Dr. Ma’s research interests in statistical methodology center on meta-analysis, causal inferences, complex survey analysis, longitudinal data analysis, machine learning, and missing data problems. He and his collaborators have used nationally representative database, such as HCUP databases, to study outcomes, trends, and disparities in major surgery. The important research findings of these studies were among the featured research activities of the AHRQ (e.g., Memtsoudis et al., Anesthesia & Analgesia, 2011). He has developed novel non-parametric statistical methods for estimation of higher order moments based on U-statistic. These methods have been applied to the estimation of inter-rater reliability (e.g., Kappa coefficient, concordance correlation coefficient) and correlation (e.g., Kendall’s tau) in longitudinal data settings with missing data (Ma et al., Biometrics, 2008; Ma et al., Psychometrika, 2009; Ma et al.,Statistics in Medicine, 2010). He has also innovatively developed robust meta-analytic methods for multiple outcomes (e.g., Ma et al., Statistics in Medicine, 2011) and rare events (e.g., Ma et al., Commun Stat Simul Comput, 2014). A software package umeta for the use of the meta-analytic method proposed by Drs. Ma and Mazumdar, is available in Stata. Recently, he has devoted extenstive time and effort to develop methods for imputing missing data in nationally representative databases such as the HCUP SID and NIS. In addition to developing innovative statistical methods, he is also interested in publishing tutorial papers (e.g., Detry and Ma, JAMA 2016; Ma et al. Reg Anesth Pain Med, 2012; Ma et al, Journal of Arthroscopic and Related Surgery, 2011) to promote improved use of appropriate statistical methods in cross-disciplinary research projects.
Dr. Ma has extensive experience collaborating with researchers from a broad range of fields in medicine, including anesthesiology, biomechanics, emergency medicine, orthopedics, pediatrics, psychiatry, radiology, and rheumatology. He provides statistical consulting assistance in study design, grant proposal, and data analysis utilizing rigorous statistical methods.