Executive Function Development in Preschool Children

DESCRIPTION (provided by applicant): Executive control is a significant cause or contributor to poor outcome in a host of neurological, psychiatric, and developmental disorders with origins in the preschool period, yet little is known about the nature or development of executive control in young children, or how to adequately measure such abilities in this age range. To address this problem, Miller and Cohen's comprehensive, yet parsimonious, context representation model of executive control was used. Because the level of analysis here is neuropsychological test performance, classic psychometric test theory was utilized to operationalize these environmental context representations as the latent task demands that are represented by the true score or common factor variance shared across tasks, i.e., latent EF constructs or factors. These latent EF constructs (labeled Working Memory, Inhibition, and Flexible Shifting factors) are separable, yet highly interrelated; and symbolize the unobservable representations of the larger environmental context that invoke executive control to produce regulated thought and behavior. A cohort-sequential, longitudinal design is used to test the utility of this postulated model. A large longitudinal cohort of 3-year-old children, stratified by sex, will be recruited (using a comprehensive strategy to assure diversity) and administered an executive test battery every 9 months through age 6-0 years, to adequately model skill development during this period of rapid brain and cognitive growth. Four smaller, sequential cohorts with staggered ages of entry also will be recruited and followed subsequently at 9 month intervals until age 6-0 years. The cohort-sequential design permits simultaneous longitudinal modeling of cognitive development, while assessing directly the impact of repeated test administration. The latent EF constructs are defined a priori by multiple tests that are designed specifically for use in young children. Sophisticated statistical analyses are used to a) contrast the fit of the hypothesized structural model to simpler and alternative models developed from test specific characteristics that have confounded previous investigations, b) determine development with age by examining individual and group differences in the level [intercept] and rates [slope] of change, and c) characterize dynamic relations to everyday behavior. Because of the potential application in clinical contexts, changes in individual executive test performance also will be examined with hierarchical growth models, and growth mixture modeling will be used to identify latent groups of children whose developmental trajectories differ from peers that may relate to parental ratings of everyday behavior. Understanding the nature and development of executive control, and its relation to everyday behavior, ultimately will yield important information regarding the processes that may be impaired in a number of neurodevelopmental disorders.