Dr. David Haziza, photo supplied
Every time data is collected, it is virtually certain that one will face the problem of missing values. Missing data is common in clinical trials, where dropout or noncompliance may lead to missing responses for some subjects. To help look at this problem, The Department of Mathematics and Statistics presents a lecture by Dr. David Haziza, Université de Montréal on An introduction to estimation in the presence of missing data, on Monday October 29, 2018 at 12:30pm in Room 1L12.
Missing data also arises in surveys and in administrative files (e.g., medical records). Missing data is undesirable because they make survey estimates vulnerable to nonresponse bias. Estimation procedures based on observed cases only tend to be biased. Also, observing a portion of the data results in a loss of information, which ultimately can lead to point estimates with larger standard errors due to reduced sample size. Haziza will review the concepts of missing completely at random (MCAR)/missing at random (MAR)/missing not at random (MNAR), and describe two approaches for dealing with missing data: weighting and imputation.
Haziza is professor at Université de Montréal, in the Department of Mathematics and Statistics. His research interests include inference in the presence of missing data, resampling methods and robust estimation in the presence of influential units. He is a consultant at Statistics Canada where he spends one day per week. Haziza has already received a number of awards in his career, including the 2018 The Centre de recherches mathématiques (CRM) and the Statistical Society of Canada (SSC) prize in statistics and the 2018 Gertrude Cox Award. He is also a Fellow of the American Statistical Association.
