Statistical Model Development in Epidemiological Research

Utilizing all SAS"s statistical and data manipulation power; Using SAS to interface with data sources -- Excel, Access, Oracle, DBII, Teradata, CSV, directly entered data; Building point and click to select data, read into application, save to output. Applying basic statistical model -- logistic, general linear, conditional logistic, ANOVA, MANCOVA, repeated measurements, survival/Poisson analysis , descriptive analysis.; Approaching complex statistical methods -- Propensity score (applied via matching, stratification, regression adjustment), instrumental variables, doubly robust estimation, local control to handle bias correction. For missing values, applying Baseline Observation Carried Forward, Last Observation Carried Forward, Completer"s Analysis, Multiple Imputation using MCMC, Multi-Imputation under MAR (Missing at Random) and MCAR (Missing Completely at Random) missing mechanisms in nonparametric regression, Mixed Model Repeated Measures, Marginal Structural Model with Inverse Probability of Treatment Weighting. To handle correlated longitudinal data, using innovative models such as mixed Model with appropriate variance-covariance structure to identify and correct the co-linearity; Applying small-area estimation techniques including synthetic method, spatial smoothing, regression to combine small area data. Developing software with features of Speed, flexibility, intuitive and easy; Enabling interact with operating system, and host system application such as SAS, SPSS, S-PLUS, STATA, R, JMP, JAVA, C++, SQL, VB, VBA, HTML.