Hoff, Erika

Learning to read is a fundamental skill that is understudied among children who hear more than one language at home. A particular limitation in the extant literature is that the samples are often Spanish-dominant and come from low-income households. This literature has evidence that Spanish skills – both language-specific (i.e., vocabulary) and language-general (i.e., phonological processing) – predict English reading ability. In the current study, we investigated the effect of oral language skills in both English and Spanish, as well as other pre-literacy skills, on the English reading skill of Spanish-English bilinguals who are English dominant and received English only instruction from school entry. The oral language skills and other pre-literacy skills of 101 Spanish-English dual language learners were assessed at 5 years. English reading skill was measured annually from 6 to 10 years. Latent growth curve analysis was used to model initial English reading skill at 6 years and the growth of English reading skill from 6 to 10 years. Four sets of hypothesized foundational skills measured at child age 5 years were tested as predictors of subsequent English reading skill: (1) English oral language skills, (2) Spanish oral language skills, (3) English and Spanish oral language skills, and (4) English and Spanish oral language skills, with other pre-literacy skills. The results of structural equation modeling indicated that English vocabulary and phonological awareness measured in Spanish were significant predictors when English and Spanish skills were entered separately. When English and Spanish oral language skills were included together, Spanish oral language skills did not explain English reading growth better than when oral language skills were modeled independently. The best model of predictors of English reading for bilingual 5-year-olds included only English vocabulary and English letter recognition.

Code-switching is a common feature of bilingual language use and has multiple factors that influence the frequency and type of code-switching. 56 Spanish-English bilingual children recorded sessions of Spanish-designated and English-designated interactions with a caregiver at 2.5 and 3.5 years. These sessions were transcribed and coded for all code-switched utterances. At both ages, we found: (1) Children switched to English more frequently than they switched to Spanish. (2) Their degree of English dominance was a positive predictor of their frequency of switching to English, but a negative predictor of their frequency of switching to Spanish. Between 2.5 and 3.5 years, children became more English dominant, and their rate of switching to English increased while their rate of switching to Spanish decreased. The present findings suggest that the strongest influence on bilingual children’s code-switching is their relative proficiency in their two languages and as that proficiency changes, their code-switching changes.

Adult-child interactions vary between cultures. For example, Hispanic parents are characterized by a more adult-centered style of interaction with children, while European American parents are more child-centered. Little is known about the influences cultural differences may have on the ways that Spanish-English bilingual parents speak to their children in each language. To address this question, 17 monolingual Spanish-speaking Hispanic American mothers, 22 monolingual English-speaking European American mothers, and 33 Spanish-English bilingual mothers were videorecorded in toy-play interactions with their children. The bilingual mothers and children were recorded in two sessions, one in which they were instructed to speak English and one in which they were instructed to speak Spanish. Using CHILDES programs, these interactions were transcribed and coded for properties of parent-child conversation known to be related to child language outcomes and hypothesized to reflect parent-centered and child-centered styles of interaction. The parent-child conversations of the two monolingual groups were compared in order to obtain baseline cultural differences in interaction style. The parentchild conversations of the bilingual mothers when speaking Spanish and when speaking English were compared in terms of the properties that showed differences between the monolingual groups. The conversations of the monolingual Hispanic American mothers were characterized by fewer maternal word types, and proportionately fewer maternal questions, and fewer child utterances than the conversations of the monolingual European American mothers. These differences were reflected in the comparisons of the bilingual mothers’ Spanish and English interactions with the exception of number of word types. The results are consistent with the hypotheses that (1) Spanish-speaking Hispanic American mothers use a more adult-centered style of interaction with their children compared to European American mothers, who use a more child-centered style of interacting with children and that (2) Hispanic American bilingual mothers reflect aspects of these cultural differences when speaking each language with their children.

Extensive evidence indicates that oral language skills at school entry predict later reading development among monolingual children. It is not clear if the effect is the same for bilingually developing children and whether their oral skills in one language can transfer to reading comprehension in the other. The current longitudinal study followed 72 Spanish-English bilingual children (42 girls, 30 boys) and examined the extent to which early oral language proficiency in English and in Spanish were related to later reading comprehension development within- and across-languages. Multilevel models revealed significant within-language relations between oral language skills at 5 years and reading comprehension growth from 6 to 8 years in both English and Spanish. Additionally, English oral skill predicted Spanish reading comprehension, whereas Spanish oral skill was unrelated to English reading comprehension. Theoretical and practical implications of the findings are discussed.

The current studies tested the hypothesis, that early exposure is sufficient for
nativelike proficiency. Study 1 compared the English skill of 116 5 year olds who had
been exposed to English and Spanish from birth with English monolingual and found that
the bilingual children had significantly lower levels of vocabulary skill. Study 2 assessed
65 adult bilinguals, comparing them to 25 English and 25 Spanish monolinguals on a
battery of language measures. The bilinguals had lower scores in Spanish in 7 of the 8
domains of language skill measured. The bilinguals were not different from the English
monolingual speakers in most, but not all, aspects of language proficiency. These
findings provide evidence that the monolingual-bilingual gap observed in childhood is no
longer evident among adult bilinguals, but that despite early exposure and continued use
into adulthood, second generation immigrants are not native-like in their heritage
language skills.

Research indicates that there are numerous factors related to dual language
proficiency. Furthermore, research suggests that the language skills of bilingual children
are very heterogeneous. In this study, we will focus on four outcomes for those receiving
dual language input. Those four patterns of outcome are as follows: bilingual children can
have high levels of skills in English and Spanish, high skills in English and low skills in
the Spanish or vice versa, or low levels of skills in English and Spanish. The present
study focused on the prevalence of these four patterns and their concurrent correlates
amongst 30 month old (N = 87, 48 females) and 48 month old (N = 74, 40 females)
simultaneous English-Spanish bilinguals with special emphasis on the distinguishing
factors for those with high dual language skills. Overall, for those with high dual scores,
the findings reveal that gender, birth-order, and maternal level of education are predictive of high dual language scores so that members of this group are largely first-born females
of mothers with an average 15.8 years of education. English dominant scores were
predicted by percent English input in the home and birth-order so that children in this
group were mostly later-born children with 65% English input in the home. Spanish
dominant scores were predicted by percent English input in the home and Spanish bookreading
behaviors as children in this group received 23% of their in-home input in
English and were read to for more than one hour per week in Spanish. Low dual scores
were marginally predicted by gender and Spanish book-reading behaviors, children in
this group were mostly male and received less than an in hour of Spanish reading per
week.

In order to explain the development of English and Spanish skills in children from Spanish-speaking homes in the United States, it is necessary to identify their sources of language exposure. Most research to date has focused on home language use. The aim of this study is to identify sources of English and Spanish exposure outside the home that bilingual children experience between the ages of 2 and 5 years. The present study focuses on 3 potential sources: grandparents, extracurricular activities, and early childcare and education settings. We ask how much English and Spanish exposure children receive, how that changes from 2.5 to 5 years, and whether family variables influence those changes. Participants were 149 children from Spanish-speaking homes in southeastern Florida, with at least one parent an immigrant from a Spanish-speaking country. Measured out-of-home sources of input included hours per of week of English and Spanish from a Grandparent, during Extracurricular Activities, and during Preschool at 30, 36, 42, 48, and 60 months. Parents’ Native Language Background groups were either both native Spanish-speaking or one native, Spanish-speaking and one native, English-speaking. Maternal Education was treated as a dichotomous variable: mothers whose highest level of education in English is less than a four-year college degree and mothers whose highest level of education in English is equivalent to or greater than a four-year college degree. Child Birth Order was also treated as a dichotomous variable: only children and first-born children or later born children. Results revealed that for these children from Spanish-speaking homes, Grandparents are primarily a source of heritage-language (Spanish) input and Extracurricular Activities and Preschool are primarily a source of societal-language (English) input. Findings suggest English exposure from out-of-home sources increases over time possibly at the expense of Spanish exposure. Parents’ Native Language Backgrounds and Maternal Education influenced children’s exposure to both languages from these outside sources of input; Child Birth Order did not. Implications for future research and practical application are discussed.

This study examined the relation between home language exposure from different
household members, specifically older siblings, and English and Spanish vocabulary
development in bilingual toddlers. The English and Spanish vocabularies of 38 toddlers
(19 boys and 19 girls; Mage = 2.14 years, SD = .14) were measured. Parental EI?-glish use
was found to be the strongest predictor of English and Spanish proficiencies. Among
households in which the parents use less than 10% English, having older siblings had no
effect on English language proficiency but was associated with lower levels of Spanish
proficiency.

Previous studies have found that bilingual children’s vocabulary development
benefits more from child-directed speech from native speakers than child-directed speech
from nonnative speakers. The current study compared the native English child-directed
speech of 20 English monolingual mothers, the nonnative English child-directed speech
of 20 Spanish-English bilingual mothers, and the native Spanish child-directed speech of
the same bilingual mothers in terms of three aspects of input previously associated with
children’s language development: data-providing properties, topic contingency, and
speech function. There were significant differences between native English and nonnative
English child-directed speech, and between nonnative English and native Spanish. The
results suggest two sources of influence shaping child-directed speech: quality
differences related to native speaker status and cultural factors primed by the language
being spoken.

Bilingual children sometimes respond to their interlocutors using a different language than the one in which they were addressed. These language choices, their concurrent correlates, and relations to subsequent language growth were examined in 91 Spanish-English bilingual children (44 girls, 47 boys). Children's language choices were assessed at 30 months, and their English and Spanish productive vocabularies and receptive language skills were examined at the ages of 30, 36, and 42 months. Children's language choices were concurrently related to English and Spanish productive and receptive language scores and to mothers' English and Spanish proficiency levels. Longitudinal multi-level modeling analyses indicated that children's language choices at 30 months were uniquely related to language growth on measures of English and Spanish productive vocabulary and Spanish receptive language, when controlling for language expo sure. These findings suggest that language use plays a causal role in language development. The findings of this study have implications for the maintenance of heritage languages in the U.S. and for the development of children's English language skills.