Course: Didactic Session: Missing Data


Course: Didactic Session: Missing Data

CHIMb.ca
[img]

Course: Didactic Session: Missing Data

Team: Events, Training & News

Date: This is not a timed event.

Description


The goal of this short course is to provide students with a basic knowledge of the potential implications of missing data on their data analyses as well as potential solutions.

The course begins with a discussion on different types of mechanisms that can generate missing data. This will lay the groundwork for discussions of what types of missing data scenarios can be accommodated by each missing data method discussed subsequently. Simple missing data procedures such as list-wise deletion and last observation carried forward method will be described next as well as the problems they can create in terms of bias and loss of efficiency. Furthermore, the course will include an exploration of slightly more complicated procedures, such as likelihood based approach, mean imputation, multiple imputations and the assumptions required for valid inferences for each. More detailed focus will be spent on implementation of the procedure MI package in SAS.  

The course will end with a discussion of missing data mechanisms that are not missing at random (NMAR) and how to address these concerns as you prepare your own results for publication. This will include a review the newest guidelines from the US Food and Drug Administration and National Academy of Sciences on the prevention and treatment of missing data in clinical trials (N Engl J Med 367;14 2012).

Session 1:
Monday, November 17, 2014
Room 214/218- Basic Medical Sciences Building
(730 William Ave)
1:00 p.m.- 3:00 p.m.

Session 2:
Tuesday, November 18
RM 231- Paterson Computer Lab, NJM
Library (727 McDermot Ave)
1:00 p.m. - 2:00 p.m.


Registration Fees: $40
Instructor: Rashid Ahmed

Click here to register

Click here for the syllabus