Family risk factors that are associated with poor performance in school-aged children are also linked with lower proficiency in early reading and mathematics skills and general knowledge among children as they enter kindergarten. These risk factors are low maternal education, welfare dependency (as a marker of family poverty), having only one parent in the home, and having parents whose primary language is not English. As with previous studies (Zill et al. 1995), the ECLS-K data show that there is a cumulative effect of the number of risks to which a child is exposed early in life. While children with one risk factor do not fare as well as those with none, children who have two or more risk factors exhibit greater achievement lags, poorer health, more problem behavior, and less positive approaches to learning than do children with a single risk factor. Further examination of the data are required to reveal whether each risk factor is of approximately equal importance or whether some are more significant than others. Some researchers have theorized that the number of risk factors in a child's background may be more important than the nature of the particular risk or risks (Meisels and Wasik 1990). Others believe that low parental education or family income are far more significant than growing up in a single-parent family or having parents whose primary language is not English (Scott-Jones 1996). Multivariate analyses of the ECLS-K data should help evaluate these positions.

The results also show that the risk factors have no or relatively slight negative associations with children's physical growth or gross motor development. What these results suggest is that the health conditions affecting at-risk children are more apt to be developmental and emotional, rather than physical.

Many of the children with multiple risk factors have attended Head Start or prekindergarten programs. How does participation in these programs affect the early achievement and behavior of at-risk children? This is a question that remains to be investigated with the ECLS-K data and that can be better addressed by the companion birth cohort study to the ECLS-K. The results of such analyses will assist researchers and policymakers in determining whether such programs as Head Start and prekindergarten have their intended effects and what can be done to improve children's preparation for school.

Although many children from multiple-risk families lag behind their classmates in early academic skills, some can overcome the odds and perform at advanced levels when entering kindergarten. This finding seems to argue against stereotyping children from educationally disadvantaged families and assuming that they are all behind when they begin school. Education researchers can examine these children further to understand better the individual, family, and preschool program factors that are associated with such high performance in the face of adversity.

American children show considerable variation in skills and knowledge as they enter kindergarten. The ECLS-K results demonstrate that children are neither alike at school entry, nor ready to be stretched and molded by the varying qualities and demands of different kindergarten programs. In other words, for kindergartners, one size does not fit all. How do kindergarten programs and teachers meet the instructional needs of children whose skills far exceed or greatly lag behind those of the average child? The ECLS-K data will provide a rich and detailed profile of the progress of groups of children who enter kindergarten at different levels.

It is common to attribute the achievement difficulties that educationally disadvantaged children experience in elementary and secondary school to the inferior schools that they are required to attend (e.g., Kozol 1991). What the ECLS-K shows is that these difficulties cannot be attributed solely to bad schools: many children are already behind when they open the classroom door. Does kindergarten help disadvantaged children catch up to other children? If so, does it do so at the expense of children who come to school with more advanced skills? Do the advanced children just mark time while the class reviews things that they already know? Or are kindergarten teachers able to work effectively with children at different skill levels? These are important questions that researchers will address with data from future rounds of the ECLS-K.