James V. Zidek FRSC, O.C.

Description of Research Interests

My research has two major themes. The first concerns the issue of how best to use statistical information in decision making and inference. Producing an estimate would be an example of such a decision. I and my co-investigators consider the problem in the abstract. The work takes account of the inevitable uncertainty in the information generated provided by the data. It incorporates the knowledge gained from previous data or from prior experience. And it takes account of the potential consequences (or value) of any specific decision that might be taken.

One specific problem of interest concerns a team or committee of experts mutually selecting a decision when the members of the team (of, say, medical specialists) have different value scales, different information and different levels of uncertainty about their information. A surprising conclusion is that under some circumstances the team is better off not getting additional information. A second problem deals with the incorporation of the relevant information in data obtained for purposes other than those in a current investigation. New methods are being explored.

Environmental health studies provide the second source of my current research topics. Again abstract theory forms the core of the work, with excursions into the implementation of the resulting theory. My initial interest was prompted by the need to develop a statistical design for assessing the negative environmental impact, if any, of wildcat (exploratory) drilling in the Beaufort Sea. That was followed by a search for trends in the acidity of acid rainfall. This led to the development of methods of designing networks for monitoring spatial pollution fields, for interpolating between existing monitoring sites to find out what goes on in between, and finally to methods of smoothing the noise out of the data obtained form such network data. Now in turn, the use of this methodology in applications is being explored. In particular, a current study is exploring air pollution and the effect of human exposure to such pollution, both long and short term. Of special interest are airborne particulates that appear to be associated with respiratory disease and with mortality.