language environment ofmonolingual and bilingual children and their language proficiency. Applied Psy-cholinguistics, 31, 117–140.

Shah, P. G. (2011). Asian Americans’ achievement advantage: When and why does itemerge? (Unpublished dissertation). Columbus, OH: Ohio State University.

Shivers, E., Sanders, K., & Westbrook, T. (2011). Measuring culturally responsiveearly care and education. Quality measurement in early childhood settings. InM. Zaslow, I. Martinez-Beck, K. Tout, & T. Halle (Eds.), Quality measurement inearly childhood settings (pp. 191–225). Baltimore, MD: Paul H. Brookes.

Shweder, R. A., Goodnow, J., Hatano, G., LeVine, R. A., Markus, H., & Miller, P. (2006).The cultural psychology of development: One mind, many mentalities. In W.Damon, & R. M. Lerner (Eds.), Handbook of child psychology (Vol. 1) (6th ed., Vol.1, pp. 716–792). Hoboken, NJ: Wiley.

Smart, J. F., & Smart, D. W. (1995). Acculturative stress of Hispanics: Loss and chal-lenge. Journal of Counseling and Development, 73(4), 390–396.

Solórzano, R. W. (2008). High stakes testing: Issues, implications, and remedies forEnglish language learners. Review of Educational Research, 78, 260–329.

Suárez-Orozco, C., & Suárez-Orozco, M. (1995). Transformations: Immigration, familylife, and achievement motivation among Latino adolescents. Palo Alto, CA: StanfordUniversity Press.

Suárez-Orozco, C., & Suárez-Orozco, M. (2001). Children of immigration. Cambridge,MA: Harvard University Press.

Suárez-Orozco, M. M., & Paez, M. (Eds.). (2002). Latinos: Remaking America. Berkeley,CA: University of California Press.

Tabors, P. O., Paez, M., & Lopez, L. M. (2003). Dual language abilities of Spanish-English bilingual four-year olds: Initial finding from the Early Childhood Study ofLanguage and Literacy Development of Spanish-speaking children. NABE Journalof Research and Practice, 1, 70–91.

Tabors, P. O., & Snow, C. E. (2001). Young bilingual children and early literacy devel-opment. In S. B. Neuman, & D. K. Dickinson (Eds.), Handbook of early literacyresearch (2nd ed., pp. 159–178). New York, NY: Guilford.

Takanishi, R. (2010). Introduction. In E. L. Grigorenko, & R. Takanishi (Eds.), Immi-gration, diversity, and education (pp. 1–6). New York, NY: Routledge.

Tamis-LeMonda, C. S., Song, L., Smith Leavell, A., Kahana-Kalman, R., & Yoshikawa,H. (2012). Ethnic differences in mother-infant language and gestural communi-cations are associated with specific skills in infants. Developmental Science, 15,384–397.

Tang, S., Dearing, E., & Weiss, H. (2012). Spanish-speaking Mexican-American fam-ilies’ involvement in school-based activities and their children’s literacy: Theimplications of having teachers who speak Spanish and English. Early ChildhoodResearch Quarterly, 27, 177–187.

U.S. Census Bureau. (2010). Language use in the United States: 2007. Retrieved fromhttp://www.census.gov/hhes/socdemo/language/data/acs/ACS-12.pdf

U.S. Department of Education. (2008). National assessment of educational progress inreading and mathematics. Washington, DC: Author.

Van Broekhuizen, L. (n.d.). Parents and bilingual learners: The impor-tance of fostering first language development at home. Retrieved fromhttp://www.prel.org/products/paced/nov02/re bilingual.htm

Vygotsky, L. (1986). Though and language. Cambridge, MA: MIT Press.Wasserman, G. A., Brunelli, S. A., Rauh, V. A., & Alvarado, L. E. (1994). The cultural

context of adolescent childbearing in three groups of urban minority mothers:Age and ethnic variations. Child Development, 61, 566–580.

West, J., & Andreassen, C. (2002). Measuring early development in the Early Child-hood Longitudinal Study – Birth Cohort. In Paper presented in the selectingmeasures for young children in large-scale surveys workshop Washington, DC, May.

Winsler, A., Fernyhough, C., & Montero, I. (Eds.). (2009). Private speech, executive func-tioning, and the development of verbal self-regulation. New York, NY: CambridgeUniversity Press.

Winsler, A., Tran, H., Hartman, S., Madigan, A., Manfra, L., & Bleiker, C. (2008). Schoolreadiness gains made by ethnically diverse children in poverty attending center-


Yates, L. (2011). Interaction, language learning and social inclusion in earlysettlement. International Journal of Bilingual Education and Bilingualism, 14,457–471.









































































































































































































































































http://www.urban.org/UploadedPDF/412205-early-education.pdf









































































































http://nces.ed.gov/pubs2007/2007084_C6.pdf





































































































































































































































































































































































































































































































































































































































































































































































http://www.census.gov/hhes/socdemo/language/data/acs/ACS-12.pdf














http://www.prel.org/products/paced/nov02/re_bilingual.htm





































































































































g model article in press - george mason universitywinslerlab.gmu.edu/pubs/winslereclsb2014.pdf ·...

Documents