CLOSE MENU
February’s Epi-Bios Forum will feature presentations from CUNY SPH Distinguished Professor, Mary Schooling, and PhD Candidate, Rehana Rasul.
Dr. Schooling will discuss; The joys of Type 2 selection bias. The talk will explain, with examples, the newly defined type 2 selection bias and potential uses. It will also illustrate the connection of type 2 selection bias to related concepts.
Rasul will present; The internal and net-external bias in Mendelian randomization (MR) studies. MR has been considered nature’s randomized controlled trial by using genetic variants as instrumental variables.
This is a hybrid event. You can join the discussion in-person, in room 707, or virtually by using the Teams Meeting link.
Keynote Speaker: Mary Schooling
Title: The joys of Type 2 selection bias
Abstract: Selection bias has recently been reconceived as having two distinct types. Type 1 selection bias is well-known as biasing the causal estimate in the analytic sample. The newly defined type 2 selection bias affects the generalizability or transportability of the effect measured in the analytic sample to a particular target population.
This talk explains, with examples, the newly defined type 2 selection bias and gives potential uses of type 2 selection bias. It also illustrates the connection of type 2 selection bias to related concepts, such as the classification of effect measure modification by Vanderweele and Robins and the use of Pearl’s selection diagrams to facilitate explication of underlying causal structures, including mechanisms, so as to complement rigorous identification of causal effects with aetiological insight.
The talk will show why type 2 selection bias is such a useful concept and how it helps explicate and interpret findings from different types of epidemiologic studies.
Speaker Bio: Mary Schooling began her career in public health in 2002 as a part-time teaching assistant at the University of Hong Kong, after earning a PhD in Epidemiology from University College London and an earlier career in technology and operations research at IBM. She joined CUNY in 2010 and has been a Professor at the CUNY School of Public Health since 2013.
Dr. Schooling is an Editorial Board member of the journal PLoS ONE, an Associate Editor of the Journal of Epidemiology and Community Health (BMJ Publishing Group), and an Advisory Editor for Social Science and Medicine. Her research program assessing the health effects of lifespan trading off against growth and reproduction crosses traditional boundaries of individual disciplines or fields of enquiry and has yielded several translatable mechanistic insights.
Student Speaker: Rehana Rasul
Title: Internal and net-external bias in Mendelian randomization studies
Abstract: Mendelian randomization (MR) has been considered nature’s randomized controlled trial by using genetic variants as instrumental variables. This design is robust to unmeasured confounding and reverse causation but may be subject to selection bias. We show how net-external bias and internal bias can inform the interpretation of Mendelian randomization (MR) studies and demonstrate how this is different in an RCT. These biases should be considered as potential reasons for different findings from MR and RCT study designs.
Speaker Bio: Rehana Rasul is a PhD candidate in the Epidemiology department at CUNY Graduate School of Public Health and Health Policy. She also conducted research with the Haiti Cardiovascular Disease Cohort Study through CUNY ISPH. She is currently applying her academic and research training to study causes of cardiometabolic diseases for her dissertation. She is an adjunct assistant professor at Hofstra University and previously worked as a biostatistician at Northwell Health.
