Upcoming CANNeCTIN Events
Next Biostatistics Methodology Videoconference will be on Friday December 13, 2013 from 2:00-3:30 pm EDT
Please join us for the next CANNeCTIN Biostatistics Methodology Videoconference on Friday December 13th. For information about joining the videoconference from one of the participating sites please click here.
The challenge of simulating high-dimensional data in methodology research
Tibor Schuster, McGill University
Methodology research in statistics and epidemiology commonly involves large-scaled Monte-Carlo simulation studies assessing the performance of different data analysis approaches. However, particularly in high-dimensional data settings, as common in pharmacoepidemiological studies, simulation of realistic data scenarios is not straight forward. In this presentation, I will demonstrate considerable issues in the simulation of high-dimensional data. In the context of parametric modeling, desirable conditions for the generated data and model parameters will be defined. The implementation of such conditions is demonstrated using a simulated observational study evaluating the performance of different treatment effect estimators. Using the statistical software R, helpful features for initating and monitoring simulation process are presented.
The Global Summit on Combination Polypharmacy for Cardiovascular Disease was held September 25-26, 2012 at The Population Health Research Institute in Hamilton, Ontario, Canada. Please click here for slides and audio from the Summit.
The 2012 Sensible Guidelines for the Conduct of Clinical Trials was held May 24-25, 2012 in Toronto, Ontario, Canada. Please click here for slides and audio from the Symposium.