Increases in maternal education and young children's language skills.(Report)

By pursuing more education, do mothers set their children on better academic courses? Or do such pursuits do little to alter children's achievement trajectories? Children of more highly educated parents enter school with higher levels of academic skills and continue to perform better than other children (Entwisle & Alexander, 1993; Lee & Burkham, 2002). Social scientists often attribute this achievement gap to the mix of biological, economic, and social advantages afforded by higher levels of parental education, but debate continues regarding the extent to which parental education is causally associated with children's achievement (Sobel, 1998).

Researchers have studied parental education largely as if it were static. Reflecting a broader trend of diverging life course patterns, it has become common for adults to accrue education in a discontinuous fashion and to attend school well into adulthood (Astone, Schoeni, Ensminger, & Rothert, 2000; Jacobs & Stoner-Eby, 1998). Nationally, more than 20% of adult women pursue some type of education (Rich & Kim, 1999), and about 27% of female college students are over age 25 (Shin, 2005). Economically disadvantaged mothers are especially likely to return to school; close to 50% of low-income mothers attend school after the birth of their children (Love et al., 2002; McGroder, Zaslow, Moore, & LeMenestrel, 2000; Rich & Kim, 1999).

Nearly 20 years ago Furstenberg, Brooks-Gunn, and Morgan (1987) found that some children born to poor, urban, adolescent mothers seemed to flourish despite experiencing the same adverse early environments as peers who did not fare so well. More than 50% of mothers in their study had attended school after the birth of their children, and Furstenberg and colleagues argued that this additional education contributed to children's subsequent scholastic success. This was a compelling yet speculative hypothesis. They could not conclude that it was the mothers' education per se that mattered rather than other characteristics that were correlated with mothers' schooling.

Understanding whether changes in mothers' education affects children's developmental trajectories is important not only because such changes are common but also because it may illuminate whether parental education, more generally, influences children. One key difficulty in this research is that maternal education completed before the birth of children is likely to be confounded with other important mother, child, and family characteristics that are difficult to untangle. For example, isolating the effects of a mother's schooling from that of her intellectual ability or academic motivation is challenging. If a mother's education is already completed, it is likely that her educational attainment reflects these factors and also that these factors independently affect her child's achievement. Thus, researchers end up in the position of having to overcontrol or undercontrol for individual and family characteristics (Newcombe, 2003).

When a mother's education improves after the birth of her children, this provides an opportunity to estimate the association between maternal education and children's development with fewer concerns about confounding factors. Although unobserved characteristics may differentiate mothers who return to school from those who do not, if the associations between these unobserved characteristics and child outcomes are constant over time, then lagged dependent variable regression analyses provide a way to reduce omitted variable bias (NICHD Early Child Care Research Network & Duncan, 2003). As with all nonexperimental research, omitted variable biases may remain if the effects of unobserved characteristics on child development vary over time.

Existing theory and research on how levels of maternal education affect young children provide some suggestions about how increments in maternal education may matter. To the extent that a mother's additional schooling provides her with positive learning experiences as well as increases in basic skills, knowledge, and higher-order thinking, it may shape her expectations for her children's education and enable her to create better home learning environments for her children (Alexander, Entwisle, & Bedinger, 1994; Corwyn & Bradley, 2003; Davis-Kean, 2005). Parents with higher levels of education engage their children in more learning-related activities both in the home (e.g., reading books) and out of the home (e.g., music or art lessons) (Davis-Kean, 2005). Parents' educational expectations and opportunities for children to engage in learning activities have been linked to children's cognitive development and their academic skills (Taylor, Clayton, & Rowley, 2004).

Of particular importance for young children may be the quality of mother-child interactions and mothers' verbal responsiveness, which are important for the emergent language and cognitive skills (Raviv, Kessenich, & Morrison, 2004). More highly educated mothers are more verbally responsive to their young children and tend to use teaching strategies with their children that mimic formal instructional techniques, such as asking questions and offering feedback, rather than issuing directives (Richman, Miller, & LeVine, 1992; Tracey & Young, 2002). Hoff-Ginsberg's (1998) research with mothers of 2-year-old children found that compared with high school-educated mothers, college-educated mothers talked more, asked more questions, and used fewer directives and more contingent responses. Although these specific observed differences in learning activities and speech patterns have not been directly linked to children's language development, there is evidence that socioeconomic status (SES) differences in maternal speech explain SES differences in children's language development (Hoff, 2003; Huttenlocher, Vasilyeva, Cymerman, & Levine, 2002).