skip to main content

Title: Rank Overspecified Robust Matrix Recovery: Subgradient Method and Exact Recovery
; ; ; ;
Award ID(s):
Publication Date:
Journal Name:
Advances in Neural Information Processing Systems 34 (NeurIPS 2021)
Sponsoring Org:
National Science Foundation
More Like this
  1. Disasters may have significant and lasting impacts on educational programs and academic achievement, yet the examination of differing patterns of school recovery after disasters is understudied. This paper focused on two aims: (i) identification of school academic recovery trajectories; and (ii) examination of potential risk factors associated with these trajectories. We used latent class growth analysis to identify school academic recovery trajectories for a cohort of 462 Texas public schools that were in the path of Hurricane Ike in 2008. Using Texas Assessment of Knowledge and Skills (TAKS) data from 2005 to 2011, we found that attendance and percent of economically disadvantaged youth emerged as significant risk factors for two identified academic recovery trajectories (High-Stable and Low-Interrupted). Higher levels of economically disadvantaged youth were associated with lower likelihood of falling in the High-Stable trajectory, relative to the Low-Interrupted trajectory. Higher levels of attendance were associated with higher likelihood of membership in the High-Stable trajectory, relative to the Low-Interrupted trajectory. These findings are consistent with the notion that disasters do not affect all people or communities equally. Findings highlight the need for policy initiatives that focus on low performing schools, as these schools are at highest risk for adverse outcomes post-disaster.