sustained attention during an individual task. Eachmeasurewas administered
at the age of 54 months, before the start of formal schooling. The
outcome measures were test scores on a standardized assessment of
mathematics achievement, administered at the age of 54 months, and
again in grades 1, 3, and 5. In examining the relation between our
main predictors and the outcome,we controlled for several demographic
factors, namely, the child's ethnicity and socioeconomic status,which
was represented by a composite of the mother's level of education and
the family's annual income. Our preliminary analyses also included
the child's gender, which we did not ultimately include in our main
analyses. To examine the relation between our cognitive predictors
and mathematics achievement, while minimizing the contribution of
general knowledge, intelligence, or test-taking skills, we included a
measure of verbal intelligence.