Project Summary: Executive functions are high-level supervisory cognitive skills vital for a child’s social and academic success. Their core components include working memory, inhibitory control and cognitive flexibility. The development of executive functions in children can be adversely affected by many early life risk factors, such as lead exposure and low socioeconomic status. However, there are two difficulties to precisely measuring and modeling executive functions, which hamper our understanding of how risk factors can predict shifts in executive functioning: 1) It is common to use a single child performance task to tap into an executive function component. While this can yield important information, it can also lead to bias, because executive functions can be difficult to measure precisely, since they are interdependent and also rely on non-executive function skills. 2) It can be challenging to model the longitudinal development of executive functions because the set of tasks used to assess executive functions can change as a child grows. Some tasks may be specific to a particular age range and not appropriate for other ages. To address the first difficulty, we will improve the precision of measuring an executive function component by using well-established latent variable approaches which integrate output from multiple tasks. Preliminary analyses demonstrate that this approach can yield improved interpretability of a risk factor’s impact on an executive function component, compared to separate analyses of each task. To address the second difficulty, we propose to use advanced psychometric and item response theory methods to create a longitudinal scale of executive functioning with optimized precision and construct validity that span all the ages of interest. This will allow us to identify how risk factors can predict changes in executive function over time (e.g. developmental trajectory). This proposal leverages data from two large, NIH-funded prospective birth cohorts with extensive longitudinal data and repeated assessments of many executive function tasks, and detailed measures of risk factors (e.g. lead exposure, low socioeconomic status). Open-source, user-friendly software for the developed approaches will be developed, including a web application for non-statistical users. I (Dr. Shelley Liu) am a biostatistician and an Assistant Professor in Population Health Science and Policy at the Icahn School of Medicine at Mount Sinai (ISMMS). ISMMS provides an excellent environment for this K25; all mentors and advisory committee members are based at ISMMS. My training will involve 1) formal graduate level coursework on executive function development, child neuropsychology, item response theory and psychometric scaling; 2) tutorials in neurodevelopment and psychometrics with my mentoring team; and 3) clinical rotations to observe neurobehavioral assessments. This training will complement my existing expertise in biostatistics. I aim to bec...