題目: The application of confounders, causal graphs, and propensity scores in health and business research.
主講人: Dr. Yu-Ming Shen (Syneos Health Inc., Germany)
時間: 2018年3月9日上午09:30-11:00
地點: 主樓216
主講人介紹:
Yu-Ming Shen received his bachelor degree (BSc) in Occupational Therapy of Chang Gung University in 2006, his Master degree (MSc) in Epidemiology of Ludwig-Maximilians-Univerist?t München in 2012, and his Doctor of Philosophy degree (PhD) in Epidemiology and Biostatistics from Institut für medizinische Informationsverarbeitung, Biometrie und Epidemiologie of Ludwig-Maximilians-Univerist?t München in 2016.
Currently, He is working at Syneos Health Inc. as a biostatistician. Prior to Syneos Health Inc., he worked for 4 years at the Ludwig-Maximilians-Univerist?t München as a scientific researcher. He was responsible for the statistical analysis in clinical trials and epidemiological & public studies. His specialization was quantitative epidemiology and statistical evaluation of predictive biomarker in personalized medicine. He also had 3 years teaching experiences in study designs, confounding & bias, causality, regression models, diagnosis, and R programming for PhD and MSc students. Additionally, he had 3 years working experience in performing the National Scale Health Survey in Chang Gung University.
In his current role at Syneos Health Inc., he works with five international randomized-controlled trials in drug development phase II & III and is responsible for all biostatistics related activities. His research interest is focused on developing innovative methods to evaluate predictive continuous biomarker that can guide treatment options for the subgroup patients. Currently, he is also a member of The International Society for Clinical Biostatistics.
內容介紹:
Accounting for confounders is particularly important when undertaking surveys in health and business research because it is harder for researchers to control variables in the same ways as they do in experimental studies. If researchers do not consider confounders, the results of their research might not be valid or close to true effect size. In my talk, I will introduce a model of causal path diagram for identifying confounders and tell you how to control confounders by using propensity score.
(承辦：組織與人力資源系，科研與學術交流中心)
