Multiple three-decision rules for 2/sup k/-factorial simple effects. Technical report No. 8

TL;DRAbstract

One of the problems confronted in the investigation of environmental health and in biological research more generally is the fact that organisms often respond to combinations of treatments in ways which are not predictable from the way they respond to each of these taken separately. One air pollutant may, for example, disable the respiratory tract's self cleaning mechanisms and thereby increase its sensitivity to the effects of other pollutants. It is important that laboratory experiments be able to simultaneously detect the effects of the various possible combinations of treatments in a way that utilizes the data as effectively as possible. This paper deals with the optimal use of data from such experiments when the response variable can be regarded (possibly after undergoing a suitable transformation) as having a normal distribution with mean determined by the combination of treatments it has undergone.

Chat with Paper

AI Agents for this Paper