Files
Abstract
Falls are the leading cause of fatal or non-fatal injuries among older adults. Evidence-based falls prevention programs are delivered nationwide for years. This study was motivated by the analysis on the falls efficacy from falls data collected by the Administration for Community Living (ACL). In this study, a large number of missing values in ACL-falls data lead to potentially biased statistical results and make it challenging to implement reliable variable selection. Multiple imputation is used for dealing with missing values. In order to obtain a consistent result of variable selection on multiply-imputed datasets, multiple imputation-stepwise regression (MI-stepwise) and multiple imputation-least absolute shrinkage and selection operator (MI-LASSO) methods are used. Simulation studies concerning missing at random (MAR) and missing not at random (MNAR) mechanisms, various signal-to-noise ratios, missing structures, and sample sizes are conducted to compare the performances of MI-stepwise and MI-LASSO.