predict later reading and other literacy skills, David Dickinson reported. For instance, several studies and a recent meta-analyses conducted by the National Early Literacy Panel (2008) have reported long-term relationships between aspects of vocabulary in early childhood and measures of reading comprehension in middle school and high school (see Dickinson and Freiberg, 2009). Beyond vocabulary, length of utterances and diversity of words at age 3 has predicted language, spelling, and reading skills in kindergarten and 3rd grade after controlling for SES and school attended (Walker et al., 1994). Similarly, a longitudinal study of 626 children from low-income homes showed that oral language at age 4 predicted reading comprehension in 3rd and 4th grade (Storch and Whitehurst, 2002). The National Early Reading Panel (2008) also recently reported from a meta-analysis that although expressive and productive vocabulary predicted reading comprehension, several measures of complex language, such as complexity of grammar, were better predictors.
Dickinson turned next to studies documenting group differences in language and achievement associated with SES, race, ethnicity, and home language. Such group differences emerge early and only increase with age (see Dickinson and Freiberg, 2009). Early language differences have been documented for each of these groups in nationally representative samples including research on Head Start (FACES, 2006), the Early Child Care Study (National Institute of Child Health and Human Development Early Child Care Research Network, 2003), and the Early Childhood Longitudinal Study (Chernoff et al., 2007). In addition, recent data collected by Dickinson and Ann Keiser at Vanderbilt University with 440 African American children in a Head Start program showed that standard measures of expressive and receptive vocabulary and broader language skills were all about one standard deviation below national norms for all children in the United States. Likewise, as mentioned earlier by Hoff, individual differences in language processing speed begin early, affect language learning, and are associated with poverty. More specifically, Anne Fernald and her colleagues (e.g., Fernald, Perfors, and Marchman, 2006) have shown that the speed of children’s lexical access at 12-25 months correlates with parents’ reports of children’s vocabulary size and complexity of language use at 25 months. Children’s rate of productive vocabulary growth also correlates with speed of lexical processing. In other research, Maria Vasilyeva and colleagues (2006) examined the use of complex syntax in relation to SES: they found that, at 22 months, all children used complex syntax at least some times, showing the potential to develop language complexity. However, by 42 months, the number of complex sentences used by children with low SES, while increasing, was much lower than the number used by more advantaged children (Vasilyeva, Huttenlocher, and Waterfall, 2006; Vasilyeva, Waterfall, and